Asking a person what they actually thought of the advertising has stayed almost exactly where it was: a survey fielded at the end, a report weeks later, a deck that lands when the budget is spent.
So the industry optimised on what it could measure rather than on what it wanted to know. Clicks, viewability, completion rates, last-touch attribution. Every one of them is a guess at human response, never the response itself. A completed view is not a changed mind. A click is not an intention to buy.
That was a tolerable trade when media plans were locked for a quarter. It is an expensive one now, because of what has been put in charge of the optimisation. The risk with AI-driven planning and buying is not that it fails. It is that it succeeds, brilliantly, at the wrong thing. Give an engine a click target and it will find you the cheapest clicks in the market, from people who were never going to buy. Give it completion rate and it will find the environments where ads are hardest to skip. It will hit the number faster and more ruthlessly than any human team. It does not know your proxy is a proxy.
The fix is not slower automation. It is a better target: independent, people-based brand impact, arriving quickly enough to steer the campaign that is still running. That is what in-flight brand lift does.
What is in-flight brand lift?
In-flight brand lift is the continuous measurement of brand lift KPIs while a campaign is live: ad recall, awareness lift, consideration lift, brand perception and message association, read throughout the flight rather than after it — the same core practice of measuring brand lift, just on a faster clock.
The science is unchanged. A brand lift study compares an ad recalling group with a matched control group and reports the gap between them. That gap is the incremental lift attributable to the campaign. In-flight measurement keeps ad recalling vs control intact; what changes is the cadence of collection and reporting.
It is worth separating this from the “real time” metrics the buying platforms already show you. Those are engagement signals — impressions, views, clicks, completion — reported by the same platform selling the media. Useful for delivery. They are not independent measures of brand impact, and they answer a different question.
In-flight brand lift vs traditional brand lift studies
A traditional brand lift study is a post-mortem. Field the survey at the end, weight the data, write the report, present it weeks after the flight closed. Genuinely useful for the next campaign. Useless for this one.
In-flight brand lift collects from launch and produces rolling reads across the flight. Teams see directional movement early, watch confidence firm up as sample builds, and refine while spend is still in market. Same underlying methodology, different reporting rhythm. The distinction matters because the usual objection to real-time measurement is that speed costs rigour. It does not. Nothing about the experiment design is relaxed; the data simply stops sitting in a queue.
Why real time brand lift matters for in-flight campaigns
The traditional model trains teams to learn lessons for next time rather than this time. Closing the feedback loop during the flight changes three things.
Optimising campaigns while they are still live
This is the headline use case. When you can see which creative, channel or audience is actually moving perception, you can act on it: rotate out the execution that is not landing, shift weight toward the placements producing lift, tighten targeting, cap frequency where returns flatten, pause what is not working.
Without in-flight data, those same decisions get made on engagement proxies. The cheapest impressions and the most persuasive impressions are frequently not the same impressions, and CTR will never tell you which is which.
Proving value earlier in the flight
Brand budgets are defended in front of finance teams and clients who want evidence, not a promise that the wrap deck will be encouraging. A rolling read gives you directional brand impact while there is still time and budget to build on it.
That matters most in the moment a campaign needs to justify continued or increased investment mid-flight. “We will know in six weeks” is not an argument. A live consideration lift, with a widening ad recalling vs control gap, is. That’s the kind of evidence a proper campaign roi analysis needs to make its case.
Learning faster across creatives, channels and audiences
Continuous measurement turns every live campaign into a test environment. Creative variants, channel mixes and audience segments can be compared against each other on brand lift rather than engagement — the comparison that actually reflects the objective.
The compounding effect is the real prize. Learnings arrive while the next brief is being written, not months after planning has closed.
How real time brand lift measurement actually works
Three mechanics do the work.
Continuous collection from ad recalling and control groups. Surveys are served throughout the flight, not in a burst at the end. The ad recalling group has genuinely been served the campaign; the control group has not, but is otherwise matched on demographics, market and media propensity. A control group that is systematically lighter on media will quietly inflate every result you report, so how exposure is identified and how the control is matched should always be explicit.
Rolling reads on core KPIs. Ad recall, awareness, consideration, brand perception and message association are recomputed as responses accumulate, producing a time series rather than a single average. That distinction has teeth: a campaign that hit its awareness target in week two and flatlined, and one that climbed steadily to the same endpoint, produce identical post-campaign numbers and demand opposite decisions.
Statistical thresholds that tighten over time. Early in the flight, confidence intervals are wide and movement is directional. As sample builds, differences become significant. The discipline is knowing which stage you are at — and pre-registering your KPIs and decision thresholds before launch, so you are testing a hypothesis rather than fishing in noise.
Because the method asks people rather than tracking them,it applies the same way across digital, social, video, CTV, audio, retail media and OOH — with ctv ad measurement as demanding a discipline as any other channel — which makes cross-channel measurement a genuine comparison rather than an apples-to-oranges exercise.
What marketers can actually do with in-flight brand lift insights
- Creative optimisation. Rotate toward the variant showing stronger lift on the KPI that matters — the best recall driver and the best consideration driver are often not the same asset.
- Channel and placement optimisation. Move budget toward the channels and placements delivering measurable brand impact, and pause the ones delivering impressions only.
- Audience optimisation. Identify the segments responding most strongly and lean into them while there is budget left to lean with.
- Frequency and pacing. Read lift against exposure frequency to find where additional impressions stop moving perception, and redeploy everything beyond that point.
- Message correction. If you are buying “premium” and the ad recalling group is lifting on “affordable”, that is a creative brief problem you can fix mid-flight.
- Stakeholder reporting. Give leadership and clients a live view of brand impact rather than a media delivery report.
One caution: brand perception moves on a slower clock than a bid. Change one substantial variable at a time, and resist treating a two-point wobble as a crisis.
Common misconceptions about real time brand lift
“Real time means less rigorous.” The ad recalling vs control design is unchanged. Only the reporting cadence is different.
“The platforms already report this.” They report engagement signals, and they are not independent of the media being measured. When platform-reported lift and an independent partner’s read diverge, that gap is the reason.
“It is just a post-campaign report delivered earlier.” No. It is a continuous read designed to be acted on, which is a different product with a different purpose.
“You need a huge budget to act on it.” Modest adjustments — rotating creative, pausing a weak placement — produce disproportionate gains precisely because they are informed by brand impact rather than engagement proxies.
Getting real time brand lift insights with Happydemics
Happydemics has always measured brand lift the same way: independent of the platforms being measured, people-based, ad recalling vs control, across every channel in the plan. In-flight measurement is an extension of that methodology, not a separate product line. The rigour is identical. The delivery is continuous.
The practical effect is that creatives, audiences and channels can be compared live, on brand impact, by a partner with no stake in which of them wins.
From post-mortem to live input
In-flight brand lift changes what measurement is for. It stops being a verdict delivered after the fact and becomes an input to the decisions being made today — both for optimisation, changing the campaign while it is running, and for proof, demonstrating value before the flight ends.
Every other part of the machine already runs at that speed. The only question is whether the target you hand it is a proxy or the real thing.
