Caution: Read only if you have a stomach for big, and I mean REALLY BIG data, and revolutionizing how your company does attribution.
What if your company had something magical for each of them, but a single message wouldn’t appeal to ANY of them? In this blog post we talk about what’s involved in people-based measurement - why it’s important, and how it can be done. Digital advertising introduced the ability to track and measure each individual’s activities -no longer limited to market or channel-level decisions. Wherever a pixel could be fired, marketers could now implement a “lift” test to understand the effectiveness of a given ad impression or series of actions. Massive performance improvements became available - for digital. But this granular “people-based” measurement hasn’t been feasible in environments that don’t fire pixels – such as television, magazines, or billboards - until very recently. This has meant that marketers have had to maintain two separate models – a digital attribution model, and media mix modeling. Maintaining both was not only cumbersome, but how were marketers supposed to reconcile, or use them together?
Conclusions were often contradictory, didn’t account for interdependencies between each other, and didn’t align in any useful way. Today, people-based measurement across all media channels is a reality for those using Unified Marketing Measurement platforms (the category first coined by Forrester in 2016), and is in reach for those who aren’t yet. Our research shows that 30%+ gains are available to customers who embrace people-based measurement and active optimizations -soon, person-based measurement and optimization will be table stakes and necessary for a marketer’s survival. So what enables people-based measurement across all channels? First, you’ll want to have sales transaction and brand lift results available at a person-level. Sales data can come from your company’s CRM file, or your attribution platform may be able to acquire it from any number of third parties who measure transactions, or proxies for transactions such as store visits. If you don’t own your customer files, this data doesn’t come cheap, but it’s increasingly available in most categories. Having both sales and brand lift data will be important for marketers who want to understand their spend holistically, and who want a “leading indicator” of success to enable nimble optimizations.
Next, you’ll want a way to deduplicate all digital exposure data. Attribution providers with deterministic user graphs are able to map cookies and mobile IDs back to a single person – what previously may have appeared to be 4-7 different people, now will appear as a single person after digital deduplication. The user graph can also address challenges of cookie deletion – a very real problem given that 20%-40% of users are estimated to delete their cookies on a regular basis (giving the perception that twice the actual number of users have seen your ads)! Increasingly, exposure data for non-digital channels such as TV is becoming available, too. There are several sources of TV viewership data including TV manufacturers, set-top box services, or companies who provide software to those services. TV exposure data is usually provided as IP data which is then translated into person or house-hold -level data. Location-based data is increasingly abundant, helping understand even billboards’ effectiveness at a person-level. Several panel-based methods that allow marketers to understand a user’s exposure in formats such as radio and print, then extrapolate those results to a larger population. The best marketers are as focused on measuring creative at person-level as they are media, as they know that different creatives work differently for driving different responses for different people. Marketing Evolution’s term “Bulls-Eye” measurement for example, refers to the ability to understand optimal frequency for a specific creative message, for a given person or audience target, at a specific media property or combination of media properties. This is the type bottoms-up measurement and optimization that seems natural to many digital natives, but is mind-boggling to traditional marketers who are newer to person-based measurement.
Lastly, person-level measurement benefits from the addition of demographic or behavior data, weather data, proximity-to-store, allowing marketers to get more granular than ever before in understanding what drives success. We never said it was easy: Making sense of all of this data requires a LOT of machine learning and technology to understand what drives results. But you don’t have to worry about that – you just need a vision and passion for more accurate measurement and optimization.
© 2019 Marketing Evolution, Inc.