1. Effective Digital Advertising:
Making Sense of
Data, Targeting and
Measurement
Bryan Schroeder
Senior Director, Global Advertiser Product Marketing, Yahoo!
Linda Young
Director, Addressable Advertising Solutions, Acxiom
2. Agenda
• Big Data Overview
• Data Challenges
• Proving ROI Through Offline Sales Attribution
• Case Studies
• Future Trends
4. Multi-Dimensional Data View
SEARCH DEMOGRAPHIC
EMAIL
INTEREST
SOCIAL
IN-MARKET
VIDEO ENGAGEMENT
CRM
APP DOWNLOADS
PURCHASE
BEHAVIOR
LIFESTYLE
5. Data is Growing Exponentially
Quantity of global digital data, exabytes
130 1,227 2,720 7,910
2005 2010
2012 2015
6. Data Challenges
Technology to Harness Large & Diverse
Data Sets
Navigating Privacy Concerns
Making Your Data Actionable
Bridging Online and Offline Sides of the
Business
7. Technology to Harness Large and
Diverse Data Sets
Big Data is not totally new
Volume, velocity and variety of data increasing
Complex data and combinations
Needed:
New technologies
New methods
New guidelines
New knowledge
Trusted partners
8. Navigating Privacy Concerns
Consumer: Guidelines to leveraging
behavior data
Marketer: Protecting your own data
Publisher: Protecting their user data
9. The Value of Addressable Data
Allows for persistent targeting
Combats cookie churn
Ensures accurate reach and frequency campaign
management
Facilitates multi-screen campaign delivery
Solves for shared computers
Enables offline sales measurement
10. Making Data Actionable
Leverage your own CRM files to reach your customers
online and measure the offline sales impact
Matched
Audience
Advertiser Yahoo!
CRM File User File
1 Advertiser 2 Acxiom 3 Ads reach
provides CRM matches matched
file to trusted Yahoo! users
advertiser file
across screens
partner, Acxiom to Yahoo! including
addressable PC, mobile and
user file tablet
11. True ROI: Offline Sales Attribution
1 2 Advertiser customer 3
Yahoo! campaign A report detailing
exposure data is offline sales data is sales or conversion
shared with trusted shared with Acxiom impact across
third party data and matched to channels is sent
Yahoo! exposure
partner, Acxiom data back to the
advertiser
12. Use Case: Department Store
Objective:
Mirror offline direct mail campaigns to reach existing customers
online and understand the offline impact
Solution:
Acxiom and Yahoo! partnered to target advertiser’s
customer list online and provide multi-channel insight
Results:
Drove $20M in in-store sales and $2.5M in online
sales
Experienced 11% incremental lift per shopper who
received both online and direct mail (vs. control
group, direct mail only)
Incremental 4:1 return-on-ad-spend for online
campaign
13. Use Case: Insurance Company
Objective:
Mirror offline direct mail campaigns to acquire new customers
Solution:
Acxiom and Yahoo! partnered to target advertiser’s
prospect list across Yahoo! and measure the impact
on offline sales
Results:
6.3x return on ad spend
$407,190 in incremental premiums
Users exposed to the campaign were converted at
a rate of 2x more than the control
14. Key Takeaways
Develop a strategy around using your CRM driven insights
online. Don’t just test once and stop
Addressable data enables persistent targeting and offline
sales measurement
Make your traditional and digital business work together
through partnerships
You can safely use anonymous data and PII data, while
staying privacy compliant and consumer friendly
15. Future Trends
Marriage of data and creative
Increased frequency and depth of campaign
insights
Tools to inform overall media mix and cross channel
attribution
17. FAQs
What segmentation strategies work best online?
What's the most efficient way to get started on a
CRM data-driven display campaign?
How long does it take to get this kind of
addressable display program up and running?
18. Thank You!
Bryan Schroeder
Senior Director, Global Advertiser Product Marketing, Yahoo!
bryans@yahoo-inc.com
Linda Young
Director, Addressable Advertising Solutions, Acxiom
linda.young@acxiom.com
Twitter: @YahooAdBuzz, @Acxiom
Editor's Notes
- Highlight a number of trends associated with the proliferation of consumer data- Dive into the challenges of managing and utilizing large consumer data sets - With challenges come opportunities and we’ll showcase innovative ways to leverage your data, reach your customers online, and measure the offline sales impact of your display marketing campaigns- Highlight case studies that illustrate best practices - Lastly, provide a perspective on where we think data driven digital marketing is headed
In the last 10 years, data and channels have exploded and given rise to the empowered consumer. Just think about it …It is simply staggering what happens every 60 seconds on the internet. Here’s a few stats: 1.4 million searches or Facebook updates. Almost 1 million comments about products/services between consumers. About 10 million display ads served. 2 million YouTube videos viewed.And it isn’t just online data….there is a vast array of information that originates offline or from a marketers own data. The proliferation of data has given rise to a substantial commercial opportunity; the ability to better understand consumers and in the end match the right message to the right audience and seamlessly manage profitable, long lasting customer relationships. At the end of the day, to capture this opportunity marketers need to have a data driven strategy
Let’s take a look at the data view of a given consumer. Insights can be gleamed from many of the online activities that are exhibited:Searches for Cruise dealsSocial and video engagementCombined with what we know about this person in the offline world:Demographics – has children, levels of income, in a certain age bracketPurchase behaviors – buys carbonated beverages once a week The key is no single attribute defines your current or potential customers. For marketers, success is predicated on moving from simply managing the collection and storage of these large disparate data sets into understanding and defining consumer audiences that can be reached across channels and geographies.Today we’re seeing digital campaign strategies that evolved from leveraging pure online publisher data like Yahoo!’s, to combining it with offline 3rd party data and CRM data. It’s multi-channel in nature. A study this year by Winterberry Group and the IAB corroborates the multichannel statement. The study highlighted how multichannel data integration is a critical component of improved media-buying capabilities. So much so, that they saw in some cases the integration of offline and online data resulted in a 30 percent increase in online display performance.
The average consumer is generally familiar with the term gigabyte…as in my iphone has 32GB of memory. A gigabyte is a billion bytes. “Show of hands in the room – how many of you know what an Exabyte is?” You were all going to say an Exabyte is a skin condition right? Kidding aside, it’s a BILLION gigabytes According to IDC in 2010 there were 1,200 Exabytes of data created. The figure jumped nearly 50% last year to 1,800 Exabytes To put it in perspective – that is enough data to fill 57 Billion iPads And the number is expected to grow 62 percent annually The size is growing exponentially….but it’s the types of data that is evolving. It used to come in neat and structured format, but the most unique data is coming in unstructured. Ultimately you need a technology to be able ingest and organize it. (Transition into slide 6 & 7 for Linda) There is growth in all possible dimensions of data # sources of dataTypes of sources of dataTypes of dataFormats of dataSize of data
Data challenges marketers face today: Technology to ingest and act on it How to enable Types of partners /options – Data partners, DSPs, DMPs, Publishers, NetworksPrivacy concernsLeveraging it while protecting consumer and advertiser privacyMaking data actionable Leveraging an advertiser’s own CRM data – citing examples for retail and finance advertisersSorting through and understanding data Finding the most valuable audience, the right people Importance/value of addressable data Actionable insights and analytics – Offline measurement Overcoming the existing hurdles between the offline and online parts of the organization
Acxiom has been handling “Big Data” long before others 5700 clients in 40 countries300,000 campaigns/year engaging 5 billion consumers1 trillion customer recognition events/yearBut the world has gone well beyond millions and billions of individual records and demographic data and structured data What are you going to do with the 98,000 tweets per minute or the 79,000 facebook wall posts per minuteThe size is growing exponentially….but it’s the types of data that is evolving. It’s used to come in neat and structured format, but the most unique data is coming in unstructured. You need technologies and methods to be able ingest and organize it.Add to that the complexity of knowing what data can be combined technologically because some data is tied to an individuals personally identifiable information, some data is known only at an unidentified cookie level, or a geographic level. You need strategies and methods for combining that data. Even when it is possible to combine that data, it is not always prudent, or ethical to combine the data and maintain the consumer’s preferences and rights. You need guidelines, knowledge and trusted partners. This combination of volume, velocity, variety, complexity makes it impossible to use traditional methods to capture, store, combine, manage and analyze. What is being employed to meet these challenges is:
Consumer – Take a moment and think of yourself as a consumer. You consume products and media every day. You are on your phone, your tablet, your computer and are bombarded with messages from all directions. Think about the fact that you DO want the messages you receive to be relevant and meaningful. But you DON’T want it at the tradeoff of information about your behavior or actions that you did not know or did not want to be tracked. The guidelines and the laws will continue to get tighter. By putting yourself in the position of the consumer, you more quickly understand what you will want to do and what data you will want to use for marketing, especially in the digital world. As a marketer you have the responsibility to treat your customer and data about their business with you with great care. Ask yourself - What is reasonable to the common man?If the reasonable man would not expect it, you must even more careful to outline it.The earlier the disclosure the better even at registration. If you send PII data to a 3rd party, say so.Disclose that you bring annoymous on-site behavior data together with PII post consumer registration.Provide the option not to participate in a program.Marketer – You go to great care as marketers to know as much as possible about your customers and your best prospects. You protect that data and analyze that data and make good use of that data to extend the relationship. We find that too often, the care taken in the offline world to protect your own data, does not extend into the online world. It is important to be certain that any company given access to this valuables data asset can be trusted to safeguard it, understand its importance, and only use it as it is intended. Marketers must take responsibility and control of their own data especially when marketing in the digital space. The digital space is blessed with hundreds of startup companies that enter with good ideas and capabilities. However, those same companies may not have the same risk for mishandling data that a giant brand has. Think of your brand on the front page in a not-so-positive story about the use of your customer’s data. Publisher - The protection of user data is paramount and a reason why publishers should work with trusted partner who understands the privacy landscape. Equally as important is transparency and control over the collection and use of data. Yahoo! in particular offers a robust privacy center and several tools to empower consumers to manage their experience on Yahoo!.
When we talk about proving ROI through offline sales attribution it is important to look at audience addressability. Why?...Persistency and accuracy of the Targeting.Let’s talk about persistent targeting: The two ways to target against your matched audiences are through what is called a Persistent ID or through a user’s more temporary browser based cookie. Both conduct the upfront matching against your customer file, but because of the differences in how they identify your customer before serving your message, they have varying degrees of accuracy. The persistent ID is the more accurate of the two because it employs targeting technology that doesn’t rely on the user’s cookie. Publishers like Yahoo!, have users who registered with the site. In fact,150MM are addressable. That registered user relationship provides the publisher with a persistent connection to their audience. This is beneficial because: It avoids cookie churn. According to ComScore, 30% of users churn their cookies 5x a month Provides greater confidence that we are reaching a group of consumers not browsers who could be shared by more than one consumer Opens the door to multi screen delivery – PC, mobile phone, tablet since we know that registered user is access content across devicesLastly, allows us in a privacy compliant way, to measure offline sales as a results of the advertising those consumers see online. (transition to Linda on how DB match and measurement works)
Intro with the variety of ways we can use dataBut today we are going to focus on leveraging your data. How you can make data you already have today actionable. Yahoo! Is a trusted entity with consumer data and information that must be protected. The advertiser is the same. These two entities want to share their data and market to those consumers who are customers of both. However, as a protection to the consumer, Yahoo does not want the advertiser to know details about who and who is not a Yahoo! Customer so they never want individual information to go back to the advertiser, not even the fact that the advertiser’s customers matched. Likewise the advertiser does not want Yahoo! to know the full set of their own customer base or any data that is unique to one individual. A trusted third party, like Acxiom, who has been providing a safe haven for marketers to work together for years, can take both sets of data, and provide a “safe” place to match the data and allow for Yahoo! to reach the advertiser’s consumers with the confidence of knowing segment level information about that consumer with confidence and thereby delivering a more targeted and relevant message to the known audience. This is extremely exciting for the advertiser because with the combination of the significant consumer set at Yahoo! and the superior matching techniques at Acxiom, most advertisers using this service today are able to reach 30 to 45% of their intended audience with confidence.
Yahoo: Now…Take that one step forward. For the first time ever, the impact of a digital campaign can be measured in terms of the results that occur not only online but via other channels as well. You will see in the case studies we will share the incredible insights that are now possible. In an industry where 80% of all sales occur offline. It is important to understand the impact and influence of a digital campaign outside of the online channel. Describe process.
Now… Let’s take a look at a couple of examples to demonstrate the impact. Goal:Look at impact of online campaigns in combination with direct mail campaignUse offline insights to segment and target Measure results of online PLUS offline salesApproach:Two groups – Direct mail only, direct mail and online; Control groups for both Ran the campaignMatched the online and offline sales back to those that were served both and those that received only direct mailResults:Those that were part of either direct mail only or online and direct mail drove huge salesFor those that received both there was an incremental 11% per shopper averages sales liftOverall 4 to 1 ROI ad spend when
Problem:A large insurance company typically used direct mail to advertise to prospects about all types of insurance policies, e.g. car, life, home. The company was looking for alternative channels and wanted to target potential new customers at an individual level with the goal of selling those prospects insurance policies. Goal:- Analyze and understand the performance of individual segments and associated creative messaging. KPIs = incremental lift and return on spend measuring offline signupsSolution:Utilized prospect database for targeting home and auto insurance products. The prospects were divided into multiple consumer segments for targeting + test and control groups to understand performance. A specific creative was used to target each segment group on the publisher’s network. Joined campaign exposure activity from the publisher to the conversion file; reported back segment level results, including which prospects were exposed and which messages they received Results:- Matchback reporting was key since 97% of policies were purchased offline (for both test AND control groups). Just because we market to consumers online doesn’t mean they will respond via the same channelPositive return on ad spend (6.3X)The campaign drove $407k in incremental premiumsUsers exposed to the campaign were converted at a rate 2X more than the control
The tests are working. Now… do that co-ordinated channel targeting you have been thinking of doing for years and do it with certainty. Segmentation works. Take advantage of your existing segments in new channels. Use your offline models and customer market segment definitions. Use the tremendous amount of data available from interactive channels to enhance and optimize and extend your segmentationPartner to utilize 2nd party and 3rd party data and all types of data such as interest, in-market knowledge, and demographics along with your own customer insights to both reach your target audience with relevant messages but also to enhance that message and respond to differences in the online audience and EXTEND your reach to people who look like those you Vary your offers and creative for relevance Reach those consumers with certainty and measure the profound impact by measuring all results, not just clicks or conversions online. PartnerBring your offline and online organization together through working with partners Treat your data and the consumer’s data with care and work with partners that are known and trusted
Marriage of data and creative Creative and media buying decisions are often done independently. In early results, we’re seeing that the combination of audience targeting combined with data driven creative can yield up to 100% lifts in conversionIncreased frequency and depth of campaign insights Data used for pre-campaign planning & post campaign insights is common place today. Move toward in flight campaign optimization and improve campaign performance. Tools to inform overall media mix and cross channel attribution Historically it has been a challenge to understand how digital marketing impacts offline sales rendering comparisons to traditional direct marketing a challenge. With addressable advertising and more effective attribution models, it is becoming easier to understand which channels are most efficient at delivering campaignperformance.
The amount of data is staggering and the insights can lead you to incomplete conclusions. This year, let’s find the more complete picture of your marketing effortsThank you for your time today. Both Bryan and I are available after the session.
Q. What segmentation strategies are working best online? A. We've seen the strongest incremental lifts coming from the mid-tier customer value segment. We initially thought we'd see the strongest lift from the best customer segments, but given they are already buying and engaging with your brand – strongest improvements are consistently coming from the lower value customer segments. Secondarily, we have seen advertisers find solid performance in display programs that target those customers that have opt'd out of email. While this particular audience may find email as a less desirable communication channel, they are responding to messages in display ads.Q. What's the most efficient way to get started on a CRM data driven display campaign? A. The CRM team reaches out to their marketing services partner to talk about segmentation strategy, and partners with their media team for new digital budgets to test effectiveness.Q. How long does it take to get this kind of addressable display program up and running? A. If you already have your data with Acxiom (or other MSPs), you can launch a display campaign within 3-4 weeks.