2>3: Why Second-Party Data Is More Valuable Than Third-Party But Activated Less Today

The Sell Sider” is a column written by the sell side of the digital media community.

Today’s column is written by Ameet Shah, senior director of global technology, publisher and data strategy at Prohaska Consulting.

Marketers and publishers want access to more quality data to scale beyond their respective first-party data, yet they often overlook second-party data.

Sure, using third-party data is always an option, but it is often of an unknown quality and a commodity where the decent segments and individuals get overbought and overused. Second-party data is far more valuable but traditionally has been harder to activate and manage.

Data-centric ad tech solutions have not traditionally focused on second-party data, so the management of second-party data connections has been challenging. It can be difficult to find partners, assess data attributes, physically make the data connections and then figure out the financials. Those are certainly sizeable barriers that require deliberate efforts to overcome. For most, the trade-off has not been worthwhile, so far.

But ad tech and mar tech ecosystems have changed and evolved to the point where data giants, such as Google, Facebook, Amazon and Verizon, have vastly larger insights than virtually all other companies, inhibiting their ability compete effectively. Publishers need to rethink their data strategies and pursue second-party data relationships. Data sharing creates opportunities to more effectively compete in scale and diversity of data attributes.

Establishing second-party data requires more effort than third-party data, but focusing on partners, platforms and people can unlock this powerful data source.

Partners

The goal for each publisher and brand is to maintain control of their data as their unique advantage. I agree with that approach wholeheartedly. But including strategic data partnerships can be extremely valuable.

Data partners need to be selected to be complementary rather than competitive relationships. For example, restaurant chains and beverage brands may seek a similar audience to create brand preference. And a sports site and men’s lifestyle site may share common audience demographics and interests. Existing partners should also be considered. 

Platforms

Advancements in ad tech have improved the capabilities and reduced complexities in leveraging second-party data. Data platforms, including data management platforms (DMPs), data exchanges and data lakes, can facilitate and simplify considerations for previously unobtainable opportunities.

Companies should work with their existing data platforms or establish relationships to explore opportunities. Solutions can make it easier to establish partnerships, maintain controls over data access, manage financial revenue shares, build data segments with data from multiple companies and distribute them to various platforms for targeting activation.

Despite how these data platforms have simplified activation of second-party data, it is still minimally used. Fortunately, the days of using a DMP only as a third-party open data exchange and to duplicate auto-intender segments are numbered.

People (Leadership)

Too often, publishers do not consider partnerships because they are instead focusing on competitive differentiation. However, differentiation is difficult to achieve given the scale of the industry giants. Establishing data partnerships actually increases publishers’ overall data capabilities. This improves their ability to sell targeted campaigns and more effectively target marketing campaigns. The value is clear but the time investment to foster these relationships is overlooked.

Activating second-party data also requires companies to have individuals who are tasked with data relationships. It can’t be the ad ops contact who is already building and managing segments and campaigns; it should be someone tasked with taking a strategic focus on partnerships and relationships.

These partnerships should lead to data co-ops (or data consortiums). These collaborations often already occur for other activities and collaborations, but not data-related ones. I know of one fantastic marketing data co-op run by a top travel marketer for more than three years with 10-plus noncompeting marketers. There are also industry associations, industry groups and conferences where like-minded companies share and collaborate.

However, this has not translated to many data-sharing deals yet, or maybe those aren’t the forums to produce data alliances. Having a company take a leadership position and reach out to partners will lead to benefits for everyone.

The First Step

Second-party data is a valuable, high-quality data source that will improve a company’s data-targeting abilities and generate more scale to capture higher spend and deliver more scalable results. Focusing on people, platforms and partnerships can increased success for publishers and brands. Taking the first step is the hardest but often the most important one.

Follow Prohaska Consulting (@TeamProhaska) and AdExchanger (@adexchanger) on Twitter.

3 Comments

  1. Some months back, Chris O'Hara of Salesforce - Krux made similar observation about the potential value of second party data linkage, along with operating issues to be addressed. From my own experience, I'd just note that firms getting involved should treat these arrangements as formal, defined projects, and have them manged accordingly by professionals with the expertise and experience to do so.

    Reply
  2. Well written. Advertisers have been very caught up in the convenience of third party data and have failed to recognize that it is directional at best. The closer you can get to the data the better and 2nd party data is a fantastic way to target accurately although scale and convenience may be less of an advantage.

    Reply
  3. Ted McConnell

    Ameet. Nice, and a timely call to action. I agree with the point, but I will quibble with a couple of ways you got there. 1st party data has more than reach issues, it has recency issues, and, for lack of a better term, inference issues. For example, most people who go to CPG websites (in my experience) are looking for a coupon. They are not there because they are in loyal, in fact, its the opposite. Also you question the quality of 3rd party data. You should, but the quality issues with 3rd party data are not usually because the data, by itself, is bad. What is bad is how we infer meaning for use in a campaign. For example, if a 3rd party dataset says "people who looked at content about a new TV", they probably did that. There are reasons the data might be compromised, but the dataset will be reasonably dense with people who did that. If its not, that's robbery. But, where it falls down is how we use it. That dataset may very well contain people who are in market for a TV, and it is likely to be used in a campaign selling new TV's, and it possibly should be. However, it will inevitably be vastly less dense with "in market" people, for a lot of reasons. They only answer is to measure, and to understand the relationship between the behaviors that generated the data, and the attitudes we infer upon those same people. What you find is, behaviors only weakly predict attitudes. The key to the deals you suggest (and its a great suggestion) will be not which Party, but the portion of the audience dataset that is likely to yield if exposed to an ad. The job of the industry is to provide that measurement.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>