How should we measure creative performance across channels on the same basis?
Our answer
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.
- Confidence
- Low
- Strongest proof
- 6 official rules
- 4 practitioners, 10 opinions or claims
- Sources
- 17 sources
- 20 statements
- MTG
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentPlatform returns don't line up on their own. Segwise, a creative analytics software maker, advises matching the window and conversion definition before comparing two platforms' ROAS. TikTok offers 1 to 28 day click windows and Google defaults to 30 days (each platform's help pages).
Each platform counts a view differently. Meta's ThruPlay counts 15 seconds or a full play, TikTok's 6-second view counts 6 seconds, and a skippable YouTube view counts 30 seconds (each platform's help pages).
A shared ID is what lets you add results up. AdManage.ai, an ad launch tool maker, argues that one creative ID kept in the file name, each platform's ad name and the UTM content tag is the only way to roll results up across platforms.
Our judgment
We built the concept-level method from Segwise's method of putting platform results on the same basis first, AdRiseLab's concept-by-concept comparison and the different definitions of a view in the platforms' help pages. Keep budget between channels on tests rather than creative scores, since Fospha says a mix model cannot reliably judge single ads and Measured says multi-touch credit breaks down for that split. A naming convention and a UTM content tag joined to store orders do the core of this without a paid tool. Paid creative analytics mainly saves tagging time.
Our judgment, built from Segwise, AdRiseLab, Meta, TikTok, Google, Fospha, Measured, AdManage.ai and MarketerTools: our reading of these sources, not a finding from any one of them.
What the sources say
20 statements from 17 sourcesMeta counts 2-second continuous plays, 3-second plays (or nearly the full length of shorter videos), and ThruPlays at completion or 15 seconds.
“The number of times your video was played to completion, or for at least 15 seconds.”
Google counts a skippable YouTube in-stream view at 30 seconds or the end of the video, whichever comes first, or a click; Shorts ads count at 10 seconds. Bumper and non-skippable ads count no views.
“A viewer watches 30 seconds or until the end of the video, whichever comes first”
Google's default conversion windows are 30 days for clicks, 3 days for engaged views and 1 day for views. It recommends at least 7 days for clicks, and changes apply only going forward.
“Note that windows of at least 7 days are recommended, because they provide a richer set of conversion data.”
Miranda Ivy judges creative by taking the top 25% of spend, split into video and static, and counting how many metrics, such as hook, hold and click-through rate, each ad does well on.
“And we see how many crossover hits each one got.”
To reconcile Meta and third-party conversion counts, Meta says to match attribution windows, time zones and conversion date settings, then check raw pixel fires for duplicates or tag setup problems.
“use the Meta Ads Data Advisor, a Chrome extension, to check for duplicate pixel fires, which can sometimes be caused by redirects.”
inBeat, which sells UGC and length testing, says to judge video length on CPA, since completion rate favors short videos, and to test 6 to 15 against 16 to 30 seconds in feed. It found no published study.
“Judge length on cost per acquisition, because completion rate favors the shortest video by construction”
Improvado suggests tagging creatives and reading results by tag only once a value has 16 or more creatives (5 to 15 with a note). It labels these thresholds its own starting points, not tested.
“Our starting gate: under 5 creatives in a value, no read; 5 to 15, read with a note; 16 or more, read.”
CTC says top TikTok Shop creative can be reused in Meta ads; one brand has 20% of its Meta budget going to its TikTok Shop content.
“20% of their meta budget right now is going towards our TikTok shop content.”
Ricardo Powell says that for complex, high-priced offers, top-of-funnel videos should be judged on engagement such as hook and hold rate, not last-click cost per acquisition.
Measured, which sells incrementality tests, says multi-touch attribution may track only 30 to 60% of touchpoints, an estimate with no method, and fails for cross-channel budgets. Start with one test on your top channel.
“Start with a single incrementality test on your highest-spend channel”
A Marketing Operators host suggests putting persona, angle and hook in ad naming conventions so performance can be grouped and compared.
“You can set all those things up in naming conventions like persona, angle, hook.”
HexClad uses Northbeam to compare channels on one basis regardless of each platform's attribution, runs geo holdout tests with Haus, and keeps a mix model.
“the backbone of our media buying decisions”
Segwise, which sells creative analytics, says ROAS from two platforms is not comparable until you align attribution windows and conversion definitions, or you will scale the platform that counts most generously.
“Pick one attribution standard and apply it consistently.”
MarketerTools says inconsistent campaign names break filtering and comparison across platforms; name the same concept in parallel on Google, Meta and TikTok so reports can match them by text.
“When campaign names are inconsistent, your ability to filter, segment, and compare data across platforms collapses.”
AdRiseLab, a creative tool vendor, says brief each concept once, edit it into Meta and TikTok versions, and compare results by concept, not by version. Creator, founder and demo videos transfer; trends do not.
“the question is whether the concept works, not whether one edit beats the other.”
Fraser Cottrell says hook rate and hold rate are soft metrics and should not be your main measure of ad performance.
“There's no point using a hook rate or a hold rate as your key performance metric.”
Fospha, a mix model vendor, says a pure mix model is not reliable at the ad or creative level because creatives move together. It uses the platforms' own signals for creative calls inside its cross-channel model.
“MMMs are not creative testing tools, and they should not be positioned as such.”
AdManage.ai, which sells an ad launch tool, says give every creative a stable ID carried in its file name, its ad name on Meta and TikTok, and the UTM content tag, since platform IDs change when ads are duplicated.
“Platform IDs change when you duplicate or rebuild. Your own creative IDs don't.”
TikTok counts a 2-second view at 2 seconds of play, excluding replays, and a 6-second view at 6 seconds, a full play of a shorter video, or an engagement in the first 6 seconds.
“Number of times your video starts playing for at least 2 seconds in an impression session. Replays are excluded.”
TikTok lets advertisers set click-through attribution at 1, 7, 14 or 28 days and view-through at off, 1 or 7 days, and the setting cannot be changed once the ad group is published.
“Attribution settings cannot be changed once the ad group has been published.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer