Checks And Balances Are Key To Clean, High-Quality Data

Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Rob Finora, senior vice president of data at ShareThis.

Marketers now use data-driven initiatives as the backbone of their campaigns more than ever. But with this boom comes the exponential concern about data transparency and quality, which are both critical for campaign success and the bottom line.

The growing need for transparent and quality data is one reason for Marc Pritchard’s rallying cry last year for full transparency. Inventory quality continued to take center stage with the launch of Ads.txt to prevent unauthorized inventory sales. When you add machine learning, which requires a massive amount of high-quality data to be effective, those accuracy and quality factors become more critical. While many in the digital media supply chain responded and much progress has been made to solve the industry’s transparency issues, data quality remains a top priority.

To ensure data quality, marketers must make sure checks and balances are in place within their organizations. The talent they hire and the tech stack they create are the support beams for their data.

I wish it were as simple as separating the wheat from the chaff. It’s not black and white. There are quality factors, such as sources, blending and expiration issues, that impact data effectiveness.

Marketers’ ROI is tied to data cleanliness, but data includes tons of noise that is challenging to cut through. We need to reduce this data noise, but this is where it gets tricky. One marketer’s trash is another’s treasure. 

Before we go further, I am imploring marketers to get back to the basics.

The data boom is similar to black gold. Oil is dirty when it comes out of the ground. So is data, until it’s refined.

Oil is refined differently for different applications, as is data for insights, analytics, segmentation, optimization or measurement. Unrefined or improperly refined data is often confused with bad data. The way to address data transparency and quality concerns is to understand and have a process for its refinement with very clear success metrics and goals. That process will be determined by a brand’s team, systems and data.

Brand marketers must have a solid foundation that begins with a customer journey plan. This goes beyond the media plan, audience definition and execution. A customer journey plan ties the data to the overall customer acquisition and maintenance strategy. Once a plan is in place, the brand can take an inventory of its assets, beginning with its team, systems and, ultimately, data.

A brand’s team is its best offense and defense. It will likely be a blended team as most marketers do not have these resources in one place, so they should include their agency, vendors and sometimes consultants. The experts they need must be able to address the full process collectively. This includes big data experts who can focus on normalizing disparate data sets and marketing experts who understand how data is collected and what specific measures mean for the brand and its customers, as well as programmatic and analytics experts. If data will fuel an outcome, brands must ensure that they have an expert to guide that journey from the start.

Next is infrastructure. It starts with how brands structure their CRM files, capture and store their first-party data and leverage their data hosting and management platforms, demand-side platforms, media agencies and other analytics or measurement resources. Build it, buy it or rent it, there are arguments for each path, but if brands are investing in data, they also need to invest in their infrastructure.

There is almost a one-to-one ratio between data used and data infrastructure costs, according to the IAB and Winterberry. Marketers must create a continuous learning loop with AI and machine learning resources. AI is not a panacea but it should help make the continuous learning loop smarter over time.

Finally, the data. We all agree that first-party insights are usually the most valuable resource, but there are often holes in strategy and execution, and there are hundreds of resources for supplementing first-party data to help scale its usability and reach. The key here is that general data transparency helps brands find the cleanest data to optimize toward it so that they can fine-tune the learnings that emerge from a mix of well-curated audience insights.

In the end, it is the people and process that allows marketers to sift the data and decide what is quality and critical for their use. But as we explore data transparency and quality, we must not lose sight of the larger transparency issues that plague the industry before any data is touched.

Follow ShareThis (@ShareThis) and AdExchanger (@adexchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!

1 Comment

  1. Before you can clean data you often need to know what it “looks” like as well. Using visualizations and breaking down the data to find what needs to be “refined” is great approach to get an idea of how to get started. Also, even though knowing your sources is key, having data from multiple sources or perspectives can help build a strong backbone of reliability.