If 80% of your influencer value is invisible, you're measuring the wrong thing
Charlie Oscar's 2026 influencer report opens with a claim that should stop any performance marketer mid-scroll: 80% of influencer value will never be captured by clicks or codes. If that's even close to right, most brands are grading the whole channel on a fifth of what it does.
It’s a bold number, and we’ll come back to how much to trust it. But the underlying problem is real, and most marketing teams already feel it. You know influencer is working. You just can’t prove it the way your dashboard wants you to. So it gets treated as a nice-to-have, funded last, and cut first.
Why last-click can’t see influence
People don’t move through a tidy funnel. Someone sees a creator they trust. They don’t click. They remember it. A few weeks later they see it reinforced, then search for the brand and buy. Last-click attribution hands all the credit to the final thing they touched (the search, the retargeting ad) and influence gets nothing.
The report frames the chain like this: attention compounds into familiarity, familiarity builds trust, trust creates memory, and memory drives the action you eventually do measure. The problem isn’t that influence doesn’t work. It’s that it works upstream of where you’re looking.
Not every touchpoint will deliver an immediate sale, but every meaningful interaction strengthens the ecosystem that makes growth possible. The challenge for brands isn't measuring the last click, it's understanding the value of the relationship that came before it.
The cost of measuring wrong
Getting this wrong isn’t neutral. It’s expensive in two directions.
First, you underfund a channel that’s actually pulling its weight, because it looks weak in a two-week attribution window. Second (and this one’s worse), you push creators to make last-click-friendly content. Hard codes, aggressive CTAs, “link in bio now.” You end up training the people who are brilliant at building trust to behave like a discount banner. The report describes exactly this drift toward sameness, and it’s partly a measurement problem in disguise.
What to measure instead
The word that matters is incrementality: the sales that genuinely wouldn’t have happened without the activity. Not the sales that would have come anyway and just happened to touch a code on the way out.
The second thing to measure is influence’s effect on everything else. The report cites influencer activity lifting branded search by up to 30% and paid media performance by 26%. Treat those specific figures as directional (they’re the vendor’s own), but the principle is sound and widely observed: influencer is often the thing that makes your other channels work harder, and last-click can’t see that by design.
The measurement toolkit, honestly
The report’s answer is Marketing Mix Modelling, which they sell as a tool called COmpass. MMM looks at all your activity together and estimates each channel’s real contribution while accounting for seasonality, pricing, and promotions. It’s genuinely useful, and it’s also what they’re selling, so read their framing as a case for the method rather than a neutral verdict.
You don’t need one vendor to start. The practical toolkit runs from cheap to serious: ask new customers how they heard about you (blunt, but revealing), run geo holdout tests where you switch influencer off in matched markets, use brand lift studies, and layer MMM on top when the budget justifies it. None of these is perfect on its own. Together they beat a last-click dashboard that’s confidently wrong.
What good looks like when you measure properly
The report shares a couple of client results that show what happens when the measurement matches the ambition. Take them as illustrative rather than guaranteed (again, the vendor is grading its own work), but they’re a useful picture of the upside.
For a supplements brand, HEIGHTS, connecting influencer activity to real commercial impact is credited with a 40% increase in influencer-led customer acquisition and a 52% improvement in paid social cost-per-acquisition. That second number is the interesting one: it’s influencer improving the efficiency of a different channel, which is precisely the effect a last-click view throws away.
For The Ritz-Carlton Yacht Collection, the report cites 160% year-on-year growth in brand demand, a 75% surge in category share, and a 29% uplift in bookings. Luxury travel with a two-week attribution window would have declared that campaign a failure long before the bookings landed. Measured over the right horizon, it’s the opposite.
How to actually apply this
Set the goal before you pick the metric. A product launch and a retention play are different jobs and shouldn’t share a scorecard. The report notes celebrities are 66% more effective at driving launches than retention, because their job is fame, not familiarity. Grading a launch creator on repeat-purchase is measuring the wrong thing on purpose.
Match the creator to the signal, then measure each against its own job. The report’s own breakdown: micro-creators over-index on conversion because credibility lives inside communities; macro creators behave more like amplifiers, extending salience; celebrities drive fame and launches. One blended ROAS number flattens all of that into noise.
And give it time. Compounding is the entire point. A fortnight-long attribution window will always make influence look weak, because you’re measuring a slow-building asset with a stopwatch.
Every number in this report comes from a company that sells influencer measurement and creator media. That doesn't make the figures wrong, but treat them as a case for the direction, not settled science. The safe takeaway isn't "the number is exactly 80%." It's "your current dashboard is probably missing most of the picture, and it's worth finding out how much."
Better measurement starts with cleaner records. The creators you work with increasingly run their deals, deliverables, and reporting through tools built for them, like Poppi Social. Cleaner data on their side means fewer gaps on yours.
See how Poppi works