Results depend on the tool, the model and what you enter.
Result preview
Example: written by our editors, not real AI output.
Recs: new SKUs never shown
Default: score 0 if < 20 events
New vs mature: exposure 0.4 % vs 12 %
Rule: explore 5 % traffic for cold SKUs
Exit: 50 impressions or 14 days
Fill in
What you type stays in this browser window. We store and send nothing.
What scores?
recommendations
lead ranking
search boost
fraud score
mixed
Who is cold?
new users
new SKUs
new locales
new sellers
mixed
What looks wrong?
never shown
always last
sudden drop after launch
prevention
mixed
Your prompt
Help me spot cold-start bias.
System: [System]
New group: [New group]
Metric: [Metric]
Steps: 1. Read how scores are built for entities with little history. 2. Flag defaults that push new entities to the bottom or top without evidence. 3. Compare outcomes for new vs mature cohorts on the same metric. 4. Propose priors, exploration or separate cold-start rules. 5. Define when an entity leaves cold-start. 6. Write a short note for product and analytics.
Output: bias findings, cohort compare, cold-start rules and exit criteria.
Take particular care to give cold entities their own rule instead of punishing zero history.
Before you answer, check: Are all instructions met? If key information is missing or a placeholder in square brackets is still in the text, ask instead of guessing. Mark assumptions as assumptions.
You have edited the prompt. Changes in the fields above no longer overwrite it.
This is an instruction for an AI, not the result itself.
Prompt that explains a cohort table, compares cohorts at the same age and separates chance differences from real trends.
Prompt · Intermediate · about 8 min
You may use our prompts freely, including at work. Please do not republish them as your own collection. We give no warranty for AI results; check them yourself. Terms of use
