• by Drew Sollberger , Columnist, July 17, 2018

Calculating attribution is one of the most challenging tasks of modern marketing. Although we have access to more data than ever before, coming up with a plan to accurately measure which marketing tactics are generating the most results is a lot of pressure. How do you make sure you’re getting it right? 

There’s a lot of information about what you should do when adopting new attribution models, but sometimes it’s what you don’t do that affects your outcome the most. To help you successfully launch a more accurate way of measuring attribution, check out these five don’ts.

1. Don’t rely on last-click or first-click.

For many years, marketers have relied on clicks as an indication of digital marketing success and assigned full credit to either the last asset or the first asset someone clicked before converting. The problem with relying on last-click or first-click attribution is it discounts everything in between.

Think about the last time you purchased something online. You likely saw an ad, Googled the item, read reviews on a third-party site, watched a product video, shopped around to see if you could find a better price, and then finally bought the item.

Your path to purchase is rarely linear, and the same holds true for the buyers you want to convert. If you’re still using first-click or last-click attribution, consider switching to a more sophisticated model that factors in all engagements leading up to a conversion. 

2. Don’t be afraid to dive in to the data.

In my experience, the No. 1 reason marketers haven’t launched a new attribution model isn’t because they don’t believe it’s worthwhile: it’s because they’re afraid they’ll mess something up. In many cases, when we see data presented in a new way, we start to panic, but you have to start somewhere. 

Begin by analyzing one new client’s purchase path. Get granular by identifying every marketing and sales touch with the client and mapping out the full buying cycle. A single example can teach you a lot about your clients’ purchase path and your organization’s sales cycle.

3. Don’t change your methods too often.

When you launch a new attribution model, consistency is key. Pick a model, pick a time frame (i.e., 30, 60, or 90 days) and go for it. Do it for a couple of quarters, and then reassess. If you can’t decide which model makes the most sense for your organization, start with the even-weighted model where each touchpoint is assigned equal value. This way, you can factor in every event that influenced a customer to make a purchase. 

4. Don’t use marketing jargon to explain results to business leaders.

Even well-aligned marketing and sales teams often speak different languages when it comes to measuring success, which can lead to frustration when you’re attempting to present results to business leaders. While you might be excited to discuss your new marketing assets and the audience engagement metrics like video views or landing page conversion rates, you might be getting a little deeper in the weeds than your executives want to be.

So instead of sharing engagement metrics, focus on how your efforts are impacting the bottom line by shortening the sales cycle, increasing deal size, or expanding the number of contacts within an account. A more advanced attribution model can you help you draw these conclusions.