We’re misattributing half of our web traffic, and for some reason, we’ve decided to be fine with it.
Nani Shaffer
The next wave of marketing measurement tools and attribution solutions have to address the challenges with misattributed direct traffic.
Direct Traffic Woes
I’ve explored a number of different functions in B2B marketing, but my heart has always been in marketing operations — which means I’m no stranger to a good bar chart. I love translating data points and benign, colored rectangles into meaningful insights that ultimately drive decision-making at a company. And of course, our websites provide one of our richest sources of data. We spend inordinate amounts of money creating valuable content, interactive experiences, and compelling reasons for prospects to continue down the path toward becoming a customer.
Inevitably, the question arises: how on earth do you get more — and, critically, the right — accounts to your site? Well, you might say, let’s turn to the data! That response warms my heart, it really does.
So, what’s the next step? Let’s pull up a breakdown of your site traffic to see which channels drive the most traffic. Some helpful, affirming channels show up (otherwise known as channels you feel like you can control) like paid search and email. But inevitably, at the very top of the list and likely accounting for more than half of your traffic, sits this one: direct.
Let’s pause for a moment to talk about direct traffic. Direct traffic is defined as any visit to your site without a known source or referrer. If someone types your domain into a browser directly, for example, that visit will be categorized as direct traffic.
In reality, direct traffic functions more as a catch-all bucket, because it really means there wasn’t enough information associated with the visit to link it to a particular channel. As a result, all sorts of visits are swept into direct — even those that aren’t direct at all and should be appropriately linked to paid channels. For example, if someone views an ad and then types in your URL into the search bar, it’s categorized as direct even though the ad (a paid channel) really drove that visit.
So what on earth does one do about direct traffic? How do you adjust your resources to impact it? The answer is something between “basically nothing” and “squint really hard and see if you can make it look like your campaign correlated with an increase in direct traffic”. You might invest your time in practicing hand-waving so that you can breeze past the part of the presentation where you acknowledge that the largest portion of your traffic — and therefore the largest influencer on your pipeline — is actually a black hole. Scary thought, isn’t it?
In fairness, it’s not completely hopeless. There have been efforts to get at least a rough idea of the portion of direct that is driven by other channels — the concept of “lift” is used to look at the incremental difference in traffic from targeted accounts before and after a campaign launches. Lift makes it clear that click-through data usually underestimates ad traffic by a factor of 3-4x. And while measuring lift is certainly better than nothing, it’s still not concrete.
As marketers, we are expected to tie our investments to business results. But if our largest channel is inaccurate, so are the rest of our channels – because they’re missing the chunks that got incorrectly hoovered up into direct – our ability to measure channel efficiency is ultimately compromised.
The next wave of marketing measurement tools and attribution solutions have to address this huge gap. Our budgets are scrutinized more than ever, so we can’t afford to stay in the dark much longer.