Draft presentation slides for my Stanford class on customer acquisition and sales. Overview on why most startups fail at acquiring customers. Covers: LTV, CAC, sales, and a number of other customer channels such as SEM, pay per install, PR, social media, and others.
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Why Most Startups Fail at Acquiring New Customers (and how you can succeed!)
1. Why Most Startups FAIL
at acquiring new customers
(and how you can succeed)
Saar Gur
Stanford Graduate School of Business
Startup Garage Presentation
January 2014
2. Me
Saar Gur
saar@crv.com
Twitter: @saarsaar
Current: General Partner at CRV
Past: VP of Customer Acquisition for one of the largest online advertisers in paid
search, display, affiliate, email and SEO (Before “Growth Hacker” was coined)
Very lucky to have learned from a lot of world-class people and companies:
50+ Million Users
200+ Million Users
150+ Million Users
3. Why Most Startups FAIL at Acquiring New Customers
#1 Reason: BAD PRODUCT
4. #2 Reason: Don’t really understand
who the customer is and their needs
5. #3 Reason: Clueless on the amount of
effort it takes to acquire new customers
Expectations
Reality
Example: A sophisticated marketer may run hundreds of A/B tests in a given year to optimize
their banner ads. And will do the same for landing page sign-up flows, email subject lines, etc.
9. Topics for Today
1.
2.
3.
4.
5.
6.
Developing a great product
Understanding your customer
Expectation setting
Making good spending decisions
Channel overview and selection
Channel examples
10. How much can you spend to acquire
a new customer?
11. Customer Lifetime Value (CLV or LTV) analysis helps
set an upper bound on what you can spend to acquire
new customers profitably.
12. Why does LTV set an upper bound on CA?
IMPORTANT
Drill This Into Your Head
When the cost to acquire customers is greater than the
company’s ability to monetize those customers,
the business model FAILS.
Common mistake: Entrepreneurs wildly underestimate how expensive it is to acquire
customers, make wrong assumptions about LTV (e.g., see unexpected churn), or make
bad assumptions about expected future improvements in their CAC and LTV.
13. Even startups with limited or
no customer data should still
think about LTV before making
marketing and sales decisions
14. Lifetime Value of a Customer (LTV) =
net present value of the profit stream of a
customer
Key things to think about:
- Average profitability per customer ($)
- Customer life (years)
- Cash cycle
15. LTV is about profitability (not revenue!) per customer
Advertising as a business model is hard
without scale, & LTVs tend to be low, but
gross margins at scale are great (70%+):
(Estimated revenue per user per year)
Google: $29.32
Yahoo: $8.00
Facebook: $4.12
Yelp: $1.26
TIP: Wrapping a low margin single sale
item with a high margin add-on product
will substantially help your LTV
Ecommerce has bigger ARPU, but
gross margin is usually much lower
(20-40%):
(Estimated revenue per user per year)
Apple iOS User: $150.00
Amazon: $245.00
Zulily: $210.00
Enterprise software has rich customers, high margins, high LTVs :
(Estimated revenue per customer per year)
Salesforce: $15,000
Sybase: $1,000,000
16. Customer Life: Often much longer than expected
Vendors at
tourist
attractions
Dentist
Average Job in US=4.6 Yrs
Customer Life
SHORT (1 Transaction) MEDIUM (0-36 Mos.)
Long (Years)
VERY LONG (10+ Years)
How many years have you been using the same razor? Using Google? Drinking Coke?
17. Why does this matter?
Every Penny Counts in LTV because:
(1) If your LTV is “low” it is hard to invest
in paid acquisition channels
(2) Many media channels don’t scale
linearly
18. Many Channels Don’t Scale Linearly
This is the front page of Yahoo. 392 Million Visitors/Month. Note the banner.
19. Many Channels Don’t Scale Linearly
Example: Y! front page 300X250 banner ad has a minimum commitment of
$250,000/day at a CPM (cost per thousand impression) of $2.00*. If you are able to
profitability acquire customers and pay $2.00 (LTV>CAC), than you get TONS of
customers through this ad placement. If your CAC>LTV by even $0.01, your business
model is upside down, and you can lose millions. Many media channels have this
non-linear characteristic.
Another example: If LTV>$2250 you can hire inside salespeople. If LTV<$1000 generally
won’t ever be able to make it work. Makes a big deal if you think a $100/month
subscription customer will have an average customer life of one year or 3 years.
*Numbers are used for example purposes and not accurate
20. CAC Payback: Not All LTVs are Equal
Example: Two customers generating $150 in revenues
YEAR:
Customer A
Customer B
Margin assumption:
Customer A
Customer B
1
2
3
4
5
$ 110.00 $ 10.00 $ 10.00 $ 10.00 $ 10.00
$ 10.00 $ 20.00 $ 30.00 $ 40.00 $ 50.00
Total Revenues
$
150.00
$
150.00
50%
$ 55.00 $ 5.00 $ 5.00 $ 5.00 $ 5.00
$ 5.00 $ 10.00 $ 15.00 $ 20.00 $ 25.00
$150.00
Customer A
$100.00
Customer B
$50.00
Gross Profits
$
75.00
$
75.00
$-
1
2
3
4
5
Assume CAC = $50. Customer A makes $55 in profits Year 1 and can reinvest those
profits in acquiring more customers in Year 2. Customer B won’t generate $50 in profits
until the end of year 4. Company B has to borrow money or raise more equity if it wants
to spend money on acquiring customers in years 2, 3 or 4.
The time to recover CAC is a capital efficiency indicator for your company and an
important metric for startups where capital is tight and expensive.
Tip: Charging upfront installation/setup fees or asking for some pre-payment
upfront (with a cancellation fee) can have a large impact on your cash.
21. Last Thoughts on LTV
- LTV is about Gross Profit per customer, not Revenue per
customer: Many websites out there are wrong! Your
COGS cannot be ignored when determining LTV.
- Overhead costs such as administrative expenses and R&D
are not included in LTV calculations. Operating at a
higher LTV>CAC ratio roughly takes this into account.
- “Attractive” investor metrics are: (a) LTV>3x CAC and (b)
less than 12 months to recover CAC
- As a startup, don’t worry initially about the LTV>3x CAC
ratio as it will take money to test and optimize new
channels. But be very mindful of your LTV and CAC so
that your business model is not upside down (LTV>CAC).
22. Setting a Sales/Marketing Strategy
(pre-channel work)
Need some understanding of:
• The Customer Profile
–
–
•
The Decision Criteria
–
–
•
Where can you find them online or offline?
Location, Age, Gender, Marital Status, Children, Income, Interests, Device (e.g., iPad? Internet
Explorer?), Magazines they read, Conferences they attend, etc.
Do they know that they need this product already?
What factors matter most to them in the purchase decision? (e.g., price? Look or taste? ROI?
speed? convenience? Brand name? Not getting fired?)
The Purchase Process
–
–
–
–
How is the purchase decision made? Is there one decision maker or multiple approvals required?
Is there an existing budget or does one need to be created? Where is this purchase taking money
away from (share of wallet)? RFP process?
Impulse purchase or a considered purchase? Does the buyer need to talk to a person before
committing? Can this be sold online without a person? Can it be sold over the phone?
What influences the decision maker? E.g., Their peers, research sites, manufacturer/company
websites?
How long does the buying process take? Immediate or months?
Helps filter high level decisions such as:
- Do you need to touch the customer in multiple channels or will one channel work?
- Do you need to hire salespeople to close the sale?
And it drives your messaging as you develop sales and marketing collateral.
23. A Single CA Channel Can Work!
Easier when:
- Low price point.
- Consumer has intent.
- Brand is not important.
- An impulse decision and easy to
purchase.
(Minimal effort required)
- Limited time from purchase to
delivery
(less likely to cancel or
change their mind)
24. Considered Purchases
Example: Buying the search term “best electric car” on Google won’t
work to sell a new startup vehicle for $40,000 to consumers.
Need to market in multiple channels to convert car intenders to buyers.
Example: The Car Buying Process = 6 Months
Months 5-6
Upper Funnel
(Demand Generation)
Months 2-4
Middle Funnel
(Maintain
Consideration)
1 Month-2 Weeks
Lower Funnel
(ROI)
TV ads, magazine
ads, manufacturer
websites, friend/relative/colleag
ue
Consumer review
sites, magazine ads, television
ads
Local newspaper ads, local
dealer sites, friends/relatives
25. Overview of Channels
• Every customer acquisition channel has positive and negative attributes.
• Most channels eventually “run out” of customers and don’t scale effectively
beyond a certain point
• Early adopters can be “extra cheap” and so CAC tends to go up significantly as you
cross the chasm
• Existing channels tend to be expensive (fully priced). New channels are created all
the time and may or may not work for your product. There are generally HUGE first
mover advantages for testing new distribution channels and platforms as their
media is not efficiently priced – especially when those channels become huge
(e.g., Google in early 2000s, Facebook, iPhone App Store). Over time prices can go
up 5-50x as a marketing channel matures and other advertisers start spending.
– Example from Groupon: “Between the first quarter of 2010 and the first quarter of 2011, the
customer acquisition cost rose 485% to more than $30 per email address.”
26. Overview of Channels
• LTV and CAC will differ by channel and often within a channel.
– Example: Buyers for “Diamond Ring” on Google will have different LTV metrics than sales from the
search term “Dimond Ring”
– Example: One salesperson on a team may close sales like crazy but have high customer churn
(selling them a bag of goods), while another may have lower sales but high LTV customers.
• In some channels you are competing with direct competitors on LTV, and in other
media channels you are competing against everyone.
– Example: In Google PPC – Zappos and Cheapshoes.com compete for clicks on the keyword “Nike
shoes” but in online display Zappos competes with Netflix, Bank of America, and others.
• Huge reductions in CAC or improvements in LTV can be achieved by optimizing
conversion funnels by each customer acquisition channel
– Not covered today, but worth mentioning that the focus of many growth hackers is on onboarding
and retention by optimizing sign-up flows and making sure customers stick
27. Channel Attributes
Attribute
Example
Ability to Target
(Intent? Females over 30? U.S. only?)
Are you spear fishing or casting a
broad net?
Minimum Test Budget
(How much capital is at risk on a given
campaign? How much capital is at risk to fully
test a channel?)
Example: Minimum buy may be
$10k and need to commit to 5
tests to adequately test the
channel
Creative Requirement
(Custom creative needed? Images?)
Making a commercial requires
money, time and creativity
Time to traffic
(get impressions)
Adwords=Immediate, A magazine
ad could take 3 months
Degree of management and coordination
required
One person can often manage
SEM, SEO requires code changes,
Telemarketing requires
management of a sales team
28. Channel Attributes Continued
Attribute
Example
Measurability (tracking)
TV, Radio, Print are hard to
measure
Scalability of the Channel
There are only a certain number
of searches and intenders on
Google on any day.
Predictability of the channel
Consistent performance of the
acquired users, no fraud/spam
Control of the channel / Platform risk
Being featured in Apple’s App
Store, Google changing its search
results
Degree of integration and schedule/pace of
the partner
Ecommerce sites freeze code in
October
29. A Few Channels to Acquire Customers
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Radio
•
Television
•
Print
•
Billboards and Out-Of-Home
•
Direct Mail/Catalog, Door hangers
Online Display
•
(Banner/Video/Mobile/Retargeting)
•
Paid Search
•
Social Media / WOM
Viral (user contacts for invites: email/facebook)
•
Incentivized word-of-mouth (tell-a-friend)
•
Search Engine Optimization
Affiliate – CPA, rev. share, coupons, bundles •
•
Pay Per Install/Download
•
Incentivized actions/downloads
•
Email
Co-Registration & 3rd party/outsourced lead gen
Platform promotion (e.g., App Store)
PR
Paid content marketing
(e.g., Infographics, white papers)
Third party endorsements (e.g.
Forrester)
Educational programs
Webinars
Offline
events, tradeshows, conferences
Contests
Freemium/Free trial
Telesales / Inside Sales
Field sales
Field sales with SEs
BD and Channel sales
30. Some Channels Won’t Work. Period.
Social/Viral:
- People don’t want to talk
about or share certain things
with friends (e.g., being in
debt)
Search:
- Can’t buy keywords if there is
no intent (e.g., selling a new
service that has no existing
demand and does not fit a
category)
But you can market to them in plenty of other channels
Target a Demographic Directly in Print
Use Infomercials for Demand Generation
32. Channel Example: Google PPC
Merits:
• Highly targeted: Great LTV performance since you are finding
intenders (know they have a problem, looking for a solution)
• Easy to create a text ad, no images required
• Easy user interface, fast setup
• Able to traffic immediately
• Can test with a small budget (<$1000)
• Consistent performance (limited fraud), regular search patterns
Considerations:
• Expensive channel, mature with existing advertisers
• Requires regular bid management and creative optimization
• Volume is limited even if you broaden the search terms
34. Pay Per Install
Merits:
• If your product is a mobile app, this can be a fantastic and highly scalable
channel
• Great targeting on Facebook (targeting in mobile used to be hard!)
Considerations:
• Expensive. Need high LTV to compete with other spenders.
• In most cases, building a mobile app is a bigger development effort than
putting up a website.
• Consumers have limited attention for apps. Just because you got a
download does not mean that your app will remain relevant in the sea of
apps on the user’s phone.
• Doesn’t work for many companies where a mobile app is not the primary
customer experience. (Example: Quickbooks)
• Platform risk if 90%+ of your spend is on Facebook for installs
36. Channel Example: Social Media
Merits:
• Low cost, high impact when it works
• Potential to go non-linear and become a phenomenon
(e.g., Gangnam)
Considerations:
• Very difficult to execute well.
• Hard to create an authentic message that consumers authentically want
to share and talk about. Won’t work for many companies.
• Requires specialization: Needs to be tailored specifically for the
channel (A Pinterest campaign will be totally different than a Twitter
campaign. There can be 10+ channels within Facebook!)
• Often hard to repeat even from the most creative teams. (e.g..,
DollarShaveClub – one trick pony?)
• Hard to forecast and control
37. PR
Twitter co-founder's startup:
Accept credit cards with your
iPhone
By Rachel Metz
Associated Press
POSTED: 01/01/2010SAN FRANCISCO —
Jack Dorsey revolutionized online
socializing by co-founding Twitter in 2006.
Now he wants to transform the way people
exchange money.
Uber Breaks Hearts With Short
Supply Of Kittens On Cat Day
The Huffington Post | By Lydia O'Connor Posted: 10/29/2013 7:49 pm
EDT | Updated: 10/30/20138:12
38. PR
Merits:
• Can get very targeted and/or very broad exposure for zero or little cost.
Considerations:
• Stories often don’t lead directly to sales.
• Hard to scale. Reporters often want to talk directly to the CEO or senior
executives.
• Constantly coming up with a story is hard unless your product is topical with a
broader world issue. As a result, many teams can’t build a sustainable PR channel
and don’t invest in PR.
– Example: Mint.com after the financial crisis when everyone was talking about saving
TIPS:
• Remember: Reporters and bloggers write stories. They are not trying to sell your
product.
–
–
•
•
Need a story arc – is your product topical? controversy? A hero? David vs. Goliath? Unique data or statistics you can share?
Is there a headline or message that can easily be shared?
Stories about the product and its features/benefits are much better than stories
about the founder, hiring events or financing events. (May help in raising money
though, or in creating the echo chamber for considered trials/purchases.)
Reporters are on their own deadlines. Be easy to work with an focus on long-term
relationships with them.
40. Channel Example: Incentivized WOM
Merits:
• Trusted referrals from friends have extraordinarily high conversion rates
• Best when the service actually improves when your friends joins (e.g., Facebook)
Considerations:
• Unless the incentive is very strong and customers feel that they are offering something
valuable to their friends, this channel usually does not scale significantly
(a) not everyone will participate (either they don’t have many friends or they don’t
feel the offer is valuable enough to send to their friends)
(b) you generally need a large existing customer base to start the campaign
• Requires tracking referrals and has the potential to create problems with existing
customers
• Hard to A/B test and offer different incentives to different customers (E.g., Tough to
offer one user $5 and another $500 for a referral)
• People will always look for ways to game the system. Need to look out for fraud.
41. Channel Example: Sales
Can you afford to hire salespeople?
Inside Sales
LTV>$2,250
Assumptions:
- Sales person will cost $75k/year in OTE
(On target earnings = base + bonus at full quota)
- Want LTV>3x CAC -> need to generate $225k in LTV
- Average inside sales teams will close 10 sales per month
- LTV per contract is $2,250 ($225k/10)
- Reminder: LTV is not the same as Revenues (the invoice
amount to the customer may or may not be close to LTV
depending on the COGS and gross margin
Direct Sales
LTV>$150,000
Assumptions:
- Sales person will cost $300k/year in OTE and
additional marketing support will cost additional
$100k = $400k
- Want LTV>3x CAC -> need to generate $1.2m in LTV
- Close 8 deals a year
- LTV per contract is than $150,000
42. Channel Example: Sales
Merits:
• Very effective IF your unit economics support it
Considerations:
• Many consumer product businesses can’t reach these LTV requirements
• Salespeople take time to get to quota and many don’t perform.
• Managing an army of salespeople requires management.
• Active cash management and sales performance monitoring are critical. There is
usually a funding gap between salespeople getting productive and the cash needs on
the business.
• Because you are hiring bodies and talking to customers, it is hard to reach the same
scale as many online customer acquisition models.
• Key sales departures or hiring gaps can impact revenues dramatically.
43. For another day…
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Radio
•
Television
•
Print
•
Billboards and Out-Of-Home
•
Direct Mail/Catalog, Door hangers
Online Display
•
(Banner/Video/Mobile/Retargeting)
•
Paid Search
•
Social Media / WOM
Viral (user contacts for invites: email/facebook)
•
Incentivized word-of-mouth (tell-a-friend)
•
Search Engine Optimization
Affiliate – CPA, rev. share, coupons, bundles •
•
Pay Per Install/Download
•
Incentivized actions/downloads
•
Email
Co-Registration & 3rd party/outsourced lead gen
Platform promotion (e.g., App Store)
PR
Paid content marketing
(e.g., Infographics, white papers)
Third party endorsements (e.g.
Forrester)
Educational programs
Webinars
Offline
events, tradeshows, conferences
Contests
Freemium/Free trial
Telesales / Inside Sales
Field sales
Field sales with SEs
Business Development and Channel
sales