3 Ways to Use Big Data for Revenue AttributionOctober 3, 2012 By John Wallace
In a complex and shifting landscape, where is a marketer to start? Good question. Instead, here are some places you shouldn't finish:
- Don't stop the analysis at the matchback stage. This is where double counting is introduced and where probably 80 percent of direct marketers get stuck;
- Don't make up rules on how to attribute revenue (7-day, last click, last touch); and
- Don't disconnect your attribution from your targeting engine.
The next step is to assemble several types of data and resolve to analyze them:
- Behavioral data
- Offline matchback data
- Online matchback data
- Contact history (search, display, catalog, DM, email, affiliate, etc.)
- Demographic/overlay data
Yes, this is Big Data. With the new tools available today, marketers are finally able to harness the power of their data through the use of predictive analytics, allowing them to optimize their marketing spend and attribute revenue to the right place. Knowledge is power, and every website click, page view, ad impression purchased/served, email sent and order taken—be it in-store or online, via mobile shopping or through a call center—can now help marketers improve every step of the decision-making process.
Through the use of powerful and scalable predictive analytics solutions, top marketers have learned a few things:
1. Google is Greedy About Taking Credit
Google search has been the darling of the online advertising world for years, but what most marketers don't realize is that Google keywords perform much worse than they think. With traditional attribution techniques like last click, Google receives a disproportionate amount of credit because the search happens near a purchase. Many customers are already trying to find your site or a specific product and only using Google as a navigation tool. A comprehensive technique looks at all of your marketing and customer activity and then assigns credit to multiple touches, rather than an all or none approach. It also understands time, knows that each marketing treatment has its own effective window and uses statistics to discern the value. One major retailer didn't believe the results and tested this by turning off search. As a result, the marketer was able to reallocate millions of advertising dollars out of search and into other marketing channels. Great results are still possible through search, but you need the proper tools to do so.