The Challenge
Digital video budgets keep growing, but the question marketers actually need answered hasn’t changed: are these channels reaching the right person, the right number of times, and moving brand metrics that matter? Platform-reported numbers don’t answer that. They answer a different, friendlier question — one asked and graded by the same party selling the inventory.
Coach wanted a deterministic, platform-independent read on a US cross-media campaign targeting women 18–34, across YouTube, TikTok, and Instagram. Specifically: true on-target reach by channel, incremental reach, frequency distribution, and whether the campaign was actually shifting brand perception — not just serving impressions.
The approach
Beatgrid measured 140 million impressions and 378 TRPs across the campaign window using its proprietary ACR (Automatic Content Recognition) technology. Audio fingerprints of each creative were matched against ambient audio captured by opted-in panelists through Beatgrid’s mobile meter apps, enabling person-level, cross-platform ad exposure verification — the same person, across every channel, deterministically.
A parallel Brand Lift Study ran against exposed and control groups (n=483 exposed, n=197 control), with a two-step weighting model that balanced the exposed sample against Beatgrid’s measured reach distribution, then twinned the control group to match.
What the data actually showed
Reach & Frequency
- Total campaign reach: 118M adults 18+ at an average frequency of 5.9
- Among Female 18–34: 24M reached at a frequency of 6.6
- YouTube delivered 53% of its impressions on-target — roughly double any other platform in the mix
- YouTube contributed 6M incremental reach against Female 18–34, with most overlap occurring against TikTok
- Against the broader 18+ population, YouTube drove 15M incremental reach
Brand Lift (Female 18–44)
Aided Awareness: +4.7 pts (90% confidence)
- Consideration: +8.7 pts (95% confidence)
- Purchase Intent: +1.6 pts (directional)
- Brand Favorability: +0.6 pts (flat)
- Unaided Awareness: −6.9 pts
The honest read: the campaign moved mid-funnel metrics meaningfully — exactly where a consideration-category purchase like a handbag gets won — but did not significantly shift unaided recall or near-term purchase intent in this window. That’s not a weakness of the campaign; it’s what rigorous, independently measured brand lift looks like when you’re not grading your own homework.
The frequency sweet spot
Exposure of 7+ times was where the lift curves bent upward on Aided Awareness and Consideration. Below that, movement was negligible. This is the kind of insight that’s operationally useful: it tells the media team where to set frequency caps and where additional spend starts earning its keep.
Creative-level diagnostics
Pastel and Flower creatives drove the strongest Consideration lift. Flower specifically over-indexed on Aided Awareness. Other creatives underperformed on specific KPIs. Because measurement was deterministic at the person level, the brand could attribute lift to specific creative executions rather than guess.
Why this matters
Three things the campaign team could do with this that platform-reported numbers don’t support:
- Re-weight the channel mix toward YouTube for on-target Female 18–34 reach, with TikTok and Instagram playing defined incremental roles rather than duplicative ones.
- Set frequency floors around the 7+ threshold for consideration-driving flights.
- Retire the creatives that underperformed and double down on Pastel and Flower in the next wave — with measurement, not instinct, making the call
The takeaway for marketers
Independent, deterministic cross-media measurement doesn’t flatter your media plan. It tells you what actually happened — including the parts that didn’t work — so the next plan is better than the last one. That’s the point.
Campaign measurement period: Feb–Jul 2024. Target audience: Female 18–34 (BLS reported against Female 18–44 due to channel-level sample sizes). Methodology: Beatgrid cross-media ACR panel + Brand Lift Study with two-step weighting.
