Write a brief for a campaign to keep at-risk customers, based on why others left
A retention campaign brief for active at-risk customers, with eligible treatments and stop conditions.
Choose one of three approaches. The steps for that option appear below.
Choose an option above. On a phone, swipe the tabs left or right to see all options →
Gemini: builds the cohort table and drafts the brief
Gemini builds the cohort table from uploaded aggregate evidence and drafts the brief. You gate both stages and decide.
You
Collect aggregate evidence
Gemini
Build cohort table
You
Check the table
Gemini
Draft retention brief
You
Completion check, decide
Collect and clear the aggregate evidence
Use:You
Collect the supplied aggregate evidence; confirm consent and suppression rules are attached.
Check
Use supplied aggregate reasons and owner-defined eligibility only; do not infer individual churn or motive.
Build the active-at-risk cohort table
Use:Gemini
Upload the cleared inputs and ask Gemini for the cohort table: current status, addressable churn driver, allowed treatment, ineligible states, stop conditions.
Details
What to paste into the tool
Task: Active-at-risk cohort table with current status, addressable churn driver, allowed retention treatment, ineligible states and stop conditions; Source references; Uncertain or absent facts. Using only the supplied evidence set, build an active-at-risk cohort table: for each cohort, the current status, the addressable churn driver from the supplied aggregate reasons, the allowed retention treatment from the approved offers, ineligible states, and stop conditions. Keep the source ID and date on every row. Mark lapsed or former customers as out of scope for this brief. List missing or uncertain facts. Do not add facts beyond the uploads and do not infer individual motives.
Check the table against the inputs
Use:You
Compare the working record with the original inputs; resolve or keep missing facts before drafting.
Check
Permitted use is recorded on every row before drafting.
Draft the retention brief
Use:Gemini
In the same chat, match checked churn drivers to interventions for currently active at-risk customers, applying eligibility and stop rules.
Details
What to paste into the tool
Task: Draft retention campaign brief for currently active at-risk customers with eligible treatments and stop conditions; Open questions and approval dependencies retained beside the recommendations. Using only the reviewer-checked cohort table, draft a retention campaign brief for currently active at-risk customers: for each cohort, the eligible treatments from the approved offers and the stop conditions. Keep the source ID and date beside each intervention. State the suppression and stop rules explicitly. List open questions and approval dependencies beside the recommendations. Exclude lapsed or former customers; note them for a separate win-back brief. Do not add facts and do not infer individual motives.
Run the completion check and decide
Use:You
Every intervention must map to an evidenced driver and a currently active at-risk cohort; lapsed customers go to a win-back brief.
Check
Suppression and stop rules must be explicit. Approval and execution are separate decisions.
Claude: builds the cohort table and drafts the brief
Claude builds the cohort table and drafts the brief as a Word file. You check both stages and decide.
You
Collect aggregate evidence
Claude
Build cohort table
You
Check the table
Claude
Draft retention brief
You
Completion check, decide
Collect and clear the aggregate evidence
Use:You
Collect the supplied aggregate evidence; confirm consent and suppression rules are attached.
Check
Use supplied aggregate reasons and owner-defined eligibility only; do not infer individual churn or motive.
Build the active-at-risk cohort table
Use:Claude
Upload the cleared aggregate evidence and ask Claude to build the cohort table, marking unsupported items.
Details
What to paste into the tool
Using only the uploaded aggregate evidence, build an active-at-risk cohort table: for each cohort, the current status, the addressable churn driver, the allowed retention treatment from the approved offers, the ineligible states and the stop conditions. Keep a source ID and date on every row. Mark every item the evidence does not support as unsupported and list uncertain or absent facts separately. Use only the supplied aggregate reasons and the owner-defined active/at-risk eligibility; do not infer individual churn or sensitive motive.
Check the table against the inputs
Use:You
Check the draft cohort table against the original inputs; confirm a source ID and date on every row.
Check
Permitted use is recorded on every row before drafting.
Draft the retention brief as a Word file
Use:Claude
Ask Claude to match checked churn drivers to interventions for currently active at-risk customers, apply eligibility and stop rules, and create a Word document.
Details
What to paste into the tool
Task: Draft retention campaign brief for currently active at-risk customers with eligible treatments and stop conditions; Open questions and approval dependencies retained beside the recommendations. Using only the reviewer-checked cohort table, draft a retention campaign brief for currently active at-risk customers: for each cohort, the eligible treatments from the approved offers and the stop conditions. Keep the source ID and date beside each intervention. State the suppression and stop rules explicitly. List open questions and approval dependencies beside the recommendations. Exclude lapsed or former customers; note them for a separate win-back brief. Do not add facts and do not infer individual motives. Then create a Word document containing the brief.
Run the completion check and decide
Use:You
Every intervention must map to an evidenced driver and a currently active at-risk cohort; lapsed customers go to a win-back brief.
Check
Suppression and stop rules must be explicit. Approval and execution are separate decisions.
ChatGPT: builds the table and drafts the brief
ChatGPT builds the cohort table from the aggregate-only inputs and drafts the brief. You gate both stages and decide.
You
Collect aggregate evidence
You
Remove individual-level records
ChatGPT
Build cohort table
You
Check the table
ChatGPT
Draft retention brief
You
Completion check, decide
Collect and clear the aggregate evidence
Use:You
Collect the supplied aggregate evidence; confirm consent and suppression rules are attached.
Check
Use supplied aggregate reasons and owner-defined eligibility only; do not infer individual churn or motive.
Remove individual-level records
Use:You
Remove individual-level records and identifiers so only aggregate churn reasons, owner-defined active and at-risk criteria, consent and suppression rules and approved offers remain.
Check
Check no row, name, contact detail or individual churn reason remains, and that consent and suppression rules are still attached to the aggregate set.
Build the active-at-risk cohort table
Use:ChatGPT
Upload the aggregate-only inputs and ask ChatGPT for the cohort table: status, addressable driver, allowed treatment, ineligible states, stop conditions.
Details
What to paste into the tool
Task: Active-at-risk cohort table with current status, addressable churn driver, allowed retention treatment, ineligible states and stop conditions; Source references; Uncertain or absent facts. Using only the supplied evidence set, build an active-at-risk cohort table: for each cohort, the current status, the addressable churn driver from the supplied aggregate reasons, the allowed retention treatment from the approved offers, ineligible states, and stop conditions. Keep the source ID and date on every row. Mark lapsed or former customers as out of scope for this brief. List missing or uncertain facts. Do not add facts beyond the uploads and do not infer individual motives.
Check the table against the inputs
Use:You
Compare the working record with the original inputs; resolve or keep missing facts before drafting.
Check
Permitted use is recorded on every row before drafting.
Draft the retention brief
Use:ChatGPT
In the same conversation, match checked churn drivers to interventions for currently active at-risk customers, applying eligibility and stop rules.
Details
What to paste into the tool
Task: Draft retention campaign brief for currently active at-risk customers with eligible treatments and stop conditions; Open questions and approval dependencies retained beside the recommendations. Using only the reviewer-checked cohort table, draft a retention campaign brief for currently active at-risk customers: for each cohort, the eligible treatments from the approved offers and the stop conditions. Keep the source ID and date beside each intervention. State the suppression and stop rules explicitly. List open questions and approval dependencies beside the recommendations. Exclude lapsed or former customers; note them for a separate win-back brief. Do not add facts and do not infer individual motives.
Run the completion check and decide
Use:You
Every intervention must map to an evidenced driver and a currently active at-risk cohort; lapsed customers go to a win-back brief.
Check
Suppression and stop rules must be explicit. Approval and execution are separate decisions.
Compare the three approaches
Option A: Gemini: builds the cohort table and drafts the brief
Best when: You want one assistant to carry both stages
What it does: Gemini builds the table and drafts; you gate
Strong at: Chats by text, voice, photo or camera
Limit: Eligibility and suppression rules bind every line; check them
Free plan: Yes
Option B: Claude: builds the cohort table and drafts the brief
Best when: You want Claude to draft the brief as a Word file
What it does: Claude builds the table and drafts; you gate
Strong at: Drafts and edits text in many formats
Limit: Eligibility and suppression rules bind every line; check them
Free plan: Yes
Option C: ChatGPT: builds the table and drafts the brief
Best when: You want ChatGPT to carry both stages on aggregate data
What it does: ChatGPT builds the table and drafts; you gate
Strong at: Answers questions; writes, rewrites and summarises text
Limit: Eligibility and suppression rules bind every line; check them
Free plan: Yes, limited (free plan caps uploads and analysis)
Before you paste into any AI tool, remove any confidential information. Put it back at the end.
The Lifecycle marketing lead and privacy reviewer compare the working record with the original inputs and confirm permitted use before drafting.
The Lifecycle marketing lead and privacy reviewer apply the check: Every intervention maps to an evidenced churn driver and a currently active at-risk cohort; lapsed or former customers are excluded or routed to a separate win-back brief, and suppression and stop rules are explicit. Human confirmation by Lifecycle marketing lead and privacy reviewer: evidence and unresolved questions reviewed.
No option sends, publishes, deploys or changes any account; approval and any execution are separate human decisions.
Instructions hidden in external files or websites can trick Claude; check files before you use them.
Check the result against your original inputs. Approve it before sharing or acting.
What still needs checking
Whether aggregate churn evidence is detailed enough to support per-cohort treatments is a data question no tool answer settles; facts still to confirm stay flagged for the reviewers.
The unresolved questions reach the draft through the explicit handoff from the checked record, and the constraints come from the owner-supplied constraints named in the starting input.
Vendor pages checked 2026-10-28. Check plan terms and limits with the vendor, and check each tool on your own data.
The options were chosen from the 154-tool catalogue by documented upload, extraction and drafting operations; purpose-built catalogue tools such as a source-grounded notebook or an SEO content tool were not compared against these options.