Creating Unique Data Assets to Drive Marketing Predictions and Actions The Second-Party Data Advantage 1
Industry Trends The second-party data movement has quickly emerged as an innovative solution for marketers looking to get smarter about identifying unique, high-value audiences at scale. Marketers have more data than ever, but extracting unique insights ones that provide a competitive advantage and drive meaningful action has become both increasingly difficult and important for brands. As third-party data becomes a commodity, brands are more and more turning to proprietary assets to distinguish their audience outreach efforts. First-party (marketer owned) and third-party (externally acquired) data remain relevant, but private second-party data sharing gives them an extra edge. This whitepaper provides marketers with background information, tactical uses cases, and sample performance of other Modern Marketers as they create proprietary data assets from second-party data sharing and add look-alike modeling into their marketing mix. 2
The Challenge and the Solution Since the emergence of data-driven marketing, first-party data has been the primary currency and the gold standard brands relied upon when targeting users. Why is that? For starters, first-party data is free and 100-percent unique to the marketer, and it provides signals that are not available to competitors. For example, first-party data would include the date a user s subscription is due for a renewal. The marketer knows this is a primary time to message and the user may be more at risk to being lost to a competitor. With this proprietary intelligence advantage, the marketer can message and keep the customer. In addition to its proprietary importance, marketers use first-party data because it s easy to access. For example, marketers can generate first-party data from: Website Media Social Video Search Email Direct Mail CRM Prospect Database Mobile On the other hand, third-party data provides the scale that marketers covet while offering a myriad of targeting options including demographic, intent, and interest attributes. But gaining access to that scale comes at the expense of uniqueness. Because that knowledge is available to the entire ecosystem, its value is reduced. Second-party data essentially creates a truly integrated co-marketing ecosystem that puts a premium on relationships marketer to marketer, publisher to publisher, marketer to publisher and teaches brands how to better execute one-on-one deals in a cacophony of noise. Question: What type of 1st party data do you use? Data: Uniqueness vs. Scale Website Data CRM/Registration Data Digital Campaign Data Search Data Social Data Mobile Site or App Data Email Data Lead Scoring Data 14.95% 24.91% 41.64% 52.67% 44.48% 55.16% 64.41% 0% 100% 77.22% Uniqueness Highly Unique 1 ST 1st Party Data Credit Cards 2 ND 2nd Party Data Hotel Chain L A Look-alike Models 3 3rd Party Data Expedia sold to Citibank RD From a recent study of over 280 respondents, marketers share the types of 1st party data they rely on. Scale Massive Scale 3
Top Five Areas to Discuss in an Exclusive Data Sharing Partnership 1 2 3 4 5 Paid or trade. The value of an audience can vary widely based on the type of data (demographic, geographic, or purchase intent) to a specific vertical (auto, travel, financial services). Discuss data pricing that makes sense against your partner s conversion goals. This may take some time to refine, but having a realistic starting point is ideal. Perhaps, a data trade with no money exchanged is the best approach. In that case, the most important discussion should surround the details of the trade of value between the two companies. Level of access. Are both parties providing data for respective audience profiling or is it one way? How deep is the information provided and does the level of data satisfy the need for audience targeting or profiling? What other partners are privy to this data? How long can you have exclusive access to the data? As a marketer sharing your private data, you may consider reserving a deeper level of targeting attributes for elite customers, or offering it as a paid option. Compliance with consumer privacy policy. Ensure that both parties data-sharing practice is clearly outlined in respective privacy policy language. This is mandatory for the company from which the data originates. Win-win offers. Once you have the data sharing defined, be sure to discuss the details of mutually beneficial offers that satisfy the business objectives of both parties. This is especially important for non-paid relationships. In the example of airline and credit card companies, the joint parties may define a special offer that drives credit card applications from business travelers, where one of the perks is special discounts for the airline. How to measure success. Early on, set a mutual benchmark to define success. Is it mutual customer acquisition? Is it successful cross-sell or in-store purchases? Determine your success criteria mutually particularly if it impacts payouts or data use. Test to refine pricing, offers, or program details. Electronics Company Bank Publisher Telco Software Company Retailer Social Platform Hotel Chain 4
Look-Alike Modeling to Create Proprietary Assets Another strategy for creating proprietary data assets is custom look-alike modeling. How does that work? Simple. Identify the audience you want more of such as purchasers or converters and create a custom model that identifies people who look like them. This works for all audiences that you want more of, including site converters, form submitters, and frequent visitors to new product pages. Look-alike modelers take the world of third-party and first-party data attributes, and layer on algorithms to find new users who display these key behaviors. Through a custom look-alike model, you can expand audience targeting or discover new personas that you may not be exploring or messaging to as part of your overall targeting strategy. And the best part: this look-alike model is now your custom dataset. Because you used your own signal to create the model, no other competitor can replicate it or identify this audience. Here are two real-world examples where look-alike modeling was leveraged to increase campaign performance: Software Company A top software company was looking to efficiently increase sales and improve the ROI from their display advertising campaigns. To effectively increase audience scale, they created look-alike models based on their top customers and targeted these consumers with their media buys. Result: Look-alike model targets increased ROI by 118 percent, outperforming the campaign average by 64 percent and the control group by 104 percent. Telecommunications Carrier A top US telecommunications carrier found that although their typical customer converted more in store than online, much of their in-store traffic could be directly attributed to visitors of the store locator page on the company website. To effectively prospect and drive traffic to the store locator page, the company leveraged look-alike modeling to generate audience segments based off of their store locator page visitors. Result: A 0.78 percent and 0.84 percent increase in traffic yield quarter over quarter, surpassing their yield rate by 40 percent. 5
Conclusion Marketers are constantly clamoring for ways to gain better access to intelligence that can help them stay one step ahead of their competition. It s the survival of the fittest data. So how can Modern Marketers gather those unique insights that are going to help them better predict user behavior, create ideal customers, and increase revenue? Through more data and better algorithms. By forging unique second-party data relationships and using look-alike modeling, marketers are breaking new ground in the area of data-driven marketing and creating proprietary opportunities for themselves. A central audience data strategy will enable you to combine data, algorithms, and a custom second-party partner network to derive the best actionable intelligence on your customers and prospects. 6
About Oracle Marketing Cloud Modern Marketers choose Oracle Marketing Cloud solutions to create ideal customers and increase revenue. Integrated information from cross-channel, content, and social marketing with data management and dozens of AppCloud apps enables these businesses to target, engage, convert, analyze, and use award-winning marketing technology and expertise to deliver personalized customer experiences. Visit oracle.com/marketingcloud Copyright 2014, Oracle and/or its affiliates. All rights reserved. 7