Lead sources and cost per lead
Lead source is 'WhatsApp'. Which campaign? That column is blank.
Tested Claude Opus 5.5, medium effort · 15 Nov 2026 · on the practice files
Prepare channel working
Use lead ID, source and approved channel spend in same period. Do not include names or contacts.
Privacy Remove personal names, phone numbers, email IDs, photos and customer identifiers. Use approved business facts, anonymous codes or aggregate data.
Whenever I send the next brief, use this format. Period: [Period] Counting rule: [Counting rule] I'm attaching lead-source.xlsx and channel-spend.xlsx. A. Deduplicate for review: hold every duplicate occurrence, count valid unique leads by source and Unknown. B. Match supplied channel spend and calculate CPL per source where count positive; flag zero/missing. C. Facts and suggested tracking fixes, no invented attribution. Use only supplied facts. No invented statistics, awards, testimonials, customer names, No. 1, guaranteed, health, finance or legal claims. Do not copy competitor copy or use real people's names or photos. Write regional-language versions only if I request them in a box. A person checks before publishing or deciding.
Set it up once. Save the approved brand facts, tone and format, not customer details. Pick your tool at the top to see how.
What is that?
A Project is a folder in ChatGPT that keeps your instructions and files for every chat inside it. It's on the free plan too. Checked 14 Nov 2026.
What is that?
A Project in Claude is a workspace with its own instructions and files that every chat inside it uses. Checked 14 Nov 2026.
What is that?
A Gem is a custom Gemini that loads your instructions before every chat. Free on every plan; you create it on a computer. Checked 14 Nov 2026.
What is that?
Signed in with a work account, Microsoft's Copilot Chat applies enterprise data protection: your prompts aren't used to train its models. Checked 14 Nov 2026.
After that, each time: send the new brief and current approved facts; check the result.
First time? Try it in a normal chat: paste and send.
Which model? Use your tool's strongest model with thinking turned on.
In your AI tool: attach the files with the + button, paste, send.
Keep actual business records in your own workbook; ask AI for the method using column headings and rules only.
Two sheets: Leads A:B unique lead ID and source; Spend A:B source and spend. Counting rule [Counting rule]. Give COUNTIF duplicate flag for all occurrences, normalise blank to Unknown, count valid unique leads by source and look up supplied spend; CPL only with known positive count. Keep duplicates held and never invent attribution. Tell me exactly where to paste formulas or set up the local steps. Use fictional examples only; do not ask for actual company records.
Run locally, verify every result against source records, and have the responsible person or professional approve before acting.
- L2Duplicate. Both rows held.
- Meta2 valid leads. ₹1,000 spend / 2 = ₹500 CPL.
- L5Source unknown. Ask team; don't attribute.
Then send this: Recount unique IDs, source totals and CPL; keep duplicates held.
Deletes duplicate
Reply: "Hold both; verify."
Guesses source
Reply: "Unknown."
Invents cost
Reply: "Only supplied spend."
Sir ne bola urgent. Ab bolo: ho gaya.
Did it work?
If your whole team does this, we'll set it up for all of them in one session.