Here at Beckon we often talk to marketers about what we call the “Iceberg Challenge” of marketing data. As any North Atlantic ship captain knows, the most dangerous part of the iceberg is the 80% of it that lies unseen below the waterline. You can steer your ship around the part that you see, but it is the mass you can’t see that sinks you.
The reason that this is a great metaphor for marketing data is that most marketers only focus on the top “above the waterline” part of the data iceberg that they can see—the visualizations, dashboards, infographics, analysis and insights that are the visible part of any marketing performance management program.
But just try to slice the data in a way that the report doesn’t show, or drill-down into a chart to get more granular detail about a particular metric, or try to combine data from two different sources and what quickly happens is the analysis gets “sunk” by the data challenges that lurk just below the pretty pictures.
Below the waterline is a complex and confusing collection of data from many different applications, in different formats, in differing levels of granularity. They have inconsistent naming, different owners, different levels of security, and overlapping measures. And adding to the challenge is the fact that there’s no automation, no search engine, no index to help us navigate the mess. In short, the manual process of collecting, cleansing, and organizing the useful data out of the mess below the waterline is sinking us.
In fact, when asked about the particular tests we face when it comes to new marketing technology, the integration of increasingly fragmented data is the one we cite most. Fifty-four percent of us say it’s a challenge (CMO Council Quantify How Well You Unify, Oct 2014).
Marketing data management today is a complex discipline and as we add new channels and new ways of collecting data it’s only getting more complex. So what’s to be done? How can we get our arms around both the top and bottom of our marketing data iceberg? From our experience here at Beckon implementing Marketing Data Management programs, we have uncovered these great best practices in solving your iceberg challenge:
1. Get all marketing data into a single system. Make sure there aren’t other icebergs floating around just below the surface that can cause you a lot of pain later.
2. Use consistent KPIs and standardized reports. Don’t look at apples in North America and peaches in EMEA. Don’t asses the Mother’s Day Campaign by the awareness it delivered, the Back to School campaign by the customers it acquired and the Holiday campaign by the sales revenue it generated.
3. Store only the data that matters. It’s ok to leave some stuff on the cutting room floor. Use the KPI framework you developed in 2 above to guide your decisions about what data is relevant.
4. Clean up the mess on the way in. Develop a marketing taxonomy and use it to tag marketing data on the way into your data hub by channel, region, product line, campaign—any way the business reports on itself.
5. Use standard definitions. Make them up yourself if need be, but you have to standardize in order to benchmark.
6. Automate data collection. Manual processing of the data will never lead to the scale necessary to effectively use your data. Work to collect, normalize, standardize and categorize your marketing data with automation (see: The Case for Automated Marketing Data Management).
7. Make data management a mission-critical process. Your marketing data is one of the most strategic assets your company has—treat it as such! Make it as important as your finance data.
8. Invest in proper resources and systems. Like any other mission-critical process, you need to have the right people, processes, and technologies in place to get the most value out of your data.
By implementing the eight best practices above, marketers can get well along the path towards a successful program of Marketing Performance Management. By cleaning up the marketing data mess below the waterline, the visible top of the marketing data iceberg—the visualization layer—can then begin to be a reliable and scalable part of campaign optimization, planning, and management reporting.