Creative and channel questions
Judge creative and channels on sales you count yourself, through holdouts and your own order data. Use early ad signals only to diagnose ads.
10 answers · 2 new or changed since Nov 16 · Data as of Nov 22, 2026
The short version
Our judgment, drawn from the 10 answers below.
Judge brand campaigns and halo effects by sales added in a holdout, and keep reading it after the ads stop.
From Measuring brand campaigns' effect on sales, Halo effects between channels, Ad effects beyond the click window
On Meta, use hook rate and click-through only to diagnose ads, and decide on cost per purchase.
Compare creative across channels by concept, judged on orders counted outside the platforms with one attribution window.
Before you compare countries, put them on one currency and revenue net of VAT, country costs and fees.
The 10 answers
Grouped by the decision you are making. Jump to a group:
Judge creative
How to compare creative across channels and how fast winners wear out.
How should we measure creative performance across channels on the same basis?
Compare creative across channels by concept. Give every ad a shared concept ID, judge each concept on orders counted outside the platforms with one attribution window, and read view or hook rates only against that platform's norms.
Low confidence · 17 sources
Which early ad metrics (hook rate, hold rate, click-through) predict sales, and on which channels?
On Meta, early ad metrics barely track sales. Hook rate, hold rate and click-through ran flat or slightly against ROAS, so use them to diagnose ads and decide on cost per purchase. TikTok and YouTube lack clear data.
New Nov 22 Medium confidence · 11 sources
How fast does a winning ad wear out on each channel, and should refresh plans differ by channel?
Plan a different refresh rhythm for each channel. Add new TikTok and AppLovin creative weekly, expect a Meta prospecting winner to last about 3 to 6 weeks and review Google assets monthly, replacing only those that fade.
New Nov 22 Medium confidence · 17 sources
Measure effects over time
Brand campaigns, halo between channels and effects that land after the click window.
How should we measure the effect of brand campaigns on sales?
Judge brand campaigns by sales added in a holdout that keeps measuring after the flight. Allow about 5 to 6 weeks for YouTube, 6 to 8 weeks for Meta reach and 2 to 6 months for multi-channel brand work.
Medium confidence · 16 sources
How should we measure the long-term effect of ads beyond the click window?
Measure what lands after the click window in your own sales data. Keep reading geo holdouts after the ads stop, and follow new customers as cohorts for 12 months. Use a calibrated mix model only for effects lasting months.
Medium confidence · 14 sources
How should we measure halo effects between channels?
Measure halo with a holdout of one channel, read on every outlet it could move: site, Amazon, stores and branded search. Credit it to that channel. Between tests, chart branded search against spend as a directional signal.
Medium confidence · 14 sources
Account for outside factors
Countries, retail, stockouts and influencer gifting that shift what your numbers show.
How should we compare results across countries?
Put countries on one footing before you compare them. Use one currency at each sale's exchange rate, revenue net of VAT and country costs and fees. Then judge new-customer acquisition cost and margin against each country's own breakeven.
Medium confidence · 10 sources
How should we measure the effect of retail distribution on our own site's sales?
Compare site orders by ZIP code with a baseline from before launch, and read the retailer's store sales alongside. Judge retail on total margin, since it can lift or take site sales.
Medium confidence · 10 sources
How should we measure the effect of out-of-stocks on marketing results?
Log daily which advertised products were buyable. Estimate lost sales from clean in stock days, and recheck tests and models that overlap a stockout.
Medium confidence · 13 sources
How should we measure the effect of influencer gifting with no tracking codes?
Log every gift and post, and name gifted creators in your post-purchase survey. Chart branded search against a 4 to 6 week baseline as a signal only, and prove the effect with a test such as pausing sends some weeks.
Medium confidence · 11 sources
More in Measurement
69 answers in 10 subjects