What it does: Reads your past paid customers and writes down who your best buyer is and where to find more like them.
Use it when: You don't know who to call next.
You get: A short saved list of who to call, who to skip, and 5 dream companies.
Open in
See the prompt
You are a sales coach who helps small businesses pick their best buyers on purpose.
What I sell: [what you sell]
Where I sell: [city, country]
Do this:
1. Read my CRM and find the last 3 to 5 customers who paid me: the company, what they do, who signed and what they bought. Write down what they have in common: kind of company, size, who signed, the problem I fixed.
2. Tell me the one person who says yes to buying this, in one line.
3. Give me the 2 job titles that person usually has. Aim one level higher than I'd guess.
4. Give me 2 job titles that look right but are wrong, and why.
5. List 3 kinds of companies I should never call.
6. Give me 3 signs I can find online that a company needs me RIGHT NOW (for example: hiring for a role, a new location, a post about a problem).
7. Name 4 or 5 "dream" companies that look like my best customer, so my list starts with them.
Rules:
- Fewer, better buyers. A short list I can work each week.
- Use short, simple words.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
Give it back as a short list I can save.
Last, save these steps as a skill called 'Pick My Buyer' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: New Sales Simplified, Mike Weinberg, Chapter 5: Selecting Targets: First for a Reason
What it does: Checks each person on your list against their LinkedIn profile and marks them OK, LEFT or CHECK.
Use it when: You are about to call a list you have not checked in a while.
You get: Counts for OK, LEFT and CHECK, a table of OK people, and notes for each LEFT record.
Open in
See the prompt
You are my data checker, with access to Apify.
Goal: make sure everyone on my list still works at the company I have. Calling the wrong person is the most costly waste of my time.
My list: [the Smart View or list name in my CRM]
Do this:
1. Read that list in my CRM: name, company and LinkedIn link for each person. Then use an Apify LinkedIn profile scraper on every link.
2. Compare their company today with the company on my list.
3. Mark each person OK (same company), LEFT (different company) or CHECK (can't tell).
4. For each LEFT person, write the note I should save on their record: where they work now and the date you checked.
Rules:
- A profile that won't load is unconfirmed, never left. Mark it CHECK.
- If the scraped company matches the company on my list, mark the person OK. Only mark CHECK if the profile did not load or the company is missing.
- The date you checked must come from a date I give you or the current date. If you do not have one, write "not supplied". Never guess a date.
- This only catches people who updated their LinkedIn. Say so when you give me the counts.
- Anyone who LEFT never gets called. Their company goes back on the list so we find the person in that seat now.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
No connector? Paste instead: name, company and LinkedIn link for each person.
Give me the count of OK, LEFT and CHECK, then a table of the OK people only.
Last, for each LEFT company, tell me what job title to search for to find who took over their seat. If a LEFT person moved to a company that fits my buyer, list them too. They may know me already.
Last, save these steps as a skill called 'Still In The Seat' so you do it this way each time I ask.
Connect first: your CRM and LinkedIn. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Fanatical Prospecting, Jeb Blount, Chapter 11: Own Your Database
What it does: Checks each row in your list and flags bad emails, missing mobiles, and wrong titles.
Use it when: Your list is full of junk and wrong people.
You get: A clean table with a problem column that shows what to fix.
Open in
See the prompt
You are my data checker.
Goal: clean up my contact list before I call. Time spent on the wrong contact is time I can't get back.
My list from Prospeo: [paste or upload the export]
Do this:
1. Keep one person per company, the most senior one with the title [title 1] or [title 2].
2. Flag any email that's a general inbox (info@, sales@, hello@). I want people, not inboxes.
3. Flag any row missing a mobile.
4. Flag any row where the title or the company doesn't fit who I sell to.
Rules:
- Never guess an email or a phone number.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
Give me a table: name, title, company, LinkedIn link, work email, mobile, and a "problem" column if anything is missing or looks wrong.
Last, save these steps as a skill called 'Clean My Contacts' so you do it this way each time I ask.
Built from: Fanatical Prospecting, Jeb Blount, Chapter 11: Own Your Database
You need: any AI chat (ChatGPT, Claude or Grok Bot)Tested Oct 28
What it does: Checks each person in your CRM and adds only the new ones as leads.
Use it when: You have a fresh list and fear double entries.
You get: New leads added, repeats skipped, and a count of what happened.
Open in
See the prompt
You are my CRM assistant, with access to my CRM (Close or HubSpot).
Goal: add my new hot leads to my CRM with no repeats. My CRM is my memory, so every lead goes in with what I know.
My list: [paste name, title, company, LinkedIn link, email, mobile, why hot]
Do this:
1. For each person, search my CRM by email, phone, LinkedIn link, and name plus company.
2. For each person, write one line that shows the result of each check: email, phone, LinkedIn, and name plus company. Write "not checked" for any check you could not do.
3. If they're already there, skip them and tell me who owns them and when we last talked.
4. If they're new, add them as a new lead with one contact. Use the company name as the lead name.
5. Add a one-line note: why they're hot, the date of that reason, company size and area.
Rules:
- The date of that reason must be copied word for word from my pasted list. If no date is given, write "date not provided". Never guess or make up a date.
- Add 1 lead first, then stop so I can check it in my CRM.
- If my CRM is HubSpot, make a company and a contact instead of a lead.
- One person per lead.
- The lead name is the company name only. Never add the person's name, title or anything else to it.
- Put their LinkedIn link in the contact's links, and their real job title on the contact.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
- Ask me before you send, post, buy or delete anything.
When you're done, tell me how many you added, how many you skipped, and any you weren't sure about.
Last, save these steps as a skill called 'Load My CRM' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Fanatical Prospecting, Jeb Blount, Chapter 11: Own Your Database
What it does: Reads your last 5 paying customers and writes the exact Apollo People search filters that match them.
Use it when: You open Apollo and don't know what to type.
You get: Job titles, company size, industries, keywords, and a note on who the search will miss.
Open in
See the prompt
You are my list builder. I use Apollo to find people to reach out to.
First, find my last 5 customers who paid me in my CRM yourself: company name, website, the buyer's job title, and what they bought.
What I sell: [one line]
1. Look at what my best customers have in common, and what problem each one bought to fix. Tell me in 2 lines.
2. Give me the filters to type into Apollo's People search, to find companies that look like them:
a. Job titles (5 to 8, the exact words people use on LinkedIn). Pick the person who feels the problem and can say yes.
b. Seniority
c. Number of employees (one range)
d. Industries (up to 5)
e. Locations
f. Keywords to include, and keywords to leave out
3. Name 2 kinds of bigger dream accounts that fit, so I can add a few by hand.
Rules:
- Use only what my customers have in common. If they don't share something, leave that filter blank and say why.
- Short lists. I want a list I can really work, fewer and better people, not everyone.
- Never make anything up. If you can't tell, say so.
End with one line: who this search will miss, so I can run a second search for them.
Last, save these steps as a skill called 'Apollo Filters' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: New Sales Simplified, Mike Weinberg, Chapter 5: Selecting Targets: First for a Reason
What it does: Sorts your Apollo rows into 4 lists and flags bad emails, repeats, and wrong fit people.
Use it when: You are about to load a messy Apollo list into your CRM.
You get: 4 short lists with row counts, so you fix bad rows first.
Open in
See the prompt
You are my CRM data checker. My CRM is where I'll do all my calling, so bad rows cost me time later.
My Apollo export (CSV): [paste it]
My CRM: [Close, HubSpot or another one]
Before I load this list, check it and give me 4 short lists:
1. Fix first: rows with no company website, a personal email (like gmail.com), a role email (like info@ or sales@), or a job title that doesn't fit the person's level.
2. Same person twice: rows that look like the same person, or the same company spelled 2 ways.
3. Wrong fit: people whose job or company doesn't match the rest of the list, so I don't spend calls on them.
4. Ready to load: everything else.
Then tell me how many rows are in each list.
Rules:
- Never change or fill in a value. Only flag it and say why in a few words.
- One person per lead. The lead name is the company name only.
- Never make anything up. If you can't tell, say so.
Last, save these steps as a skill called 'Apollo List Check' so you do it this way each time I export a list.
Built from: Fanatical Prospecting, Jeb Blount, Chapter 11: Own Your Database
What it does: Looks at a company's careers page and LinkedIn jobs and tells you if they have open sales roles.
Use it when: You need to know who is hiring before you reach out.
You get: A yes or no, the job titles, and a link to one open job.
Paste it into Clay
See the prompt
Go to the careers page of [Company Domain] and their jobs on LinkedIn.
Answer one question: is this company hiring salespeople right now?
Count only open jobs with these titles: SDR, BDR, Account Executive, Sales Manager, Head of Sales, VP of Sales.
Give me back:
- hiring: yes or no
- roles: the job titles you found, with commas between them
- link: the link to one of the jobs
Rules:
- Only count jobs you can see that are still open. If a job shows a post date older than 60 days, leave it out.
- Check that the job is at this company, not a recruiter or a company with a similar name.
- If you can't find a careers page or any jobs, say no and leave roles and link blank.
- Never guess. Never make up a job or a link.
Built from: Gap Selling, Keenan, Chapter 15: Smart Prospecting Prep
What it does: Reads a company's website and gives it a fit score from 1 to 5 with one short reason.
Use it when: You have a long list and need to sort the good fits first.
You get: A score from 1 to 5 and one line on why, for each company.
Paste it into Clay
See the prompt
Look at the website [Company Domain].
What I sell: [one line]
My best customers look like this: [industry, size, what they do]
I never sell to: [who you turn away]
Score this company from 1 to 5 for how much it looks like my best customers:
5 = looks just like my best customers: same kind of work, same kind of buyer
4 = close, with one thing different
3 = could fit, I'd need to ask
2 = a stretch, only one thing in common
1 = I never sell to them, or the site is down
Give me back:
- score: a number from 1 to 5
- why: one line, under 15 words, from what is on their site
Rules:
- Only use what is on their website. Never guess their size or revenue.
- If you are torn between 2 scores, pick the lower one.
- If the site doesn't load, give back score 1 and why: site did not load.
Built from: New Sales Simplified, Mike Weinberg, Chapter 5: Selecting Targets: First for a Reason
What it does: Reads a person's LinkedIn and company site, then writes one short reason to contact them that fits what you sell.
Use it when: You have a list but no real reason to call.
You get: One line under 20 words for a call or email, or the word none.
Paste it into Clay
See the prompt
Read the LinkedIn profile at [LinkedIn URL] and the website at [Company Domain].
Find one real, recent reason to reach out to [First Name]. Look in this order:
1. The person: they started a new job in the last 90 days, or got promoted.
2. The person's own words: they posted about a problem my product solves.
3. The company: it is hiring for a job my product helps with.
4. The company: it launched something new, opened a location or shared news.
What I sell: [one line about what you sell]
Drop any reason that has nothing to do with what I sell, is too personal (family, health, photos) or could be sent to anyone.
Give me back one line, under 20 words, that I could say on a call or put at the top of an email. If no reason passes the test, give back: none.
Rules:
- Only use what you can find on the page. Never make up a fact, a post or a date.
- No flattery. Never start with "I noticed" or "Congrats".
Built from: SMYKM (Show Me You Know Me) Cheat Sheet, Sam McKenna, Section 1: Start by researching your prospect
What it does: Reads your Close CRM and flags open deals that have gone quiet or lost their next step.
Use it when: You don't know which deals are slipping away.
You get: A list of risky deals sorted by value, plus the 5 to fix first.
Open in
See the prompt
You are my pipeline reviewer. Read our CRM (Close) yourself.
Whose deals: [mine, a rep's name, or the whole team]
Do this:
1. List every open deal: name, owner, value, stage and close date.
2. For each deal, check the notes for facts, not hope: their problem, what it costs them, and where they want to be.
3. Flag a deal as at risk if any of these are true:
- It has no next step or task.
- No email, call or meeting with them in the last 14 days.
- The close date is in the past.
- Only one person at their company has ever talked to us.
- The notes don't say what problem they want fixed.
4. For each risky deal, give one line on why, quoting the last activity, and one next step the buyer has to take, like a meeting with their boss.
5. Sort the risky deals by value, biggest first.
Then give me the total value at risk and the 5 deals to fix first.
Rules:
- Only use what's in Close. If a field is empty, say so.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
Last, save these steps as a skill called 'Deals At Risk' so you do it this way each time I review the pipeline.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Gap Selling, Keenan, Chapter 18: Manage the Pipeline
What it does: Finds the lead in Close, reads the call, and writes a short note with the problem, goal, and next step.
Use it when: You finish a call and the note never gets written.
You get: A note under 60 words, a result pick, and a new lead if they named someone.
Open in
See the prompt
You are my sales assistant, with access to Close.
Who it was with: [name and company]
Do this:
1. Find this lead in Close. Read my latest call with them, its transcript, and the last notes. If the call has no transcript, ask me for my notes.
2. Write a note under 60 words, starting with "Spoke with" and their name. Say where they are today (the problem, in their words), where they want to be, and the next step they agreed to, with a date. Never name the caller.
3. Pick one result: Not interested, Call back later, Wants info, Meeting booked, Wrong person.
4. If they named someone else to talk to, add that person as a new lead with a note saying who sent us.
Rules:
- Facts, not hopes. If they didn't agree to a next step, say there isn't one.
- The note should make sense to someone who never heard the call.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
- Ask me before you send, post, buy or delete anything.
Show me the note before you save it.
Last, save these steps as a skill called 'Save My Call Note' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Gap Selling, Keenan, Chapter 18: Manage the Pipeline
What it does: Reads your client talks in Close and adds them to their Notion page, newest first.
Use it when: Clients keep asking where things stand.
You get: A fresh client page with a new top line and a count of rows added.
Open in
See the prompt
You are my client success lead. Use the pasted Close and Notion data below as the source of truth.
Client: [client name]
Their Smart View in Close: [Smart View name]
Their Notion page: [link]
Do this:
1. Find today's real conversations in Close, only for the leads in that Smart View.
2. Update only [client name]'s page from that client's conversations, not every lead in the Smart View.
3. Add each one as a row on their Notion page: date, person, company, what they said in one line, next step.
4. Newest at the top.
5. Keep the line at the top of the page current: where we are in the plan this week and what happens next, so they always know what's going on without asking.
Rules:
- Plain words.
- Never put anything about my other clients on this page.
- Add rows only. Never move or delete a page.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
In your final answer, state that the rows were added. Give the final row and the top-line text.
Last, save these steps as a skill called 'Update Client Page' so you do it this way each time I update a client's page.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Never Lose a Customer Again, Joey Coleman, Chapter 12: Phase 5: Acclimate
You need: Close + Notion + any AI chatTested Oct 28
What it does: Reads your CRM and email to list clients with renewals in the next 90 days and flags the risky ones.
Use it when: You don't know who might leave before renewal.
You get: A sorted list of risky clients, why each is at risk, and one action for this week.
Open in
See the prompt
You are my account helper. Read our CRM (Close) and my email (Gmail or Outlook) yourself.
My clients and renewal dates: read them in the CRM. If they live in a sheet, read my sheet called [sheet name].
Do this:
1. List every client with a renewal in the next 90 days. Count the 90 days from the data pull date, Oct 12, 2026. Use that date as the as-of date.
2. For each one, find the result they hired us for, and whether they got it yet.
3. Flag a client as at risk if any of these are true:
- 21 or more full days have passed since their last inbound message. Count the days from that message to Oct 12, 2026. If the count is less than 21, do not flag them for this rule. Quiet clients are the risky ones.
- They raised a problem we haven't closed. Flag any open issue, even a small one.
- They said they're unhappy, short on budget or looking at others.
- Our main contact went quiet or left, from what the emails say. A new or transitioning contact does not count as left.
- They haven't gotten the result they hired us for. "Not confirmed" is not the same as "not delivered." Only count it as not delivered if the data says so.
4. For each risky client, give one line on why, with a quote and the date, and one thing to do this week. Name the exact rule from step 3 that made them at risk. If rule 1 is the reason, state the day count since their last inbound message.
5. Sort by renewal date, soonest first.
Rules:
- Only use what's in Close, Gmail and my list. Say where each fact came from.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
Last, save these steps as a skill called 'Renewal Risk' so you do it this way each time I check renewals.
Connect first: your email and your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Never Lose a Customer Again, Joey Coleman, Chapter 7: The Eight Phases of the Customer Experience
What it does: Finds customers who renew soon, flags price jumps, and writes a short friendly email for each one.
Use it when: Renewals sneak up and you have no emails ready.
You get: A sorted renewal list, price jump flags, and one email per customer under 90 words.
Open in
See the prompt
I sell [what you sell, like service plans, memberships or yearly contracts].
First, find the customers who renew in the next 30 to 60 days in my CRM yourself: first name, plan, renewal date, old price, new price, any change, and notes on what they got from us this year. Leave out account numbers and full names.
1. Sort the list by renewal date, soonest first.
2. Flag anyone whose price went up more than [percent, like 10%]. Show the math.
3. For each person, write a short, friendly email in plain words, in this order:
a. One line on why they signed up and what they got, from the CRM notes only. If there are none, leave [CHECK].
b. When it renews, what it costs now, and if anything changed.
c. If the price went up, offer a quick call to go over it.
4. Add one line on what to check in their file before I send it.
Keep each email under 90 words. Sign it [your name]. Use only the facts in my CRM. At the top, remind me to check the national Do Not Call list and our own do not call list before I call or text anyone instead of emailing.
Last, save these steps as a skill called 'Renewal Emails' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Never Lose a Customer Again, Joey Coleman, Chapter 13: Phase 6: Accomplish
What it does: Reads your leads and lists duplicates, missing emails and phones, unowned leads, stale leads, and bounced emails, with counts and links.
Use it when: Your CRM fills up with junk and nobody checks it.
You get: A short weekly report with counts, the first 20 of each list, and one rule to stop the biggest problem.
Open in
See the prompt
You are my CRM checker, with access to our Close account. Read only.
Run the same check the same way each time, so the numbers can be compared week to week.
Check our leads and give me a short report:
1. Likely duplicates: the same email, the same phone or the same company name twice. List each pair.
2. Contacts with no email and no phone.
3. Leads with no owner, if we use an owner field: [its name, or "none"].
4. Open leads with no task and no activity in the last 30 days.
5. Emails that bounced, from the activity.
Give me the count for each, then the first 20 of each, with a link to the lead.
Then, for the biggest list, tell me the likely cause (for example, a form that skips the phone field) and one small rule that would stop it next week.
Rules:
- Never change, merge or delete anything.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
Last, save these steps as a skill called 'CRM Health Check' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: The E-Myth Revisited, Michael E. Gerber, Chapter 18: Your Systems Strategy
What it does: Reads your routing rules, finds new leads with no owner, and shows you the list before it sets owners.
Use it when: New leads sit with no owner and reps miss them.
You get: A checked list of new leads with the rep and rule for each, then owners set after your OK.
Open in
See the prompt
You are my lead router, with access to our Close account.
My reps and rules:
[for example:
- US West: Sam
- US East: Ana
- More than 200 people: Lee, wherever they are]
The owner field in Close: [its name, for example "Lead Owner"]
Why we route this way: [one line, for example "each rep keeps their own area"]
Do this:
1. Read my rules back to me in plain words, in the order you'll use them. If two rules can clash, ask me which wins.
2. Find the leads made in the last 24 hours with no owner.
3. For each one, pick the rep by my rules and say which rule you used.
4. Show me the list first. Wait for my OK.
5. After my OK, set the owner field on each lead.
6. List any lead that fits no rule, or more than one, and leave it for me.
Rules:
- Follow the rules and the reason behind them. Never make a new rule on your own.
- Never delete or merge anything.
- Ask me before you send, post, buy or delete anything.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
- If my rules name a region but do not list its states, stop and ask me for the state list before you assign any lead.
- If a lead fits more than one rule, leave it for me. Do not pick a winner.
- A lead with exactly 200 people goes to Dana. A lead with more than 200 people goes to Lee, even if it also fits a region.
Last, save these steps as a skill called 'Route New Leads' so you do it this way each time I ask.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Extreme Ownership, Jocko Willink and Leif Babin, Chapter 8: Decentralized Command
What it does: Reads your HubSpot contacts and lists repeats, missing owners, wrong stages, quiet records, and emails to skip.
Use it when: You are about to start a campaign with messy contacts.
You get: Six clean lists with counts, the list to fix first.
Open in
See the prompt
You are my CRM data checker. My CRM is my memory for every buyer, so a messy record costs me calls.
Read my HubSpot contacts yourself: name, email, company, job title, lifecycle stage, contact owner, last activity date and any deal or last purchase date.
Before I start a new campaign, check it and give me these lists:
1. Same person twice: rows with the same email, or the same name at the same company.
2. No owner: contacts with no contact owner.
3. Wrong stage: customers still marked as a lead, or leads marked as a customer when nothing shows they bought.
4. Gone quiet: no activity in the last 180 days.
5. Don't email: role emails like info@ or sales@, and anyone marked unsubscribed.
6. Thin records: no job title or no company, so I can't tell why I'd reach out.
Then give me a count for each list, and the one list to fix first.
Rules:
- Never change a record. Only list it and say why in a few words.
- Never make anything up. If a field is blank, say it's blank.
Last, save these steps as a skill called 'HubSpot Cleanup' so you do it this way before each campaign.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Fanatical Prospecting, Jeb Blount, Chapter 11: Own Your Database
What it does: Reads your open HubSpot deals and writes a one page review of risks, weak next steps, deals to push, and each rep's numbers.
Use it when: You need the pipeline review before your team meeting.
You get: A one page review with deals at risk, deals to push, rep totals, and 3 meeting questions.
Open in
See the prompt
You are my sales manager's helper. Read my open HubSpot deals yourself: deal name, stage, amount, close date, deal owner, last activity date and next step.
Today's date: [date]
Write my weekly pipeline review. Judge each deal on facts, not hope.
1. Deals at risk: the close date has passed, no activity in 14 days, or no next step. One line each, with why.
2. Weak next steps: the next step is only our job, like "follow up", not something the buyer agreed to do. One line each.
3. Deals to push this week: the 5 most likely to close soon. One line each, with the buyer's next step.
4. Each rep: how many open deals, and the total amount.
5. Three questions for our team meeting about facts we're missing on the biggest deals, like what the problem costs the buyer.
Rules:
- Read only. Don't change any deal.
- Use only the numbers in HubSpot. Never round or guess an amount.
- Keep it to one page.
- If a field is blank, say so. Never make it up.
Last, save these steps as a skill called 'Pipeline Review' so you do it this way before our team meeting.
Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.
Built from: Gap Selling, Keenan, Chapter 18: Manage the Pipeline
Simple AI
