YouTube view-through conversions under finance scrutiny
A streaming fitness app runs a YouTube brand campaign plus Search and Demand Gen. YouTube reports 1.8 million impressions, 0.08% CTR, and 4,200 view-through conversions (VTCs) in Google Ads. GA4 shows YouTube assisted conversions, but last-click gives YouTube little credit. The app has a 14-day free trial and 60% trial-to-paid conversion. Finance questions whether VTCs are real and wants to pause YouTube. The team has a brand lift study showing 6-point ad recall lift and 3-point consideration lift, but no holdout. YouTube spend is $60k/month, and the app's CAC target is $45. Search converts at $38 CAC and Demand Gen at $52. The team needs to decide within four weeks whether to renew the YouTube budget for the next quarter. Brand search volume is not tracked.
1. How should the team interpret VTCs and assisted conversions, and why are they not proof of incrementality?
Read the worked answer
VTCs count users who saw a YouTube ad and later converted without clicking, within the view-through window. They are not incremental by definition because some of those users would have converted anyway. Engaged-view conversions require a meaningful view and are stronger, but still not causal. Assisted conversions in GA4 show that YouTube appeared in the path, but last-click under-credits the channel; neither metric proves YouTube caused the conversion. With 1.8 million impressions and 4,200 VTCs, the implied VTC rate is about 0.23%, which is plausible but not evidence of lift. The brand lift study showing 6-point ad recall and 3-point consideration is useful but measures upper-funnel effects, not paid conversions. Finance is right to question VTCs; marketing is right that last-click is biased. The answer is an incrementality test, not a debate about definitions.
2. Recommend a measurement approach to decide whether to keep YouTube, including what metric should drive the decision.
Read the worked answer
Run a geo holdout or YouTube conversion lift study. Split comparable DMAs into test (YouTube on) and control (YouTube off or at much lower budget) for 4-6 weeks. Measure incremental free trials and paid conversions using app analytics or GA4, not platform VTCs. Use difference-in-differences and report a confidence interval. If a holdout is impossible before renewal, run a smaller geo test or use a matched-market analysis and treat the result as directional. The decision metric should be cost per incremental trial and cost per incremental paid subscriber, compared with the $45 CAC target and the 60% trial-to-paid rate. Keep YouTube only if incremental CAC is acceptable; otherwise reduce budget or change creative. Do not optimize bidding to VTCs alone.
3. What should the upper-funnel dashboard show so that finance and marketing can agree on performance?
Read the worked answer
Separate the dashboard into three layers. Upper funnel: impressions, unique reach, frequency, view rate, engaged views, ad recall lift, and consideration lift. Mid funnel: assisted conversions, path length, new users, and brand search volume once tracking is added. Lower funnel: incremental trials, cost per incremental trial, trial-to-paid rate, and incremental CAC. Add guardrails for direct traffic and branded search. Avoid combining VTCs with click conversions in one ROAS. Show lift-study confidence intervals and label platform-reported metrics as modeled or attributed. This lets finance see incremental outcomes while marketing can still manage creative and reach.