Which ads held people's attention, and how quickly each one stopped working.
ad by ad · attention and wear-out · SGMB.SGMB_All
A plain-English read of the charts on this page.
Waiting for the data to load.
Every rate on this page is one number divided by another. Here is each sum in full, with what a high or low result usually means.
Hook rate
The share of people who stayed long enough to watch the first quarter of the video.
Views to 25% ÷ impressions
High means the opening seconds stopped the scroll. Low means the start is weak.
Hold rate
Of the people who got a quarter of the way in, the share who watched to the end.
Views to 100% ÷ views to 25%
High means the middle of the video keeps people. Low means they drop off after the opening.
Completion rate
The share of people who started the video and watched all of it.
Views to 100% ÷ total video views
High means the full message landed, not just the first frame.
CTR (click-through rate)
Out of everyone who saw the ad, the share who clicked it.
Clicks ÷ impressions
Higher means the ad persuaded more of the people who saw it to act.
one point per ad · bubble = impressions · quadrants split at the median (0.0% hook / 0.0% hold)
Each dot is one ad. Going right means more people stayed past the opening seconds; going up means more of those people watched to the end. The top right corner is where the strongest ads sit.
Scroll to zoom, drag to move
Decision this unlocks — shift budget into the top-right quadrant and recut or retire the bottom-left before the next flight.
CTR and hook rate by days in flight · ads with 7+ days
How an ad's results change the longer it keeps running. Most ads decline as the same people see them repeatedly — the dip tells you when to refresh it.
Decision this unlocks — set a rotation rule — swap creative at the marked day rather than when performance has already collapsed.
Arabic against English against bilingual · CTR, hook, hold and CPM
How Arabic and English versions of the advertising compared, both in reach and in cost.
Decision this unlocks — weight the Arabic and English split by which one actually holds attention, not by an assumed 50/50.
0.30 CTR + 0.35 hook + 0.35 hold, each normalised across the set
The strongest and weakest ads once attention, cost and clicks are combined into one score. Treat it as a shortlist to look at, not a verdict.
| Ad | Platform | Spend | Impr. | CTR | Hook | Hold | Score |
|---|---|---|---|---|---|---|---|
| All 0 ads, best score first | |||||||
Decision this unlocks — brief the next round against what the top five have in common, and stop funding the bottom five entirely.
themes derived from the ad naming convention
Results grouped by what the ad was about, so you can see which subjects people respond to regardless of which ad carried them.
Decision this unlocks — fund the themes that convert attention, and give the weak ones one recut before dropping them.