How an assisted conversion works
Most enquiries are not the result of one visit. Somebody meets you through a search, comes back a week later from a social post, then types your name into Google and fills in the form. Three touchpoints, one enquiry. Under the simplest reporting only the last one is credited, and the first two look like traffic that did nothing.
An assisted conversion is the report’s way of saying otherwise: this channel appeared somewhere on the path to a conversion, but was not the final click. Analytics tools record the sequence of visits tied to the same browser or user and then count, per channel, how often it appeared as a helper rather than a closer.
The count is bounded by two things. First, the lookback window: touchpoints older than the window are simply not seen. Second, whether the visit could be recognised at all — a click that arrived untagged, or from a browser that cleared its storage, cannot be joined to the conversion that follows.
Why assisted conversions matter
They stop you switching off the channels that create demand. Awareness work rarely closes anything. Social posts, video, display and general informational content are usually met early, remembered, and revisited through a branded search later. Judged on last click alone, all of it looks like waste and gets cut, after which the branded searches quietly dry up.
They are also the fastest way to spot a channel doing the opposite job — one that closes plenty but assists almost nothing. Branded search often behaves this way. It is valuable, but it mostly harvests demand that something else created.
Where assisted conversions go wrong
The first mistake is arithmetic. Assisted and last-click conversions describe the same conversions from different angles, so adding them together invents sales that never happened. Look at them side by side; never sum them.
The second is treating assist counts as proof of value. An assist records presence on the path, not influence over it. A visitor who was going to buy anyway may pass through several channels on the way, and each will claim an assist. Only a proper test — turning a channel off, or running a holdout — shows whether it changed the outcome.
The third is missing tags. If your email, social and partner links are not tagged with UTM parameters, those visits land in direct or referral, and the channels doing the assisting never appear in the report at all.
What to do about it
Before reading any assist report, make sure everything you control is tagged and the lookback window matches how long your customers actually take to decide. A remittance sign-up and a school enrolment do not deliberate on the same timescale.
Then compare models rather than trusting one. Look at the same period under last click and under a data-driven model, and pay attention to the channels whose value changes most — those are the ones your reporting has been mispricing. Treat the answer as a guide to where to test next, not as a settled number, and keep the measurement setup stable long enough for the comparison to mean something.