When done properly, attribution programs can live up to the hype and deliver outstanding business value and insights. But for every success, there are also failures. Here are a few of the most common pitfalls to adopting an attribution solution.
1. Lack of buy-in at all levels of the organization
One of the primary reasons attribution implementations fail is not due to the data or technology. It’s due to a lack of commitment to a new way of thinking and operating your business. Let’s face it—marketing attribution data can be scary. Why? An attributed view of performance is always going to look worse than a traditional program-centric view. What does that mean for the channel manager in defining success? Or for hitting performance targets? Year-over-year measurement? You get the idea. It also changes the way programs need to be optimized. How do you tell an affiliate that you just learned they’re not ROI-positive and you need to cut their commission rate? These things are hard but necessary to leverage the data and optimize your business.
2. Failure to understand the requirements to implement and maintain the solution
Attribution implementations run the gamut from relatively easy to set up to being very complicated and cumbersome. We’ve seen setup times extend out past six months or more, which can be painful if you’re paying for a system that’s not providing any value. Maintaining custom parameters across all of your marketing links can also be challenging. Without a good plan in place, you may be surprised at the ongoing investment of time and quality assurance that’s required. Evaluate the complexity of any solution against the robustness of data output to determine the sweet spot for your business.
3. Adoption of overly simplistic modeling methods
There’s no shortage of marketing platforms that claim to offer an attribution solution. Most are relatively lightweight, don’t fully capture your data, and offer the simplest of modeling options. Last-click is the default method and a good place to start. It’s at least better than overstating your marketing performance via disparate program-centric views. Other solutions extend model options to a business-rule-based first-click or weighted distribution. Adding more sophistication, impressions are factored to capture influence of display marketing. The most robust offerings calculate true unbiased attribution rates—inclusive of every channel, display impression, and device type—by applying statistical methods to the raw data.
4. Inability to define and measure success criteria
The value of an attribution solution can be difficult to measure and can often stump executives tasked with defending budgets. Here are a few ways to derive value: First, you have a completely new set of reporting that tells a different story than your web analytics or your program-level data. Shifting the basis of your measurement and reporting alone can be a powerful change. Second, you have the ability to act on that data to answer questions and make better investment decisions. What’s the incremental value of spending media dollars on your brand keywords? How are display ads tangibly influencing consumers to purchase? What are the benefits of affiliates who use coupons to drive traffic? Third, you can leverage the data to perform sophisticated media mix modeling, informing optimal marketing spend, and distribution. Want to be cutting edge? Inject granular attribution data into your marketing optimization platforms, providing an entirely new data set to optimize against. All of these benefits then need to be weighed against the direct and indirect costs of the solution.
Careful consideration of these factors in the early stages of multi-touch attribution program development will give you a much higher likelihood of success.
This article was originally published on Chief Marketer.