See which of your own short form posts are worth making again, with the numbers behind every call. A free browser tool for creators and marketers who post short form video. It looks back at your posts. It does not predict reach.
<aside> 📊
🧭 7 steps in the tool · 🧪 6 test sets · 📂 11 files · ⚡ 1 prompt to start
</aside>
<aside> 🚀
🚀 FREE STRATEGY CALL
Want to grow your product through viral reach?
We help teams across iGaming, mobile apps, AI products, and SaaS build creator-led organic distribution generating tens of millions of views at CPMs starting from $0.03.
<aside> 🚀
</aside>
On the call you'll receive:
<aside> ⚡
Got 5 minutes? Start here is right below. Paste your rows into Claude and get a first read.
</aside>
<aside> 📦
Got 30 minutes? Download the zip, then open 🧭 How to run this workflow.
</aside>
<aside> 🧭
Looking for one answer? Open 🧩 The seven steps in the tool.
</aside>
<aside> 📂
Get everything in one download. The tool, six test sets, their expected results, the test log, a blank template and the guide.
Before using this prompt with your post data you should omit any PII (personal identifiable information) and check with your compliance team and/or your AI policy. The tool itself runs in your browser and sends nothing anywhere.
Paste the prompt into Claude, then add your goal, your label and eight to 30 rows under the last line.
You are helping me review my own short form video posts.
This is a look back at what already happened. It is not a
forecast. Do not say any result is statistically significant.
Do not say a label caused a result.
Rules:
A blank cell is missing. Never treat it as zero.
Work only with the rows I paste. Do not invent values.
Never average percentages.
Do these steps in order.
1. CHECK. Confirm every row shares one platform, one organic or
paid status and one observation window. List any row that
differs and leave it out.
2. RATES. For each post, give the goal rate as a percentage to
two decimals, with its events and its denominator, for
example 140 / 8337 = 1.68%.
3. SET. Give the number of posts with a rate, the median rate,
and the lowest and highest rate.
4. GROUPS. Group the posts by the label I name. For each group
give: posts, median rate per post, and pooled rate (summed
events divided by summed denominators). Keep the median and
the pooled rate as two separate, labelled numbers.
5. SPARSE. Mark any group with fewer than 3 posts as sparse.
6. CONFOUNDING. Flag a group whose median duration is 0.6 times
the rest or less, or 1.67 times the rest or more.
7. WORTH TESTING AGAIN. Name at most one group with 3 or more
posts, a median above the set median and no flag from step 6.
Cite its post IDs. If none qualifies, say so.
8. NEXT TEST. Write one experiment card: hypothesis, the posts
it rests on, one intended change, and how to compare.
If fewer than 8 posts have a rate, stop after step 3 and say a
reliable pattern has not been established.
Write in British English. No contractions. No dashes.
Keep every line under 70 characters.
Use plain text, no tables and no markdown headings.
Paste your goal, your label and your rows below this line:
<aside> ✅
Worked? Your next step is 🧭 How to run this workflow.
</aside>
<aside> ⚠️
A look back, not a forecast. The tool describes posts that already ran. It cannot say why a post worked. A group marked worth testing again is a lead for your next batch. Test it before you trust it.
</aside>