1. 1 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
SOCIAL
ANALYTICS
TRAINING
MANUAL
Social Intelligence for Brands
2. 2 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Loved By
The Media Loves Us, Too...
3. 3 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
This guide will walk you through the skills and steps you’ll need for your brand to be successful
at Facebook Marketing. Get ready to learn how to create performance reports on how your campaigns
are doing — not just blindly inserting numbers into a chart, but understanding WHAT worked or didn’t, WHY
that was the case, and what ACTIONS to take based on that insight.
The optimizing of brands on Facebook is a mix on science and art — art for creative inspiration and
science for traffic management and optimization. Together we can leverage your brand’s existing power and
marketing assets by understanding the math of traffic optimization, why consumers connect with brands, and
the underlying analytics rigor of making your selected metrics improve.
Do you have to be a programmer or learn to code? No, but you will learn how traffic flows based on
measurable influence, how to troubleshoot when conversion rates are lower than expected, and how we can
deliver a targeted fan count and revenue figure.
Your success is our success. Ask questions. Together, let’s bring your brand to the next level!
Sincerely,
Dennis Yu
CEO, BlitzMetrics
dennis@blitzmetrics.com
Welcome!
4. 4 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Facebook Training
Section 1 - Facebook Timeline and Wall Management 10
1.1 Facebook Timeline 10
1.2 Facebook Ads 17
1.3 Your Wall and Your Brand 21
Section 2 - Tie Your Profiles and Sites Together 24
Section 3 - The Viral Cycle 26
3.1 The Viral Cycle 26
The Engagement Funnel 27
Targeting 28
3.2 Growth 29
Ad Interest Targeting 30
3.3 Engagement 35
Sponsored Stories 36
Promoted Posts 37
3.4 Conversion 48
11 Killer Ways to Increase Facebook CTR 48
3.5 Optimization 50
Sponsored Results 50
Section 4 - Analytics 52
4.1 What's Your Brand Worth 52
4.2 Engagement 53
4.3 What's the Value of Your Fans? 58
4.4 Measuring Engagement 61
4.5 Social Media Health Check 63
4.6 The Engagement Funnel 64
Section 5 - Ads 65
5.1 How to Run an Effective Facebook Campaign for $5 65
5.2 Ad Targeting to Increase Traffic 71
5.3 Good Facebook Ads Stimulate 73
5.4 Optimization 74
5.5 Bid Optimization 80
5.6 Why B2B Marketing is Facebook’s Best Kept Secret 82
Section 6 - Apps 84
C O N T E N T S
5. 5 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
15 Level Analyst System 89
Section 1 - Intro to Blitz 90
1.1 We Are Performance Marketers 90
1.2 What Makes Blitz Different? 92
1.3 Frequently Asked Questions for New Analysts 95
1.4 Analyst Support 98
Section 2 - Process Overview 99
2.1 How to Write a Strong Process 100
2.2 Creating a Proposal 103
2.3 After Acquiring a New Lead 104
2.4 Email Triggers 110
Section 3 - Employee Relations 116
Section 4 - Qualifying Clients 117
Section 5 - Enterprise Clients 118
Section 6 - Engineering 120
6.1 Process Checklist (for landing page or website) 120
6.2 Process for Listing a New Client in Directories 121
6.3 Automated and Manual Processes, and Systems Maintenance 122
Section7 - Ad Networks 127
7.1 Banner Ad Templates 127
Section 8 - The Blitz Process 129
8.1 New Analyst Welcome Note 129
8.2 Responding to Requests 131
8.3 Communicating With Clients 127
8.4 Most folks who ask for dashboards really need ads 138
8.5 Custom Dollar Sticker 139
8.6 Shredded Currency 141
8.7 Ad Creation 143
8.8 Creating a Dashboard 144
8.9 Custom Dashboard Skins 145
8.10 Custom Skin Implementation 147
8.11 Compiling Testimonials 149
8.12 Setting Up a New Project 150
8.13 Setting Up a New Project Under a Partner Company 151
8.14 Shutting Down a Project 152
Operations Process
6. 6 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Section 1 - Introduction 154
1.1 What is Internet Marketing 154
1.2 Why Use Blitz? 155
1.3 The Secret To Getting This Done 156
1.4 The Secret To Winning 158
1.5 Self-Service For Blitz Users 159
Section 2 - SEO For Beginners 160
2.1 What Is Real And What is Snake Oil 160
2.2 How To Do An Analysis - SEO in 12 Easy Steps 162
2.3 Doing a Solid Website Analysis 164
2.4 How To Write Great Articles To Generate Business For Your Clients 166
2.5 Blog Promotion 168
2.6 Helpful Analytics Tools 169
Section 3 - PPC For Beginners 174
3.1 Make Your Ads Better: Three Powerful Techniques 174
3.2 Optimizing Your PPC Campaign 177
3.3 Further Optimizing Your PPC Campaign 180
3.4 PPC Bid Rules and Optimization 185
3.5 The Largest Player in the Space Buys a PPC Company 188
3.6 Paid Search Automation: Choosing A Vendor 189
The Secrets of Online Marketing
7. 7 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
The Social Impact
Leveraging Your Business 194
Ch. 1: Promoting Yourself in Social Media 194
Ch. 2: Multiply Like a Virus 200
Ch. 3: Anatomy 201
Pinterest 201
YouTube 202
Twitter 203
Instagram 204
Ch. 4: Twitter Friending Rules 205
Ch. 5: Google+ And The +1 Button... Should You Care? 207
Ch. 6: Further Promotion 210
Ch. 7: Advanced Analytics 212
Ch. 8: Why Your Social Analytics are Meaningless 217
8. 8 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Running Your Facebook Marketing Business
Section 1 - Responding to Leads 220
1.1 Follow Up 220
1.2 Not Knowing Info 222
1.3 Students 222
1.4 Agencies 222
1.5 Requesting Insights 223
Section 2 - Ad Accounts 224
Section 3 - How to Target B2B on Facebook 227
Section 4 - Facebook Marketing in the Travel & Leisure Industry 228
Section 5 - Managing Local on Facebook and Google 231
Section 6 - Facebook FAQ's 236
6.1 Getting Started With Facebook Ads FAQ 236
6.2 Conversion Report FAQ 238
6.3 Definitions 241
6.4 Domain Insights 245
6.5 Facebook How To's 246
6.6 Facebook Page vs. Profile 250
Section 7 - Common FAQ's 254
Section 8 - MySQL Notes 256
8.1 Channel-Specific Data 260
8.2 Social Media Data Schema 265
8.3 How to Get Tweets Pulled By Our Tracking 272
8.4 Page Related Queries and Tricks 273
8.5 Fan-Related Queries 275
Section 9 - Common Mistakes 276
Section 10 - Bonus Materials 278
Section 11 - Facebook for Your Further Education 279
About the Author 280
9. 9 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
1 800 485 4958 BlitzMetrics.com Facebook.com/blitzmetrics 718 SW Alder St. Ste 200, Portland, OR 97205
Facebook
Training
10. 10 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
As of March 30th, 2012, all Facebook business pages will be
converted to the timeline profile. Here’s a look at the changes
and how you can best utilize the new layout.
What's New?
1. Facebook Timeline
2. Facebook Ads
3. Admin & Messages
Timeline & Wall Management
SECTION 1 - FACEBOOK TIMELINE
AND WALL MANAGEMENT
11. 11 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Profile Image
• Square image only
• Needs to work alone
• Needs to work well with cover image
Cover Photo
• Always public
• 851x315 pixels
• No calls to action
Cover Images Must Be At Least 399 Pixels Wide and May NOT Contain:
• Price or purchase information, such as "40% off" or "Download it at our website"
• Contact information, such as web address, email, mailing address or other information intended
for your Page's About section
• References to user interface elements, such as Like or Share, or any other Facebook site featured
• Calls to action, such as "get it now" or "tell you friends"
Cover
• No pricing/discounts
• No "like" or "share"
• No contact info
Timeline & Wall Management
1.1 Facebook Timeline
How Timeline is Different
1. Cover
2. Apps
3. Friend Activity
4. Pinned Posts
5. Larger Story Layouts
6. Milestones
• Everyone can preview now.
• All pages were converted to Timeline on March 30th 2012.
12. 12 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Apps
Apps
• "Tabs" are now "Apps"
• Thumbnails can now be edited
• 111x74 pixel thumbnails
• 1 pixel internal border
• New navigation above app
• 810 pixels wide
• NO DEFAULT LANDING TABS
• Photos display by default
• # of Likes can be displayed
• 3 apps decided by page
• Up to 12 total apps in dropdown
Timeline & Wall Management
13. 13 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Friend Activity
Pinned Posts
• Top right of timeline
• # of friends that like a page
• Sample of friends
• Featured friend comments
• Top right of timeline
• Up to 7 days
• Calls to action encouraged
• Can be treated by location
• Social proof
• Shows to fans and non-fans
• Tags not required
Timeline & Wall Management
14. 14 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Larger Story Layouts
• Links
• Photos
• Photo albums
• Video
• Starred posts (highlighted)
• Activity
Timeline & Wall Management
15. 15 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Source: facebook.com/burberry
Edit post to highlight
them (star), change
date, hide, delete, or
pin to top.
Timeline & Wall Management
16. 16 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Milestones
• Similar to highlighted (starred) posts
• Define the event, location, date,
story and image
• Must create as page
• Will create a date link on the
timeline sidebar
Timeline & Wall Management
17. 17 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Offers
• Replaces "Deals"
• In news feed
• Can be a sponsored story
• Redeem via email or mobile
Timeline & Wall Management
1.2 Facebook Ads
18. 18 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
3. Admin and Messages
Timeline & Wall Management
19. 19 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Activity Log
• Click Admin > Manage > Use Activity Log
• All content, mentions, comments included
• Filter by date or story type
• View hidden posts, spam
• Hide, delete, highlight, change date
• Manage content, users
• Create ads
• Share Facebook and email
• Get insights
• Request name change
Timeline & Wall Management
20. 20 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Messages
• Can be turned off by page
• Fans and non-fans can message page
• Reply in admin
• Pages cannot message first
Timeline & Wall Management
21. 21 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
The key to successful brand marketing on Facebook is to produce and maintain engaging
content to keep the Facebook page wall acting as an interactive forum. Facebook, and more
specifically the Facebook Wall, offers a communal forum for brands to reach out and connect with Fans;
the very same customers that support the brand and allow it to flourish. By posting interesting and exciting
content to the Wall of a Facebook page, brands hope to attract new customers and retain existing customers.
Facebook Walls offer an easy outlet for brands to post content that they do not or choose not to include
in more traditional channels of marketing, such as television and print. One easy way to maintain this
engagement is through Wall management. Wall Management consists of three intimately related
aspects: Posting Content, Monitoring Engagement, and Rewarding Engagement.
1. Posting Content
Timeline & Wall Management
Some social mechanics to understand: Brands can increase their impressions by growing their
reach (the number of fans that see their messages) or their frequency (how many times, on
average, a fan sees these messages).
The new timeline for pages heavily emphasizes pictures and open graph actions: Combine pictures and short
text for maximum impact. The ideal post length is between 120 and 129 characters. If it's a question, keep it
shorter, and no more than one post per day should be promotional.
Post length could easily affect the number of
organic impressions. On average, page posts are
157.7 characters, while user posts are 121.5 characters
and mobile posts average 104.9 characters. Posts
between 140 and 159 characters long as, on aver-age,
13.3% less engaging than posts between 120 and 139
characters and 9% less engaging than posts between 100
and 119 characters. This demonstrates that the content
many brands are competing against for a spot in the
newsfeed, is more engaging and readable
(Figure 1.7).
Changes in content type are hurting total
coverage: Some types of content generate more
impressions than other types. For example, status posts
have weaker coverage, on average, at 29.1% for pages
with fewer than 100,000 fans and 54.8% for pages with
more than 100,000 fans (Figure 1.8).
We also see link coverage is decreasing, while
status coverage is increasing. A post with a link in
it is more likely to be a sales-oriented post, especially in
December. If true, then the increase in link posts may very
well tie with engagement dropping in December.
1.3 Your Wall and Your Brand
22. 22 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Any forum will inevitably feature benefactors and detractors, pros and cons. Respond to each
quickly. Facebook Walls act as forums and should therefore be treated no different. For every handful of
users who “Like” or “Comment” on a Wall post, there will be one who is posting negative comments that
may be contextually inappropriate or superficially offensive. It should be the responsibility of Facebook Page
Admins to monitor the Wall to make sure all user interaction is appropriate and not deterring present or future
Fans from supporting the brand.
Monitoring Engagement Responsibilities
• Checking the Wall for negative comments that are detrimental to the image of the Brand.
• Checking the Wall for positive comments that are helping to increase engagement and drive traffic to the
Facebook Page.
• Fast response time.
Being a social network, Facebook is a great marketing tool because it leverages User’s
interests and influence to gain more fans and increase exposure. Facebook Page Admins can
“Like” and “Comment” on Posts as well. It may help to think of Facebook Wall Management as a new media
version of customer service. Responding to productive “Comments” and answering questions posed on the
Wall are great ways to maintain that Brand/Fan engagement that is crucial to a successful Facebook Brand
Page.
Above is an example of a User posing a question regarding a Special Contest that Lane Bryant
is promoting. Lane Bryant quickly and succinctly responds to the query, just like any other form of traditional
customer service. This quick response shows Fans that the brand they invest their own money and time in,
cares about them as a customer and is willing to help answer questions and field concerns. Some brands
have begun to “Reward” Users for their prolonged engagement to their Facebook Page. “Rewarding” high-
engagement users with gift certificates, coupons, or other special offers is a great way for Brands to further
drive traffic and engagement to their Facebook page. Fans who are posting repeatedly to the Wall with
content that is engaging other Fans and actively supporting the brand deserve to be rewarded for helping to
sustain engagement, thus helping drive traffic.
2. Monitoring Engagement
3. Rewarding Engagement
Timeline & Wall Management
Targeted
Reach
Social Reach
23. 23 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Rewarding Engagement Responsibilities
• Rewarding engagement is an extension of monitoring the positive comments that are appearing on the
Wall.
• Rewarding Engagement consists of taking note of the Users who are the most interactive on the page.
Identify active individual Users by noting how often an individual User posts on the wall, “likes” a post, or
“comments” on a post during a span of 30 days.
• If it is determined that an individual User is extremely active on the Brand’s Facebook Wall, this User
should be noted so as to monitor further activity, and be possibly rewarded.
• Once highly active Users have been identified, coordination with the Brand is crucial to determine if
the User is worthy of a coupon or special promotion. Transparency between Blitz and the client is of the
utmost importance, so as to not make false guarantees.
What could be decreasing your organic impressions? For one, post length could easily affect the
number of organic impressions. One average, page posts are 157.7 characters, while user posts are 121.5
characters and mobile posts average 104.9 characters. Posts between 140 and 159 characters long as, on
average, 13.3% less engaging than posts between 120 and 139 characters and 9% less engaging than posts
between 100 and 119 characters. This demonstrates that the content many brands are competing against for
a spot in the news feed, is more engaging and readable (Figure 1.7).
Timeline & Wall Management
24. 24 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
If you’re promoting something you’re already an expert in, tie your blog to your page.
Social media only works when people can see what you have to say, and the more places
you’re able to show it off, the better. There’s a reason why Facebook is the leading social
network on the web – it’s just so versatile. They’ve provided an easy way for you to link your
blog to your pages wall. Check out these steps:
1. Log into Facebook, and go to the main Notes
application page by typing “Notes” into the
search box at the top of the screen.
2. On the left-hand side is Notes Settings,
click the "Edit import setting" link.
3. For the Web URL field, enter the RSS/Atom feed address for
your blog, agree to the terms, then click "Start Importing".
Profiles & Sites
SECTION 2 - TIE YOUR PROFILES
AND SITES TOGETHER
25. 25 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
4. Make sure that the feed imported correctly, and click "Confirm Import".
See? Not hard at all. It’s easy to get the word out about things like Michael Jackson’s funeral arrangements
and free funeral advice. Now your Facebook page is automatically kept up to date with your blog posts!
Cross promote your profiles on all networks, chances are your friends use other social
networks as well. Think of each connection as another possible source for people to see what
you’re talking about. Don’t just think about multiplying by yourself – There’s a reason why it’s
called Viral Marketing.
If you find yourself writing an email that could just as easily be turned into a blog post, why not do it? We
have a couple folks that format blog posts, too. My blog posts take only a few minutes to write. I just write
an email and then they format it, look for funny pictures, stick in links, and so forth. Thus, when you might
be thinking - “I don’t have an hour to mess with writing a blog post”, instead, you say “Oh, ten minutes? No
problem - let me crank one out real quick.” Consider the bang for the buck that is possible.
CONCLUSION:
It pays to reduce the time and effort required when you want to get a message out.
Profiles & Sites
26. 26 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
3.1 The Viral Cycle
Social ads are about amplifying user actions, not broadcasting creative.
When you amplify your own content, you are able to activate fans that may have joined the
page a long time ago, but haven't heard from you since, plus increase the feedback rate on
content (likes + comments).
People make decisions based on the trust of their friends. It doesn't matter what the marketing channel is.
What Facebook has done is create a layer — a social layer that exposes those relationships that have always
been there, between people's real friends. Because those relationships are visible, think of all these dots with
lines connecting them. This created an advertising system that sits on top of that and allows you to inject your
content.
Now let's say your content is testimonials, specials, videos of procedures, whatever kinds of content might
be interesting. Identify who your best customers are; not people who could be customers, but people who
already love you. Amplify what you already have that's great. That's what Facebook is; it's a word of mouth
amplifier.
SECTION 3 - THE VIRAL CYCLE
The Viral Cycle
27. 27 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Manage against the stages of the
engagement funnel and ads, apps, analytics.
T H E E N G A G E M E N T F U N N E L
If you don’t have a goal, then all roads
look as good as any other.
To say you want “greater awareness” or
“deeper engagement” is not enough. You
need to have quantifiable goals based on a
reasonable set of short, medium, and long-
term targets. But that requires understanding a
few things.
• What IS the right number of fans to have?
• What is a healthy level of engagement?
• How am I doing versus my competitors?
• If I’m a direct marketing company, can
I really drive revenue, and if so — how
much and how does this play with my
other channels?
Top Funnel References
1. Adidas
2. ComScore Power of the Like 2
3. AmEx Small Biz Saturday
4. Kia
5. O2
6. Sony Pictures
The Viral Cycle
28. 28 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Define Your Audiences by Value and Size:
T A R G E T I N G
VALUE
HIGH
LOW
SIZE
LOW HIGH
• Fans of
competitors
• Page post
ads to fans
• Broad
category
targets
• Demographic
targeting
Geo-targeted
traditional ads
• Page like ads
• The media
• Interest
targets
• Promoted
posts
• Friend of fan targets
Social media success is about pinpoint precision targets. We’re simulating
the one-on-one conversations that friends have among themselves.
• Workplace
and job titles
• Promoted results
• Mobile placements
• B2B targets
The Viral Cycle
29. 29 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
3.2 Growth
Social Analytics Is About Context
Traditional web analytics has always been about just
your site — your traffic, which focused on number of
hits, inbound referrals, conversion rates, and all kinds
of activities that occur just in your patch of land. But the
majority of social interactions occur in public — on ground
that is not exclusively yours.
A user’s news feed is a shared space where friends, brands, events,
and other objects interact. You don’t own them, you can’t easily tag
them, and you can’t barge into the conversation and force them to
hang out only at your place so you can cookie them.
Don’t be “that guy” who has poor social etiquette. Brand marketers are like bums with cardboard signs that
walk into strangers’ houses pushing their products and asking for money. If you are reporting on your social
channels just like your paid search channels, odds are this is you.
Paid search assumes that people are ready to buy — they had to initiate a search to demonstrate intent. In
social, there is no such thing. What gets measured gets managed. Or another way to say it is, “if you have a
hammer, everything looks like a nail”.
The metrics in social should
be around share of voice,
engagement levels, fan quality,
influence, and the lift to other
conversion channels.
DIFFERENT ANGLES TO MEASURE ENGAGEMENT
People Talking About This (PTAT): Total number of people who have had a brand-
related action show up in their news feed.
Coverage: Total number of impressions per 1,000 fans.
Unique Reach: Total number of people seeing impressions.
Engagement Rate: Percent of people who saw a post who commented or liked.
The Viral Cycle
30. 30 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Laser target ads down to specific interest targets
(including geographic locations).
Facebook rewards high relevancy by discounting the price you pay for traffic.
We know that relevancy leads to high CTR, leading to a lower CPC, which leads, in turn,
to a lower cost per fan and lower cost per conversion. Throughout a campaign, test
different interest targets to optimize effectiveness.
Interest Targeting Count
Internet 2,718,000
Computer 1,959,060
Technology 1,178,760
Laptop 906,720
Competitor Targeting Count
Google 2,861,280
Yahoo! 2,198,180
AOL 453,980
Skype 206,440
Lateral Targeting Count
Blog 2,081,280
Online Chat 1,269,080
Newspaper 491,640
Instant Messaging 39,180
Literal Targeting:
What are the direct interests
related to your brand?
Competitive Targeting:
Who are your competitors?
Lateral Targeting:
Who are these people? What kind
of things are they interested in?
PRO TIP: Ego ads
Create one page post ad that is targeted just at folks who like your company, plus one page post
ad against those who are fans of your page but live in your headquarters city. These are “ego
ads” to appease your colleagues and the executives.
For those of us who run Google PPC ads, this is the “ego campaign” for pesky executives that want
to make sure that we’re “showing up” on Google for a set of keywords, even if those keywords are
unprofitable.
A D I N T E R E S T T A R G E T I N G
The Viral Cycle
31. 31 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Micro-Targeted Ads
How much should you spend? Depends on your goals.
We take a 3 layered approach: to grow your audience, engage them,
and convert them.
Your Facebook ad budget should be 15-20% of your total digital budget and
mirror the campaign structures of other marketing channels.
Fan Acquisition Cost is number of fans x cost per fan
Fan Nurturing (engagement) is fan base x 0.5 cents per fan per month
Conversion Cost: target CPA x number of conversions required
TOTAL:
MONTHLY BUDGET
Campaigns drive conversions or sales.
Always on and visually represents about 20% of
the budget or growing slightly. Fan engagement
budgets are directly proportionate to how many
fans we have to nurture.
Fan Engagement
In spurts but still always on. The fan
growth phase is about 50% when on
and down to 7% when in lulls.
Fan Growth Campaigns
January February March April May June July August September October November December
Fall Contest
Ad
Spend
Summer Promotion
Campaigns drive conversions or sales.
Always on and visually represents about 20% of
the budget or growing slightly. Fan engagement
budgets are directly proportionate to how many
fans we have to nurture.
Fan Engagement
In spurts but still always on. The fan
growth phase is about 50% when on
and down to 7% when in lulls.
Fan Growth Campaigns
January February March April May June July August September October November December
Fall Contest
Ad
Spend
Summer Promotion
The Viral Cycle
32. 32 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Most of the pages you come across suffer from terrible engagement
rates. They eagerly grew a lot of fans in 2010 but forgot about
what to do with these fans. Allowed to go stale, these fans never see the
brand again. The brand may post to the wall multiple times per day, but not
have high enough EdgeRank score to qualify for showing up in the News Feed
(the user’s home page). The solution is to use the ads platform to reconnect
with these fans and to run ads so targeted, even non-fans will respond.
Ad Targeting to Increase Feedback
PRO TIP: Always run
combos to amplify
follow-on engagement
Don’t make the mistake
of going for pure
audience growth and
then letting your new
fans die. Conversely,
don’t go for conversion
right out the gate without
having engagement ads.
That results in one-off
conversions at high cost
without building social
community or the word
of mouth effect.
The Viral Cycle
33. 33 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Interests
Media & Press
Workplaces & Job Titles
Interest and Group Targeting
Your ads should always be targeted, else they risk not seeming both relevant and
personal.
The Viral Cycle
34. 34 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
We recommend 4 blocks in your ad name:
<client_name>, <type_of_ad>, <targeting>, <creative_description.
Facebook doesn’t provide previews on targeting in Power Editor or
the ad interface. Plus, when you download your ad performance,
you’ll want to be able to create pivot tables to group performance by
any of these attributes.
For example, let’s say we’re running an ad on behalf of Rosetta Stone, that it’s a Sponsored Result ad,
we’re targeting folks who like the BBC and are friends of fans, and we’re sending to the 1 million fan
celebration landing page tab.
We’d name it:
rosetta_sr_bbc-fof_1mm_celebration
Or maybe we’re the Treasure Island casino running ads to enhance our post visibility against fans of the
Wynn, Venetian, Palazzo, and other properties.
We’d name it:
ti_ppa_competitors_ongoing
We like to use underscores to separate each variable, so it’s easy to import into Excel or your favorite
tool. If you have multiple items in a block, use hyphens (treasure-island, for example), so you don’t mess
up your importing.
For ad type, use these:
sr: sponsored result (search ads)
ppa: page post ads
pl: page like ads
ppl: page post like ads
checkin: check-in ads
app-used: app used ads
trad: traditional ads (sending off-site, requiring a headline, image, body, and url)
Ad Naming Conventions
PRO TIP: Set up one campaign per ad
Normally, you’d group all ads by objective and have them live in a campaign. However,
Facebook’s Quality Score algorithm is not as sophisticated. So by having one ad per campaign,
you get greater unique reach, a higher CTR, and a lower CPC.
When Facebook introduces ad rotation and is smarter about eCPM estimation, you’ll be able to
treat social campaigns like ad groups in traditional CPC. But for now, you’ll have to deal with
having many campaigns, so have a smart naming convention
The Viral Cycle
35. 35 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
If you're an administrator of your brand's Facebook page, you can access Insights.
Here are 2 key metrics to determine your engagement:
What percentage of your fans
are interacting with you?
Is it under 1%? If so, then your fans
are saying your messages are not
interesting. Facebook, noticing this,
will also not share your messages with
friends — a double whammy.
Improve your score here by asking
questions as opposed to making
statements, keeping messages
10 words long or less (versus a
paragraph), including multimedia
versus just text, and having content
that triggers emotion. Above all, resist
the temptation to sell too hard.
The wall is where most engagement
occurs, although you can boost
your engagement rate by running
ads to your postings, plus including
engagement apps.
Interactions per week looks at how many interactions you have.
Underpost and you have no community. Overpost and you wear users out. Your ideal post frequency is
dependent upon what your users expect.
If you’re a news site, then 10 times a day can be okay. If you’re a B2B software company, a few times a week
is probably fine. Do not cheat by auto-posting your RSS and Twitter entries — fans will frown upon non-
social, automated content. Don’t promote your specials or discounts more than once in every 4 postings, and
definitely no more than once per day.
If you are getting less than 10% of your fan base as interactions each week, you’re not unlocking the value of
your community. Thus, if you have 100,000 fans, you should have at least 10,000 interactions each week.
1. Feedback Rate
2. Interactions Per Week
Measuring Engagement
3.3 Engagement
The Viral Cycle
36. 36 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Ensure your messages get into the news feed
S P O N S O R E D S T O R I E S
Amplify your organic content and positive fan behavior via Sponsored Stories.
The Viral Cycle
Sponsored Story - Page Likes Sponsored Story - Page Posts
Your message
amplified
Friends who’ve
liked your Page
are displayed.
Your message
amplified
The Page can
be liked right
from the ad.
Your message
amplified
Your latest Page
post is displayed as
a Sponsored Story.
37. 37 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Spend $7 to promote your personal status to friends
P R O M O T E D P O S T S
Pay to get your posts more visibility in the news feed.
The Viral Cycle
38. 38 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
We can target contextually
On posts with high engagement, amplify whenever possible.
39. 39 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Social Ads That Engage and Convert
Configure the right mix of Sponsored Story ad units against your chosen goals.
Publishes the latest Page post
as a Sponsored Story.
Displays the user’s friends who
have already liked your page.
In the last month your App or
game was used by someone at
least twice or for 10 minutes.
In the last week someone used
Facebook Places to check in
and/or claimed a deal at one
of your claimed Places.
Your page was liked either on
Facebook directly or on a Like
Box on a separate site.
Friends of your fans.
All of the people who
have liked your Page.
Friends of your fans who
have liked your posts.
Friends of the people
who have used the App
or played the game.
Friends of people who
checked in or claimed a
Deal.
Story Type Story Content Who Sees It
Page Like
Page Post Like
Page Post Story
App Uses /
Game Played
Check-In
The Viral Cycle
40. 40 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Current Ad: Suggested Revision:
Note that I didn’t mention in the ad that they’re a dog lover. I believe that can safely be assumed because of
the image. In some cases, where the image is not able to represent the interest, we do want to explicitly call
out the interest target.
Using your Merchant Name in the ad is a double-edged sword. In this case, I believe we shouldn’t use it
because the merchant doesn’t have brand awareness. Remember, to get a click from people who don’t know
us, we must establish trust quickly — and no brand is no trust. Using the long brand name also lengthens the
ad copy and makes it slightly harder to understand the value proposition.
Social media requires the lowest of conversion hurdles (the minimum of thought or consideration). Since
people are on Facebook to socialize, I advocate the two-step conversion for most clients — one set of ads
to get the right interest targets to be fans, then another set of ads targeted just at fans. True, it costs more to
run more ads, but the prices are so low on Facebook ads that we can get away with this. Plus, depending on
the nature of the product or service — hitting people at different points in the funnel can be far more effective
than going for a jump all the way through to conversion in one leap.
A good rule of thumb is a 3X improvement on ads when you optimize them. In other words, if your
cost per fan is 60 cents when you test the first couple days, you should be able to get it down to 20 cents on
average. The 3X rule is what we’ve discovered is reasonably possible with a moderate amount of testing - of
course, you could get far more.
* Liking as a condition of entry is not allowed, but many brands do it. Proceed at your own risk.
Ad Copy Techniques
Determining Who Is The Winner
If we have 3 images - which of them performs better?
By “perform”, I mean by the goal we’re being measured on, as well as early indicators (CTR and CPC). If we
have 3 interest targets, how do they perform? Or maybe we have a few age/sex/location targets. Unless we
have multiple variations of each ad multiplied, it will be hard to isolate a winner and isolate the performance
of a single variable.
The Viral Cycle
41. 41 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Geographic Split Test
FOF (Friend of Fan) and
Non-FOF Ads Split Test
Run one ad that is not FOF and another that is, just
so we can get an apples-to-apples comparison. In
other words, the ads are identical except for this
targeting change. This is called an A/B test and we
want to determine the differences in CTR, CPC, etc...
Don’t let it run more than a couple hundred dollars
in total — or enough to have a significant difference.
The Viral Cycle
42. 42 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Demographic Optimization
If you have an interest target that is already highly skewed towards a particular
demographic (combination of age, and gender), then it’s not necessary to also filter by
demographics, too. For example, if you are targeting New England Patriots fans (folks who
like a professional football team), it’s not necessary to specify males as a demographic. If
you do, that will cut out 20% of the audience, as there are still some females that will like
this.
Therefore, when setting up initial ads that have interest targets, demo targeting is usually
not necessary. But in subsequent refinement, you might want to multiply into combinations
of these demographics. The Demographic reports in Facebook’s ad tool will tell us that the
CTR or conversion rate may differ or not.
PRO TIP: Combination targeting
If you want to give moderately performing ads a boost and if the existing target audience is large,
use combination targeting. Duplicate the existing ad (use Power Editor if you have more than 3 or
4 ads to duplicate) and add in another filter for demographics, cut down the number of interests
you are targeting — or my favorite, use specific and broad category targeting together.
Often an audience of 200,000 that goes down to 20,000 will double your CTR. And adding in
irrelevant interest targeting of any sort will often give you a CTR boost, since the fact that these
users like anything means they’re more likely to like you. People who like pages have almost 300
friends, while the average Facebook user has 130 friends. So you get a boost in influence, too.
Fan
Targeting
Non-Fan
Targeting
Friend of Fan
Targeting
Your friend is attending,
how about you?
Segmentation
The Viral Cycle
43. 43 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Brand Logo On The Ad Image
Interest Targeting - Friends of Fans Targeting 1
Cold Targeting vs.
Fans Targeting
For many large brands, often the highest CTR
and lowest CPC/CPF is the brand image
itself. Thus, if we could make the logo larger,
it could increase the CTR. Maybe even try
a colored border, provided that the client
approves this.
When you do interest targeting, you don’t always have to put in a gender filter, too. For example, if you
specify “cooking”, you don’t have to necessarily also place “female” as a gender target. There are some men
who like cooking as well. To see how the counts change, try the targeting in the ads interface with and without
female targeting — you’ll see only a minor difference. The reason is that most of the folks who like cooking
are female.
Average CTR of ads to non-fans: .05%
Average CTR of ads to fans: .35%
Interest Targeting - Friends of Fans Targeting 2
Have you ever considered doing pure
FOF targeting in cities/states where
your brand is strong?
For instance, let’s say you’re running
a campaign for a fast food chain in
California cities since the location
density is much lower than the rest,
say cities in New Mexico. The lower
the people/location figure, the better
we’ll do, as the people seeing the
ads are more likely to have a location
near them. In addition, online ads
should follow offline popularity as a
tactic to increase conversion.
The Viral Cycle
44. 44 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Interest Targeting - Friends of Fans Targeting 3
Ad Distribution
Choose interest that
has at least 1,000
population reach. When
doing FOF, the estimated
reach should be at least
10,000 but even if only
500, it’s worth doing,
you’ll just have to create
a LOT of them.
3 Types of Audience Reach:
Targeted: Based on the targeting you’ve selected, this is an
approximate number of people your ads and Sponsored Stories
can reach.
Reach: The number of individual people who saw your ads or
Sponsored Stories.
Social Reach: The number of people who saw your ad or
Sponsored Story with the names of their friends who liked your
Page, RSVPed to your event, or used your app. If you are running
ads off Facebook, you won’t be able to see the social reach.
Ad Multiplication
Try our top ads with FOF Like cooking?,
Like [interest]?
For example, target “food network”
and say “Like Cooking Shows?” Target
“Bobby Flay” and say “Like Grilling?”
If you create enough of them, you can
do FOF targets, too FOF significantly
limits reach, of course, but the trade-off
is better CTR and conversion.
Learn how to put your Facebook sponsored stories on steroids!
http://www.allfacebook.com/facebook-sponsored-likes-2012-02
The Viral Cycle
45. 45 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Ad Burn Out
It may be that if you run an ad that was semiburned out, but gave it time to “rest”, it
gets a light refresh.
A good predictor of ad burnout is not just the trendline for CTR, CPC, and CPF, but
also looking at the audience size each ad targets. The smaller the base, the fewer
number of impressions needed
to burn out.
I’m saying that if the audience
base is small, it takes fewer
impressions total to burn
out. For example, an ad that
targets 100k people can
serve a lot more impressions
before burnout than one with
only 1k people in the target.
Smaller audiences might not
burn out faster in terms of number of days, since they will also show fewer impressions per day. Therefore, in
estimating future traffic and our ability to hit fan targets at CPF goals, we should account for audience size.
Assume we can hit users 20 times before burnout on average. So if the audience is 100k people, we can
serve 2mm impressions.
Audience size matters especially in local. If some campaign wants 100k fans in just Atlanta vs nationwide,
it will be a lot more expensive because we hit burnout faster - there are fewer people to target. We get
saturated faster. In large multiplied local campaigns, whether for a franchise or a single location, we have to
consider the bounds of each geo-target. This is true is regular PPC, too. ReachLocal has a traffic eliminator
built into their proposal generator, which I’ve shared a few times. Buggy - a good concept, but poorly
implemented.
PRO TIP: Max the bid on small audiences
If you have an audience less than 1,000 people (maybe 50,000 people for large advertisers), you
sometimes have to force the bid to get traffic. Facebook’s eCPM estimator isn’t good at estimating
which advertiser’s ad will perform, so the high bid sends a strong signal. We will often bid $10
a click, knowing that the price is often going to be pennies. This is especially useful when doing
competitive targeting (where you know the traffic is high quality) and workplace targeted ads (where
the audience is highly valuable).
Facebook doesn’t reveal average position of your ads yet, so a top bid gets the highest placement
and higher CTR, too.
The Viral Cycle
46. 46 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
If you are bidding CPM and care about CTR, then never allow your bid to be below the
suggested bid. If you do, Facebook will certainly give you impressions, but against the
lowest quality audiences at the bottom of the page.
If you bid CPC, Facebook will seek out which combination of demographics and
position will generate the most clicks, since they’re paid on clicks. If the audience is
small, as mentioned earlier, you should do an extremely high CPM or CPC bid. If you
expect a low CTR, then bid CPC.
Facebook allows you to now automatically bid based on likelihood of a user to
become a fan or click on your story. If you choose this, you cannot select a CPM
or CPC bid. Against audiences over 100,000, we find their algorithm to be highly
effective. But test and see for yourself.
The Latest on Bidding Strategies
The Viral Cycle
47. 47 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Education Level Impacts CTR
Competitive Targeting
Want to know how one small
business saw 603% growth
from effectively using sponsored
stories? Click here.
The Viral Cycle
48. 48 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Like AdWords, in Facebook the #1 factor governing your
CPC is your Click-Through Rate (CTR). While Facebook
doesn’t reveal your ad or keyword Quality Score, you bet
that it makes a huge difference.
1. Ask a question in the ad
Not only will this drive in-line likes (fans), but encourages
participation. Consider asking them if they like you or the interest
you’re targeting — but make sure you’re still relevant or you’ll be
disapproved.
2. Write short ad copy
Sure you can use up all 135 characters in the body and 25 characters in the headline, but odds are that
people won’t read it all. This is Facebook — people are likely not conducting serious business, so make it
light and easy.
3. Use a close-up face in your image
Smiling is better, as well as looking directly at you. This is FACEbook, so use FACEs in your ads.
4. Personalize the image
If they’re a 45 year old white female in the Bible Belt, we’re not going to show an urban teenager rocking
out. People usually convert better when the ad model is closer to them — the exception is dating and beauty
products. Baby products can be like that, too. Whatever the case, test it. We know if they’re married or not
on Facebook and can even guess their race — so that is something you can personalize the image with.
5. Capitalize a couple words
Saying FREE is not okay in AdWords, but we see it all the time in Facebook. You might try it.
6. Use numbers and unusual characters
This works in regular PPC, too. And if you make a claim — don’t say “We can help you save money on
insurance.” Don’t even say “We can save you 15% on your insurance.” You need to be more specific —
“Save 17.3% in just 3 minutes!”
7. Stimulate emotion
”You’ll be sorry. That’s what you’ll say if you miss Portland Honda’s Labor Day blowout sale!” Arouse
curiosity. Message it as if it was a personal friend talking to a personal friend — “Doris, you wouldn’t believe
the sale at Luckys on hotdogs this weekend.”
Relevancy leads to high CTR,
leading to a lower CPC, which
leads, in turn, to a lower cost
per fan and lower cost per
conversion.
11 Killer Ways to Increase Your Facebook CTR
3.4 Conversion
The Viral Cycle
49. 49 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
8. Send users to your Facebook page
This is where custom tabs, especially a reveal tab or engagement app positively rules.
9. Follow through on the promise on the landing page
Almost nobody does right — if you see it done right, please let me know. In PPC, we
know that we must tightly map the keyword to the ad to the landing page. That means
if someone is looking for patio furniture, we don’t say “cheap furniture” and dump
them on our home page. But that’s what most folks do in Facebook — send everyone
to exactly the same page, as opposed to differing landing pages based on what’s
targeted in the interests and ads.
10. Dayparting
We run Facebook campaigns for a number of food companies. I suppose you could promote breakfast foods
at night, but why would you? Consider how time of day may affect the messaging as well as the type of user
you see. For example, if you’re selling cold and flu medication, you might run ads between midnight and 6
am saying “If you were using X, you’d be sound asleep right now.” By the way, dayparting is not a feature in
Facebook yet, so we had to build our own.
11. Fan targeting
Once you have all these fans, you have to keep the conversation going. The fan targeting won’t give you
much volume, but I’ll bet it has the highest CTR of any ads you run. It’s the equivalent of social retargeting.
Read the entire article here.
The Viral Cycle
50. 50 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
S P O N S O R E D R E S U L T S
You might have noticed, when searching on Facebook
for a brand page, that other related companies show up
in the search results.
Facebook rolled out sponsored results in August, for those with ads
application-programming interface or power editor access. Here
are some of the basics you need to know about this new feature:
• The performance is stupidly good
Click-through rates are in the range of 3 percent to 10 percent,
and cost per click is between 1 cent and 10 cents. Thus, you’re
getting the same CTR as newsfeed or mobile ads, but paying
the price of marketplace traffic — those seven ads on the right
side that might command CPMs (cost per thousand views) of
60 cents to $1, depending on your targeting.
• You can pretty much forget about interest targeting
The volumes are so small that if you filter down further by
gender, location, or whatever else, you won’t have any traffic
left. With search, you’re only showing up when someone types
in that page, as opposed to whether they’re a fan or just talking
about it in their status. Even with all of these search terms,
called “targeted entities,” that in total have a couple of million
fans between them, you are only reaching 4,220 people in this
example.
The Viral Cycle
3.5 Optimization
What's the skinny on the new Sponsored Results?
51. 51 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
• Brand bidding is nice
If you’re Coke, you bid on Pepsi. But you could also bid on any product or brand
that is related — McDonald’s french fries, the local sports team, music videos,
or something completely irrelevant. For example, you could set up a script to
automatically geotarget any city that is over 90 degrees today, then choose the
biggest search targets, related or not. Show your 70-character message of “Hot in
<city_name>? How about some refreshment!”
• Perfect for the small guys
If you’re a small brand, you can bid on the well-known brand. It’s a David and Goliath strategy. Just be
aware that people who search in that box are actually navigating, so your message better be persuasively
distracting. I wouldn’t necessarily mention your competitor. Likewise, if you’re the big guy, you better bid
on your own name to protect it, right?
• 70 characters of awesome
Make sure it’s related to the search term. Ask a question. Choose your default landing page — perhaps
your email registration page that has your re-targeting code on it (for those at advanced levels of
sneakiness). Create one ad per target, since creating 100 ads is as easy as copy and paste in power
editor. Why not?
• Bid high
Or just choose “optimized CPM,” which lets Facebook guess what is enough to get the spot. It doesn’t
really matter, since right now, CTRs are super-high for the novelty factor and because few advertisers are
using it. You’re likely to be in the first spot. You’re going to get clicks for pennies. If you’re not getting
traffic for 90 percent off, tell me, since you’re doing something wrong.
Read the entire article here.
The Viral Cycle
52. 52 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
The most common question clients have about Facebook is how to measure ROI or what a
Facebook fan is worth. A great way to demonstrate how our valuation and measurement techniques work,
is to answer these questions.
The value of your brand on Facebook is a function of the number of fans you have multiplied by the value
of your fan. If you have twice as many fans, then your presence is worth twice as much. Likewise, if you can
double the value of a fan, then you double your brands value as well. If you have 50,000 fans and your fans
are worth $10 each, then your Facebook presence is worth half a million dollars.
Let’s start with your fan count.
A common question is, "How many fans should I have on Facebook?" That depends on two factors — how
you rank relative to other companies in your industry, and how much traffic your website receives. Your
BlitzMetrics Fan Page Score shows how you fare against peers in your industry, based on publicly available
data, using the industry classifications of the Fortune 1000. By the way, the average Fortune 1000 company
has 62,141 fans, which includes Starbucks at 13 million and various B2B brands at only a few thousand.
How do you compare?
Facebook, at nearly a billion users, now accounts for about 7% of unique visits on web properties in the
US, which is just slightly above Google. Look in your analytics to see how much traffic you are getting from
Google versus Facebook. If your Facebook traffic is less than Google, it means that you’re not getting
your proportionate share of traffic. Further, because average session time on Facebook is over triple that of
Google, even having your Facebook profile on par with Google, may still mean you’re under optimized.
4.1 What's Your Brand Worth
Analytics
SECTION 4 - ANALYTICS
53. 53 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Here are some of the engagement flags and what they mean in your social scoring report.
Custom Tab:
It is nearly impossible to build
meaningful relationships with customers
via the Wall, which is a streaming flow
of the latest things people have to
say. It’s graffiti. A custom tab allows
you to insert a landing page, play
a video, collect their email address,
create a quiz, buy something, share
with friends, or other activity. To
not have engagement apps is to
disappoint consumers who expect to
be able to interact with you and also
be entertained. It also means that
Facebook’s EdgeRank will effectively
silence you — as lively conversations
are promoted and inactive players are
not seen.
Allow Fans to Post:
The default setting of your page is to show
posts by the page only - you. That means
if a fan wants to post something, it cannot
be easily seen by others. This is the default
setting because most brands are not able to
monitor page activity in the same way they
might staff up their call center or support
desks. To shut out your fans is to tell them
their participation is not welcome. This is
one of the easiest settings to change. But
doing so then requires vigilance to respond
to what fans have to say.
4.2 Engagement
Analytics
54. 54 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
The Like Button
The Like button visually
creates desire
Viral Loop
Shares a link in the newsfeed Discussion increases reach
Analytics
55. 55 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Website "Like" Button
Most brands have a link to their Facebook page from the company
website. But what they often miss is the like button — it's not a
static link, but an intelligent button that tells users whether they
are a fan or not, as well as which of their friends are fans of the
page, too.
Folks who visit your website can become fans of your page without ever
having to go to Facebook - they can click “like” right there. This “instant
socialization” is far more powerful than just pasting in a blue F logo on your
site. Further, the XFBML version of the like button will post user comments
to their wall and show up on friends walls. You can also pass an “interest”
variable in your like buttons and then target ads based on that interest.
Which version of the like button do you have?
Reveal Tab
TARGET
Local Customers
APPS
ADS
A reveal tab is a custom tab that shows whether the user is a fan or not. It shows one type of
content to those who are fans and something else to those who are not. The most common reveal
tabs say “click like to get exclusive discounts”, “click like to watch an exclusive video, “click like to show your
support”, or some other enticement.
We see that conversion rates can be as high as 80% on these custom tabs, since users are now accustomed
to click "like" to get something in return. If you have compelling content, perhaps a funny video or an
attractive coupon, a reveal tab will be quite effective for you. Your fan page value is based upon how many
of these flags are set to true, as well as the level of engagement of your page. The engagement of your
page is how often you post, what share of users are interacting with you, and whether you are able to attract
interaction via a comment or a like.
Analytics
56. 56 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
This reveal tab converted in the 20-50% range - low
relative to fast food, but good for a $500+ considered
software package. What makes this special is how the
reveal tab extends upon the promise of the ad copy;
that you can learn a new language for free. By testing
ads and landing pages together, in the same way that
ad copy and landing page combos work in PPC, clients
have found that a lower conversion rate is worth the
trade-off for higher quality. Using ad imagery that mirrors
that of landing pages gives users a smoother experience,
especially when reinforcing the trademark yellow, in this
particular case.
This brand may have the most fans of any CPG page
because of their emphasis on featuring the fan. Giving
users each their proverbial 5 minutes of fame by being the
“World’s Fan of the Week” is one of the most effective
ways of driving viral likes. The incentives of a free ringtone
and game dynamic drive great engagement. Facebook
recently relaxed their contest rules to no longer require
permission from an account manager or have an ad
budget of at least $10k. Expect more contests in 2011 as
a result.
Lane Bryant is the leading retailer for plus size women’s
clothing. Their Facebook page capitalizes upon their
offline brand strength, leveraging the same imagery
and offers found in their print catalog and in-store
marketing units. The $250 gift card is a sufficiently
strong prize. As JVC demonstrated in their contest,
many small prizes are more effective than one big prize,
because it allows for the contest to build momentum.
Two lessons: make sure your offline and social
presences are in synch, and find ways to elevate your
users.
Rosetta Stone
Lane Bryant
Levi's
Oreo
Elements of Reveal Tabs
The pre-like page should have a blurred out version of that first chart, because the key to a landing page
conversion (whether on Facebook or anywhere on the internet) is to give users a clear visual idea of what they
will get if they take action. The more real the reward and more clearly they can see what they’ll get, the more
likely they will convert.
• Big arrow pointing to the "Like" button.
• Don't put the "Like button image on the tab because people might get confused and click on it.
Examples:
Levi’s has built 2.7 million fans by integrating Facebook
“like” buttons strongly with their site to encourage user
feedback that goes from the site to the Facebook page
and vice-versa. Their page has a comment wall below
the main image, which is effective for national brands
of sufficient size. If you don’t have a brand of significant
recognition (ask 100 random people at a mall and see
if 50 people have heard of your brand) and also have
at least 100,000 fans, the comment wall might not be
effective for you. Judge your social reach and decide
whether you can get away with this. If you’re big enough,
as we see with Carl’s Jr, you might not even need a strong
graphic pointing to the like button.
Analytics
57. 57 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Driving Ads to the Wall Post Increases Fan Engagement
When you amplify your own content, you are able to activate fans that may have joined the
page a long time ago, but haven't heard from you since, plus increase the feedback rate on
content (likes + comments).
A higher engagement rate increases EdgeRank and thus, drives greater News Feed visibility. While Google
has a “church and state” separation between ads and organic, this is not true with Facebook, nor should it be
in social channels. Items that have a higher engagement rate among your friends, whether driven by ads or
not, deserve to be shown to friends of interactors. Who doesn’t want to know what their friends are doing and
what they like?
You can drive ads to a specific wall post by choosing that wall post url or you can just use Sponsored Post
Stories, which does it automatically for you.
Facebook cites 900 million objects (events, pages, community pages, and groups) that users can interact with
on Facebook. In addition, there are 36.2 million user actions per hour showing up in the newsfeed.
Newsfeed Competition:
DIFFERENT ANGLES TO MEASURE ENGAGEMENT
People Talking About (PTA): Total number of people who have had a brand-related
action show up in their news feed.
Coverage: Total number of impressions per 1,000 fans.
Unique Reach: Total number of people seeing impressions.
Engagement Rate: Percent of people who saw a post who commented or liked.
Analytics
58. 58 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
4.3 What's the Value of Your Fans?
The beauty of the e-commerce and lead gen model is that it's simple to measure the ROI of your Facebook
presence. You need only conversion tracking that accounts for the impact of ALL marketing efforts, not just the
last thing a customer did prior to converting.
EARNED MEDIA: If you are a Consumer Packaged Goods brand or one that interacts
with customers via retain channels, then earned media value is most accurate to
determine value.
Sometimes a Facebook presence can
actually hurt your brand. Witness Dell:
when they didn’t tend to their wall, allowing
unhappy customers and ex-employees to
post dozens of comments with no response
from Dell. Just like vandalism, this is visible
publicly to anyone, but it also represents
an opportunity for Dell to demonstrate they
care. Note: since we pointed this out, to
their credit, Dell has taken corrective action.
Analytics
59. 59 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
1. Discounted Cash Flow (DCF)
Discounted Cash Flow is the most revered technique since it’s the closest to real profit and not
subject to stock market manias. Thus, BlitzMetrics can pixel the conversion page of a website to capture
not only the conversion, but whether they are a Facebook fan or not. If so, assign credit. Then calculate what
percentage of conversions are Facebook fans and what the average conversion value is for Facebook fans vs
non Facebook fans.
Now you have a real Facebook fan value by taking checkout value of the Facebook fan times what
percentage of Facebook fans buy (survey randomly on all site users who is a fan or not compared to
purchasers who are fans or not). Indisputable value which can then be plugged into our grading tool to
override the default fan value.
You can measure how many impressions
your Facebook page is generating.
For example, if you are generating 100,000 impressions per day via your wall posts and your cost of media is
a $5 CPM (cost per thousand impressions) in display, then your Facebook page is worth $500 per day. That’s
what you would have to generate the same level of traffic. Arguably, the Facebook traffic is worth more,
because it is trusted, has a higher engagement rate, and allows for sharing. But what if you post only once
per week that few people notice your comments? You might have 1,000,000 fans, but be generating only a
few thousand impressions per day.
In finance, if you want to buy a company, there are 3 primary valuation models you would use: discounted
cash flow, comparables, and net asset value. The first (DCF) is a lifetime of projected earnings discounted
back via an implied interest rate. The second is to say what similar companies are worth and apply their same
p/2 ratio or composite set of metrics- for example, if Google is worth $150 billion and has 60 percent share,
then Yahoo at 5 percent should be worth 1/12 that market cap. The third is liquidation value- if sell off all
that company’s assets and pay off all debts, what are you left with? Each technique is appropriate in the right
situation. For example, you wouldn’t use NAV with software companies, since they don’t have assets (and
goodwill via acquisition doesn’t count).
The same 3 valuation techniques apply to Facebook fan value. Let’s step through them:
Analytics
60. 60 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
3. Asset Value
If you were to sell your list to others, what would that fetch? If you were to sell ad space on
Facebook (not allowed, of course), how much could you generate? If you were to buy the same
amount of exposure via a display campaign, what would that cost you? That’s called earned
media and often based on relative CPMs or what you are paying Google for clicks/visits.
What would it cost to buy the traffic? Here we are assuming that markets are efficient and that your
advertising prowess is middle of the road. If you spent $50k in Facebook ads, it must be worth that in an
efficient market. Not a good valuation model, but it’s actually the most common model to value advertising.
Don’t believe me? Almost all agencies charge a fee based on media spend. So if you have an agency, then
you are buying based on assuming the price is the determinant of value. Even worse is buying on a rate card,
which is the equivalent of buying at “list price”, negotiating a 15 percent discount, then claiming it was a
good deal.
• When you talk to advertisers, know which model they are using. Agencies understand the percent
of spent markup. Performance players (lead gen and e-commerce) look at the first model (ROI and its cousin
NPV, or Net Present Value).
• If your fan value is under $5, it’s due to a lack of engagement relative to your user base. Most
retail brands are in this scenario, where they have invested significantly in building a fan base, but allow their
investment to atrophy from not having engagement apps that stimulate and nurture relationships. Don’t be
guilty of abandoning your customers.
• Your Facebook fan page report has flags for the presence of engagement tactics. If you do
not have any of these apps, you effectively reduce your Facebook presence to a complaints board - a place
where unhappy consumers can vent, often with significant brand damage if there is not a swift response.
2. Industry Multiples
Industry Multiples is the trickiest of valuation techniques because it’s subject to speculative
bubbles. Just because everyone thinks a fan is worth $10, then you drive off the cliff, too. ClickZ published
a “study” showing fans to be worth $3.60, as if all fans are homogenous. Adidas said fans are worth $90
based on the faulty logic that these people who are Facebook fans wouldn’t have bought otherwise - they
didn’t measure incremental value, but took full credit.
There is no lifetime value of a Facebook fan because nobody has perfect vision into what
changes Facebook will make, what the shift in marketing channels will be over time, and so
forth. So it’s tricky to compare against direct mail, TV, or other more stable media. Perhaps the closest is
email, which is semi-interactive and an initial acknowledgement of permission based marketing. I’ve used
2x as a rule of thumb- to say that a Facebook fan is worth 2x an email subscriber. Why 2x? Because emails
go stale faster- people check their Facebook often multiple times per day and aren’t about to discard their
identity for a new one each month. They don’t have multiple Facebook accounts, but will have one Facebook.
And there is no social power is in email, since replies are not broadcast into the feed. The recommendation
power of a Facebook fan, properly leveraged, is tremendous. The third, we will discuss in more detail.
Analytics
61. 61 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
4.4 Measuring Engagement
If you're an administrator of your brand's Facebook page, you can access Insights.
Here are 2 key metrics to determine your engagement:
What percentage of your fans
are interacting with you? Is it
under 1%? If so, then your fans
are saying your messages are not
interesting. Facebook, noticing this,
will also not share your messages with
friends — a double whammy. Improve
your score here by asking questions
as opposed to making statements,
keeping messages 10 words long or
less (versus a paragraph), including
multimedia versus just text, and
having content that triggers emotion.
Above all, resist the temptation to
sell too hard. The wall is where most
engagement occurs, although you
can boost your engagement rate by
running ads to your postings, plus
including engagement apps.
Interactions per Week looks at how many interactions you have. Underpost and you have no community.
Overpost and you wear users out. Your ideal post frequency is dependent upon what your users expect. If
you’re a news site, then 10 times a day can be okay. If you’re a B2B software company, a few times a week
is probably fine. Do not cheat by auto-posting your RSS and Twitter entries—fans will frown upon non-social,
automated content. Don’t promote your specials or discounts more than once in every 4 postings, and
definitely no more than once per day. If you are getting less than 10% of your fan base as interactions each
week, you’re not unlocking the value of your community. Thus, if you have 100,000 fans, you should have at
least 10,000 interactions each week.
1. Feedback Rate
2. Interactions Per Week
Analytics
62. 62 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Deeper Measurement Options
Track your engagement apps and understand more about how you're engaging your fans and
who are the most influential.
Analytics
63. 63 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Ads Apps Analytics
4.5 Social Media Health Check
Ads, Apps, Analytics
See how well you're doing tying this all together for your brand.
• Do you rotate your ad creatives
at least twice a week to prevent
burnout?
• Are you running separate ads for
fans, friends or fans, and non-
fans?
• Do you have at least 20 separate
interest targets for which you have
personalized messaging to each?
• Is your CTR over 0.050% and/or
your CPC under $1.50?
• Do you know what in-line likes are
and how this Facebook feature
drives massive fan growth?
• Do you leverage the endorsement
of existing fans via FFO targeting?
• Do you have a reveal tab?
• Do you allow fan postings to
be visible on your page?
• Do you use the XFBML like
button on your website to be
"instantly social", revealing
which of a user's friends are
already fans?
• Do you have at least one
engagement app such as
coupons, voting, quizzes,
sharing, and badging?
• Do you know how your Facebook
marketing effors dovetail with other
marketing channels?
• Have you calculated the value of your
fans based on earned media, conversion
value, or as assit?
• Are you able to track interactions from
Facebook all the way through to a
conversion event?
• Do you integrate reporting of what
happens on your Facebook page with
your website into a single dashboard?
• Do you know which of your fans are the
most valuable based on demographics,
interest, and engagement?
Analytics
64. 64 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Competitive Benchmarking
More Fans
Increased Reach
Search Ranking
Earned Media
Total Interactions
Coupon
Redemptions
E-commerce
4.6 The Engagement Funnel
If you don’t have a goal, then all roads
look as good as any other. Yet more often
than not you’ll find that few brands have specific
goals with social media. Those that do are
unlikely to have them quantified. To say you want
“greater awareness” or “deeper engagement” is
not enough. You need to have quantifiable goals
based on a reasonable set of short, medium, and
long-term targets. But that requires understanding
a few things.
• What IS the right number of fans to have?
• What is a healthy level of engagement?
• How am I doing versus my competitors?
• If I’m a direct marketing company, can I really
drive revenue, and if so — how much and how
does this play with my other channels?
Analytics
65. 65 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Take advantage of the power of microtargeting on Facebook — at a crazy
cheap price!
In the “Demographics” section, we targeted males ages 30 to 40.
5.1 How To Run An Effective Facebook Campaign For $5
Not long ago there was a buzz in the Webtrends and BlitzMetrics
offices. One of our employees was trying to get my attention. He
did so by creating a Facebook ad targeting anyone who lived in
Portland, was between 30 and 40 years old and worked at either
Webtrends or BlitzMetrics. Of the nearly 600 million users on
Facebook, only 80 people met that criteria.
It cost him only 6 cents to do it.
And for that price, he was able to bombard our people with ads.
The cost of that inventory is a 30 cent CPM, which means it costs
30 cents to show a thousand ads. So he was able to send 200
highly targeted messages. Using the “Location” section in the ad
tools, he entered Portland, Oregon which is where BlitzMetrics
headquarters is located.
Ads
SECTION 5 - ADS
66. 66 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
For “Education and Work” we typed BlitzMetrics LLC and WebTrends.
The result ended with an estimated 80 people targeted.
The total cost is 6 cents.
Ads
67. 67 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Social media success is about pinpoint precision targets.
We’re simulating the one-on-one conversations that friends have
among themselves.
Sounds less like advertising and more like super-targeted email marketing, doesn’t it?
And, in fact, it is, except for this:
• You can send these messages without needing someone’s email address.
• You pay only when someone clicks it (yes, it’s cost per click advertising).
An impression is guaranteed when the person next opens Facebook (whereas in sending an email, you can
only hope that someone will open it).
Now imagine that you’re a software company like Webtrends, building relationships with other agencies that
resell your social analytics software. The founders of the data visualization agency JESS3 come to visit and
you’d like to strengthen that bond. Maybe you spend $5 on a micro-targeted campaign like the one above,
but slice it up to put the ad image more compactly next to the stats. You absolutely bombard 50 people,
that’s 60 ads per person. Who cares that we got only 9 clicks (of which 4 happened to become fans)? The
goal is not the click, but the awareness.
Total cost : $5.67 in Facebook ads
Ads
68. 68 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
But you could take it a step further, since those folks who do click through on the ad can come
to your landing page. So imagine that we send all employees of the email marketing company ExactTarget
to this Facebook landing page. (Warning: there is sound.) And how much did this landing page cost? Only
$5. We have a network of dozens of freelancers that will do voiceovers, take photos, sing songs or do
whatever for a few dollars.
While each of these examples might be clever or interesting, the question becomes: How do
you scale this? Social media success is about pinpoint precision targets — ultimately, because
we’re simulating the one-on-one conversations that friends have among themselves. But if you
want to have 1,000 conversations, you need 1,000 different ads and 1,000 different landing pages. Who
has the infrastructure, staff, or the budget to do that?
This is where smart automation comes in. Here’s an example of our scoring platform at work:
Our partner Webtrends sells analytics software to the big boys who don’t mind paying $100,000 per year.
Trouble is that every website needs some form of analytics. Maybe they’ll use Google Analytics — it’s free
and pretty good. But our company wants to talk to only those clients who have the money and need for
enterprise analytics software. It would be suicide to buy the keyword “web analytics” on PPC because of all
the players that offer web analytics for free or super cheap.
So we took the Fortune 1000 and ran a script that collected a wide range of data — market cap, their
industry, annual revenue, P/E ratio, website url, homepage pagerank, pages indexed, Facebook page,
number of fans, company logo from Google images and so forth — dozens of metrics. See the detail from
our spreadsheet/CSV file below.
And then we ran this data through our
scoring algorithm to calculate their
Social Score — how well they did versus
peers in their industry. We might say,
“Shell, you got a 56 and rank 7 out of
9 in Oil and Gas.” Or we might say,
“Shell, why do you have only 53,548
fans while others in oil and gas have
184k on average?” Then we target
people who work at Shell — not just
everyone, but those people who have
titles of VP of Marketing, Chief Financial
Officer, Public Relations and so forth.
Ads
69. 69 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
There might be only a couple dozen people and not everyone puts their
information on Facebook, but it’s enough. And you can bet it gets their
attention! They come to a landing page that has their social scoring report,
which shows a portion of the metrics that we’ve gathered. But they have to
click "Like" to see the rest of the report, which is grayed out.
Now what happens when that person clicks “Like”? Of course, some of their
friends and co-workers see it. And as all curious co-workers will do, they
want to check out what you found to be so interesting. And then when these
people see our ad, it shows that their friend liked it, which makes our offer of
a report that much more social relevant (image at right).
Move to Quality Targeting Over Mass Media Blasts
Now do you see how this works? It’s quality over quantity, folks. Think about who you want to
target as precisely as possible. Where do they work? Where do they live? What kind of car do
they drive? What TV shows do they watch? What industry conferences do they attend?
Can’t afford $15,000
to exhibit at your
favorite conference,
plus the $3k to ship the
booth out, the cost of
the people to have to
man the booth during
Expo Hall hours, the
schwag you have to give
out and so forth? Then
run an ad for the three
weeks leading up to the
conference targeting
fans of the conference.
Bingo, you’ve now spent $5 to target this audience with your message and you have plenty of time to set up
in-person meetings with those folks who are worth talking to, as opposed to any random people who might
wander up to visit you at the show. And then you can thank them later.
Need some PR help, but can’t afford a New York PR agency for $10,000 a month? Then let
Facebook do the work for you, running ads that target journalists who write for the Wall Street Journal,
Mashable, Forrester, VentureBeat, the New York Times or whoever. What would you like to say to them?
Can’t afford to hire a big sales staff to cold call people who don’t want to talk to you? Easy. Just
run ads targeting the competitors of your existing customers. Let’s say that Marriott is your client and you’ve
got a great case study there. Run ads targeting the folks who work at Hilton, Starwood, Motel 6 or whoever.
You can bet they want to know what their competitors are doing. Inquiring minds want to know!
Ads
70. 70 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
By now, I hope to have shown you that with some ingenuity and $5 in your pocket, you can make some
serious waves on Facebook. If you’re a small business or start-up, learn how to master some of the techniques
mentioned here. If you’re a big brand and looking to scale, then you’ll need some process and software
automation to make this happen across thousands of conversations.
Know of any companies that offer software that will do mass personalization of ad and landing page
content? Ad agencies are good at throwing bodies at client accounts — great service, but no scale. Software
companies are good at building code based on a predefined set of rules that can be repeated. But success
for your company can’t be solved by either a pure agency or a pure software company. The agency can’t
throw enough people at the problem and the software company can’t offer a one-size-fits-all solution to
everyone.
Only you can work the magic at your company. As much as we’d like to sell you some software, vendors
like us can only assist you in coming up with the creative strategy that resonates best with your customers,
the PR strategy that gets the press talking about you, a unique way to position how you solve your client’s
pain. Ultimately, these $5 campaigns, whether you run just one of them or 10,000 of them, boil down to a
marketing strategy — a unique, compelling message — that we can multiply out to your customers and get
those customers to spread on your behalf. (Again, if you’re a smaller company targeting just a few potential
or existing clients or partners, go for it yourself!)
Making Waves With 5 Bucks In Your Pocket
Ads
71. 71 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Ads
Feedback
Visibility
Fans
Micro-targeting on Facebook
5.2 Ad Targeting to Increase Feedback
Most of the pages you come across suffer from terrible engagement rates. They eagerly grew
a lot of fans in 2010 but forgot about what to do with these fans. Allowed to go stale, these fans
never see the brand again. The brand may post to the wall multiple times per day, but not have high enough
EdgeRank score to qualify for showing up in the News Feed (the user’s home page). The solution is to use the
ads platform to reconnect with these fans and to run ads so targeted, even non-fans will respond.
Ads
72. 72 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Interests
Media & Press
Workplaces & Job Titles
Interest and Group Targeting
Your ads should always be targeted, else they risk not seeming both relevant and personal. If the client has
run ads before, ask them if they recall their average CTRs. Usually they won’t know, but if they do and it’s
below 0.050%, then something is wrong. Likely, they are not choosing highly relevant interest targets and then
creating dozen, if not hundreds, of such ads.”
Ads
73. 73 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Increasingly aggres-
sive ads being ap-
proved > better CTR
> more profits
5.3 Good Facebook Ads Stimulate
Social Media - You’d think that would be an oxymoron, such as “open secret”, or “deafening
silence”. So many people are getting into the social media stream, and so few are profiting.
But it is possible to get practical value out of social media - let me provide a few tips:
Piggyback on trends: When Michael Jackson died, we had ads up in minutes. The ads were at first
tame: “Is MJ dead?” and then increased to vulgar - “See the King of Pop’s grand exit”. By the time
everyone else jumped on the bandwagon a day later, it was too late.
Automate things that work: If you spot something that works, systematize it.
Listen to what's popular: Tapping into what everyone is talking about is crucial to social media.
Imagine you’re back in school and you’re trying to keep ahead of the curve with what’s hip. Luckily,
almost every single social medium has a feed. This handy feature allows you to see a condensed version
of what’s new on someone’s blog, Twitter account, or Facebook page. Check out Google Reader
and subscribe to a few news websites, and keep an eye on what topics are popping up the most. For
example, compare the popularity of Michael Jackson versus Billy Mays versus Farah Fawcett over the
last 30 days. Clearly, MJ is the winner of the dead celebrity contest. So that ought to give you an idea of
what hot trend to be riding.
CONCLUSION: It's easier to ride someone else's wave than to have to create your own.
Ads
74. 74 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Current Ad: Suggested Revision:
5.4 Optimization
Note that I didn’t mention in the ad that they’re a dog lover. I believe that can safely be assumed because of
the image. In some cases, where the image is not able to represent the interest, we do want to explicitly call
out the interest target.
Using your Merchant Name in the ad is a double-edged sword. In this case, I believe we shouldn’t use it
because the merchant doesn’t have brand awareness. Remember, to get a click from people who don’t know
us, we must establish trust quickly - and no brand is no trust. Using the long brand name also lengthens the
ad copy and makes it slightly harder to understand the value proposition.
Social media requires the lowest of conversion hurdles (the minimum of thought or consideration), since
people are on Facebook to socialize That’s why I advocate the two-step conversion for most clients - one
set of ads to get the right interest targets to be fans, then another set of ads targeted just at fans. True, it
costs more to run more ads, but the prices are so low on Facebook ads that we can get away with this. Plus,
depending on the nature of the product or service - hitting people at different points in the funnel can be far
more effective than going for a jump all the way through to conversion in one leap.
A good rule of thumb is a 3X improvement on ads when you optimize them. In other words, if your
cost per fan is 60 cents when you test the first couple days, you should be able to get it down to 20 cents on
average. The 3X rule is what we’ve discovered is reasonably possible with a moderate amount of testing - of
course, you could get far more.
* Liking as a condition of entry is not allowed, but many brands do it. Proceed at your own risk.
Ad Copy Techniques
Determining Who Is The Winner
If we have 3 images - which of them performs better? By “perform”, I mean by the goal we’re being
measured on, as well as early indicators (CTR and CPC). If we have 3 interest targets, how do they perform?
Or maybe we have a few age/sex/location targets. Unless we have multiple variations of each ad multiplied,
it will be hard to isolate a winner and isolate the performance of a single variable.
Ads
75. 75 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Demographic Optimization
FOF (Friend of Fan) and
Non-FOF Ads Split Test
Run one ad that is not FOF and another that is, just
so we can get an apples-to-apples comparison. In
other words, the ads are identical except for this
targeting change. This is called an A/B test and we
want to determine the differences in CTR, CPC, etc...
Don’t let it run more than a couple hundred dollars
in total - or enough to have a significant difference.
If you have an interest target that is already highly skewed towards a particular demographic (combination
of age, sex, and gender), then it’s not necessary to also filter by demographics, too. For example, if you are
targeting New England Patriots fans (folks who like a professional football team), it’s not necessary to specify
males as a demographic. If you do, that will cut out 20% of the audience, as there are still some females that
will like this.
Therefore, when setting up initial ads that have interest targets, demo targeting is usually not necessary.
But in subsequent refinement, you might want to multiply into combinations of these demographics. The
Demographic reports in Facebook’s ad tool will tell us that the CTR or conversion rate may differ or not.
Ads
76. 76 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Fan
Targeting
Non-Fan
Targeting
Friend of Fan
Targeting
Your friend is attending,
how about you?
Segmentation
Brand Logo On The Ad Image
Interest Targeting - Friends of Fans Targeting 1
Cold Targeting vs. Fans Targeting
For many large brands, often the highest CTR
and lowest CPC/CPF is the brand image
itself. Thus, if we could make the logo larger,
it could increase the CTR. Maybe even try
a colored border, provided that the client
approves this.
When you do interest targeting, you don’t always have to put in a gender filter, too. For example, if you
specify “cooking”, you don’t have to necessarily also place “female” as a gender target. There are some men
who like cooking as well. To see how the counts change, try the targeting in the ads interface with and without
female targeting — you’ll see only a minor difference. The reason is that most of the folks who like cooking
are female.
Average CTR of ads to non-fans: .05%
Average CTR of ads to fans: .35%
Ads
77. 77 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Interest Targeting - Friends of Fans Targeting 2
Interest Targeting - Friends of Fans Targeting 3
Ad Distribution
Have you ever considered doing pure
FOF targeting in cities/states where
your brand is strong? For instance,
let’s say you’re running a campaign
for a fast food chain in California
cities since the location density is
much lower than the rest, say cities in
New Mexico. The lower the people/
location figure, the better we’ll do,
as the people seeing the ads are
more likely to have a location near
them. In addition, online ads should
follow offline popularity as a tactic to
increase conversion.
Choose interest that has at least
1,000 population reach. When
doing FOF, the estimated reach
should be at least 10,000 but even
if only 500, it’s worth doing, you’ll
just have to create a LOT of them.
3 Types of Audience Reach:
Targeted: Based on the targeting you’ve selected, this is an
approximate number of people your ads and Sponsored Stories
can reach.
Reach: The number of individual people who saw your ads or
Sponsored Stories.
Social Reach: The number of people who saw your ad or
Sponsored Story with the names of their friends who liked your
Page, RSVPed to your event, or used your app. If you are running
ads off Facebook, you won’t be able to see the social reach.
Ads
78. 78 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Ad Multiplication
Ad Burn Out
Try our top ads with FOF Like cooking?,
Like [interest]? For example, target
“food network” and say “Like Cooking
Shows?” target “Bobby Flay” and say
“Like Grilling?” If you create enough
of them, you can do FOF targets,
too FOF significantly limits reach, of
course, but the trade-off is better CTR
and conversion.
It may be that if you run an ad
that was semiburned out, but
gave it time to “rest”, it gets a
light refresh.
A good predictor of ad burnout
is not just the trendline for
CTR, CPC, and CPF, but also
looking at the audience size
each ad targets. The smaller
the base, the fewer number of
impressions needed to burn out.
I’m saying that if the audience base is small, it takes fewer impressions total to burn out. For example, an
ad that targets 100k people can serve a lot more impressions before burnout than one with only 1k people
in the target. Smaller audiences might not burn out faster in terms of number of days, since they will also
show fewer impressions per day. Therefore, in estimating future traffic and our ability to hit fan targets at CPF
goals, we should account for audience size. Assume we can hit users 20 times before burnout on average.
So if the audience is 100k people, we can serve 2mm impressions.
Audience size matters especially in local. If some campaign wants 100k fans in just Atlanta vs nationwide,
it will be a lot more expensive because we hit burnout faster - there are fewer people to target. We get
saturated faster. In large multiplied local campaigns, whether for a franchise or a single location, we have to
consider the bounds of each geo-target. This is true is regular PPC, too. ReachLocal has a traffic eliminator
built into their proposal generator, which I’ve shared a few times. Buggy - a good concept, but poorly
implemented.
Learn how to put your Facebook sponsored stories on steroids!
http://www.allfacebook.com/facebook-sponsored-likes-2012-02
Ads
79. 79 BlitzMetrics
TM
Social Analytics Training Manual
BlitzMetrics.com 1960 Joslyn Pl. Boulder, CO. 80304 CONFIDENTIAL AND PROPRIETARY V2.9
Naming Conventions
Please name the ad something descriptive, since the default ad
name (the internal name we give it, which the users can’t see), will
default to “(campaign_name) (number)”, which is not descriptive or
helpful when we want to go back to view stats. Look at the name
of the ad you are multiplying from and just copy that ad name,
appending “image-2” to show we are testing that same ad but with
another image. Better yet, append “Steak_food_image.women.close.
up” to the ad name to be even more descriptive. If you use the ad
multiplication tool, the ad is automatically named for you. With new
clients, run ad so that you can get a solid base to multiply from. Do
not just multiply up 1,000 ads to start.
Education Level Impacts CTR
Competitive Targeting
Want to know how one
small business saw 603%
growth from effectively
using sponsored stories?
Click here.
Ads