Run log

Every test run, including the weak ones.

Each row is one idea taken from the Make button to a finished, AI-generated video, on our made-up test product. We add every run of this kind and don't take any away. None of these has run as an ad, and one made-up product can't tell you what yours will get.

Time to a finished video
7:56
Median. Range 6:23 to 11:37
Spent per video
$3.58
Median. Range $3.16 to $4.55
Things flagged per video
6
Median. Range 3 to 8
DateIdeaTimeSpentChecksFixedFlaggedScores hook · story · picture · sound · captionsVideo
Nov 2One Square Already Done7:23$4.55244535 · 50 · 45 · 90 · 90 14 s
Nov 2One Square Already Done6:23$4.10228462 · 65 · 68 · 90 · 88 15 s
Nov 1One Square Already Done7:22$3.34215640 · 45 · 60 · 80 · 85 14 s
Nov 1The Ring Around Drain10:48$3.72236370 · 45 · 50 · 80 · 90 14 s
Nov 1Three Heads, Three Jobs7:56$3.58223740 · 20 · 55 · 85 · 85 12 s
Nov 1One Square Already Done7:46$3.16225768 · 65 · 72 · 85 · 90 11 s
Nov 1One Square Already Done10:05$3.42242865 · 75 · 60 · 90 · 65 11 s
Nov 1One Square Already Done10:59$3.42193368 · 78 · 82 · 85 · 72 11 s
Nov 1One Square Already Done11:37$3.7251635 · 45 · 55 · 75 · 20 11 s
Oct 31One Square Already Done4:13$2.36Before the check existed 14 s
Time
From the first step to the last, leaving out idle gaps of over two minutes. Since November 2 the plan is written three ways and scored first, which adds about two minutes before the first step.
Spent
What the steps cost at canvas prices, with checks, redos and any later re-run on that canvas.
Checks, fixed, flagged
How many times a model or a measurement looked at a file, what got fixed on the way, and what's left for you to judge.
Scores
One model's opinion of the finished video, out of 100, for hook, story, picture, sound and captions. It isn't an ad result. The pipeline changed between these runs, so the scores aren't a trend.
Not listed
The 26-second ad and the reel on the home page, which were built by hand on the canvas, and two early experiments that never reached a video.