Coming Clean: 4 Steps to Resolve Martech’s Struggle for Data Transparency

We’ve all heard the martech industry leaders across advertisers, agencies, data providers, platforms and publishers rally for increased transparency. But is the industry listening? More than half of advertisers still note data transparency as a significant programmatic challenge, according to eMarketer. A recent Forrester study found 75% of marketers report a lack of transparency into data sources when it comes to their data providers.

These percentages are really no surprise when you consider that a marketer asked to build a campaign that targets “auto intenders” for an automotive client’s latest model launch will yield hundreds of search results. How is a marketer expected to choose the best segment?

Marketers want access to more information so they can make the best decision for their campaigns. But in order for a marketer to know which of the hundreds of thousands of segments will drive their campaign KPIs, we first need to create standards, common definitions and a verification process to ensure adoption.

I’ve seen many opinions on the state of transparency; most agree it needs improvement. But I’ve seen less ideas on what we must do as an industry to instigate a cleanup. With marketers expecting adjustments be made in the next two years, I believe we must move faster.

To that point, I recently joined the DMA’s Council for Data Integrity, a cross-sector martech group committed to increasing transparency by simplifying and standardizing the complex lexicon behind identity in order to enable better  understanding and trust. It’s a process similar to how the Food and Drug Administration (FDA) standardizes food labels. We have a lot of work to do, but there’s a agreement across the board members that it starts with these three steps.

1. Take Action

First, industry constituents need to actively participate in change, and we’re starting to see this. The DMA’s Council for Data Integrity just had its first meeting in early March, which featured a lot of genuine discussion on how to move forward as a more transparent industry.

I want to note other great forms of active participation from other groups as well, such as:

    • Other formal groups like the Trustworthy Accountability Group, a collaborative effort from the American Association of Advertising Agencies (4A’s), Association of National Advertisers (ANA) and Interactive Advertising Bureau (IAB)
  • Platform-initiated programs like:
  • The Adobe and Appnexus effort to reveal all DSP fees

All of these examples of active support are meaningful, but a conscious, concerted effort is just step one.

2. Agree on the Facts

The next step—actual simplification, standardization, guidance, verification and education—is more complicated. But it starts with what most of us can agree on—the facts.

Let’s return to our marketer running her automotive campaign and the hundreds of audience options that show up for her search for “auto intenders”. It’s unclear from names and descriptions the different classifications or characteristics of each option. Is an audience:

        • Based on third-party service records
            • Declared audiences
            • Modeled audiences
            • Based on profile data
            • Based on cookie intent data
            • Probablistic
            • Deterministic

When armed with this information, the marketer can better understand the different options and select an audience that best fits the marketing objectives and KPIs. For example, if her campaign’s goal is brand awareness, starting 2 months prior to the launch, she may need a modeled audience based on profile data in order to reach a more expansive group of prospective buyers. If she’s launching a campaign aimed at sending users back to the auto brand’s website one week before the launch, she may prefer to use an audience of users who recently searched for the auto brand online.

I think groups like the DMA’s Council for Data Integrity are a great opportunity to develop and agree on common industry names and definitions, much like the FDA defines ingredients used on food labels. A standardized vocabulary will enable the next step.

3. Build Trust with Verification

Once standardization and adoption is in place, the ecosystem will have a common set of definitions and companies will have the foundation to self-police. This will be a good first step to rebuilding the trust between marketers and vendors.

Quality—It’s in the Eye of the Beholder

When I hear industry constituents discuss transparency, I often feel the conversation is more geared toward quality. Quality is important when it comes to data transparency because the most high-quality data will inform the best models, campaigns, and, ultimately, ROI.

Quality is the most difficult component dig into because it’s subjective. For example, would our marketer say that the audience based on behavioral signals or offline profile data is higher quality? Does her answer change if the data is deterministic or probabilistic? Or does it depend on when each dataset was last refreshed? Would the data vendors and platforms have a different standard of quality?

However, once definitions are standardized, we’ll see more parties advocate for specific data types, sources, modeling methods and so on, and develop educational tools to drive understanding and adoption in the marketplace, much like how groups such as the Beef Board and the National Pork Board invest in marketing to encourage consumers to buy their products.

Advocation will happen in place of a formal quality metric to drive industry trends. Just as the definition of “healthy eating” has changed over the past decades to encourage more organic and antibiotic-free, for example, I think we can expect to see data advocacy groups push for industry changes—moving from segments to people-based audiences, or from cookie-based to profile-based data, for example.

Of course, there are details that will lead marketers to make a qualitative decisions. But while quality may seem like the best metric, there’s a lot more to unpack when it comes to defining the best ways to recognize, define and measure quality. Stay tuned for my next blog, which will delve into the issues of how to measure quality. In the meantime, marketers must be armed with the right information first, which is why standardization, education and verification are key to building and increasing transparency.

David Dowhan

About David Dowhan

As CEO and founder of TruSignal, David leads business strategy and product development efforts and is keen to create innovative, customer-centric solutions using data and analytics.
This entry was posted in Articles, Data & Audiences and tagged , , , , , . Bookmark the permalink.

Comments are closed.