Misunderstood Metrics #1: GA Attribution
(or how someone came to your site today from that email you sent out 4 months ago)
GA does a great job of making web data accessible and intuitive but there are a few dimensions and metrics that tend to trip up users, including marketing sources.
Users visit your site from a variety of sources over time. Maybe they discover your site via search, sign up for your email newsletter, then return to your site after clicking on the emails. Distributing credit to your different traffic sources is a question of attribution. Most frequently, companies use last-click attribution, where the most recent traffic source gets credit for the visit. That’s also the model that makes the most intuitive sense – when you ask “what traffic sources drove today’s visits” you’re usually not asking “how did they come to the site 2 months ago”.
Google Analytics, however, uses a default attribution model that is a variant of last-click. If a user comes in directly (as explained above), GA will NOT credit that user as direct if they ever came to the site from some other traffic source in the past 6 months (assuming they didn’t delete their cookies). Essentially GA is trying to make the “none” bucket as small as possible.
In practice this means every traffic source except direct/none is overstated if you’re looking at it from a last-touch perspective. This is also why you’ll see old emails and outdated links in your reports.
Quick links to other misunderstood GA metrics: