Data is everywhere. But what about quality?
Much has been said of this data revolution that is taking place — that “data is the new gold” of the media, advertising and technology industries. In fact, there isn’t an area where so-called “big data” isn’t a talking point, be it for business analysis and strategic planning, pre-ad campaign planning, in-flight campaign optimization or post-campaign analysis, with actionable insights gleaned from the data smattered in between.
In fact, media buyers are getting more familiar with using data in their media planning and buying process, with the Global Alliance of Data-Driven Marketing Associations (GDMA) & the Winterberry Group finding that 91% of U.S. marketing and ad professionals segment data to target their consumers. The research and insights yielded from this data are then being used to fine-tune (or “optimize”) media planning so that a return on investment (ROI) can be determined and benchmarked.
However, as brands and marketers get more comfortable using data and as more companies come on board with new data sets and refinements to already existing data sets, how will the consumers of these data overlays be assured that what they are paying for and ultimately implementing is of good quality? Does the product work as advertised? Do the marketing claims align with the results? Let’s think about this concept for a moment, applying it to a real world situation…
If you were in market for a new car, would you purchase an automobile if you knew it was going to break down right after you drove it off the lot? Of course not! There are lemon laws and reasonable expectations of a warranty when you buy a car, so that you as the customer don’t get burned if the car breaks down right after you drive it off the lot. Applying the same logic, why can’t we have the same sort of guarantees for data as we would for car buying?
At the present moment, the scary thing to consider is that any entity can claim they have some sort of worthy data and load it into a DMP/DSP, making it available for targeting for advertising; it isn’t necessarily vetted for quality or meeting any standards. Plus, you as the consumer in most cases don’t get to open the “black box” and see how the data is sourced, collected, modeled or put together, so how do you really know what you are getting under the hood? Wouldn’t you think we as an industry should have our own version of a “lemon law” for bad data? Quality is key to long-term health of any business, and data is no exception.
To prove my point, Experian did a study back in January 2015, dubbed The Data Quality Study (which can be accessed here) and found that 32% of organizations found that their data was inaccurate. You heard it right — almost a third of companies with data had a sense that it was inaccurate on some level. To me, that seems like too high for a rate of bad data.
In the same breath, Experian found that 97% of companies turn to first-party data for business insights and that virtually all (99%) of businesses view data as being essential to marketing success. The fact that virtually all companies use data for insights and for gauging marketing success, yet close to a third think the data is inaccurate in some form shows that there are major gaps that need to be addressed.
At the present moment, we as an industry are tackling a myriad of larger issues such as fraud/non-human traffic (NHT), ad blocking, native advertising disclosures, improved measurement (both in terms of overall measurement quality and cross-media measurement) and viewability. From my perspective, as these problems are winnowed down, figured out or otherwise mitigated, the next frontier will most likely turn to a focus on data quality. But what is the answer to the data quality issue? How can data companies preserve their trade secret while providing some modicum of transparency to their customers? The answer is simple and consists of two words: Data Audits. What do I mean by this and why is it necessary? Allow for me to explain…
Quality standards in data and research products span back to the early 1960’s in the television business, which faced negative backlash from the game show fixing scandals by advertisers and networks (watch Robert Redford’s 1994 film Quiz Show if you want some history on this debacle). As a result, the entire TV business at that time was marred under a cloud of untrustworthiness and the U.S. Congress intervened, looking into not just TV network practices, but also the practices of research firms that provided audience guarantees to the networks, ultimately leading the Media Ratings Council (MRC) to be established at the behest of Congress (the MRC continues to operate today, but as an independent, standalone non-government entity). Since then, the MRC has conducted audits for just about every research and consumer insights tool, ranging from the Nielsen TV ratings system and many other products. More recently, they have taken the lead on drafting standards for ad fraud/non-human traffic (NHT), viewability and location data.
Given the push for transparency and a desire for universal standards, my belief is that you will see brands, agencies and marketers rally for data audits on companies who provide data to them. So why not have the Media Ratings Council (MRC) and/or theAlliance for Audited Media (AAM), two independent bodies that help to set and uphold quality standards across the industry, come together with interested parties, draft standards and begin data provider audits? Or, why not involve the recently formedTrustworthy Accountability Group (TAG) get involved as well?
While the aforementioned is an ideal scenario, until we have formal standards in place, there are some basic tenets you should probe when employing the use of data in campaigns, consumer insights and/or media planning. Ask yourself the following:
- What is the methodology and can I understand it? Are you able to clearly understand how the data is being sourced, collected and processed? Is it being done in a valid, effective and reliable manner?
- Are there any privacy compliance issues that might arise? This is key when dealing with data that contains PII (Personally Identifiable Data).
- What other modifications are being made to the data set and why? Are you able to understand if other data is being appended, or if the data is being modeled? Do you have a reasonable understanding of the sample sizing, base and/or reach? What about benchmarking and norms?
From my point of view, data audits are the only way we can ensure that the data supply chain meets quality standards and is fully transparent to all parties involved — to the marketer, the agency and most importantly, the customer. If everyone has a reasonable sense of what they are buying and can be confident that it has been independently vetted for quality and standard conformity, there is more trust for the ecosystem. As the old saying goes, “a rising tide lifts all ships,” so let’s hope that a focus on quality will help the data industry grow and flourish.
Karl Brautigam is a brand strategist and MBA student with expertise in the media and technology industries across the sales, research, marketing and strategy functions.
