The challenge: when a user is involved in multiple campaigns across multiple sessions on different devices, how do you know which of these different touchpoints actually influenced the sale? Thankfully, modern tooling can solve this problem across teams. The problem is, these types of attribution cannot deliver the deep insights necessary for a holistic, cross-channel understanding of marketing performance across the whole consumer journey from the first touchpoint to the last click. Because those are not core skills in marketing, most marketers fall back on last-touchpoint models or outsource complex attribution to third-party attribution tools that don’t have a complete picture of the attribution data. The first (comprehensive data collection) is a data engineering problem and the second (statistical modeling) is a data science problem. Neither of these marketing measurement challenges should be addressed by the marketing team. First, it involves collecting and normalizing an enormous amount of data from a lot of different sources, and second it requires the application of statistical modeling techniques that are typically outside the skillsets of many marketing departments. To understand which marketing efforts are contributing more to a successful objective (conversion event) we have to evaluate it relative to all of the other campaigns, and this is very complicated. Instead of binary outcomes (this digital marketing campaign brought someone to our site or it didn’t), a user can have many marketing touchpoints, which introduces the idea of influence along a buyer's journey, and that’s where things start to get tricky. In short, moving beyond basic single-touch attribution introduces a number of complexities. Even so, advanced attribution is still hard. Today, though, data teams and marketers have more data than ever, increasing their appetite for more advanced attribution solutions. These limitations are due to the fact that many ad platforms were built when basic marketing attribution was enough for most companies. Upon first glance, it’s fairly straightforward to review metrics in ad platforms like Google ads to examine ROI on paid ad campaigns, and you may even be sending conversion data back to that platform, but you’re still restricted to evaluating only a fraction of the user interactions (one or two steps of the conversion path at best) that led up to and influenced a sale. The answer to this age-old budget allocation question boils down to determining which campaigns are working and which aren’t, but with the amount of data we collect on the customer journey today, it’s not always crystal clear which campaigns are which. Where should we spend more marketing budget? An overview of the architecture, data and modeling you need to assess contribution to conversion in multi-touch customer journeys
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |