B2B SaaS / Customer Research Beginner

31 ChatGPT prompts for B2B startup founders to uncover hidden ICP pain points

If you’re a B2B founder still writing your ICP from a slide deck, you’re guessing. You’re picking a firmographic shape - “VP of Sales at a 50–200 person SaaS company in North America” - and hoping the pain lives inside it. In 2026, the buyers inside that shape have changed faster than the shape itself. Forrester’s June 2026 research shows that 94% of B2B buyers now use answer engines like ChatGPT, Claude, and Perplexity during their search - and many of them lock in a vendor shortlist before a single human sales conversation (Forrester). If you can’t describe the buyer’s pain better than an LLM can scrape it from a G2 review, you’ve already lost.

These ChatGPT prompts for B2B ICP pain points are the 31 I keep coming back to. They cover ICP persona reconstruction, review and community mining, call and transcript mining, jobs-to-be-done (JTBD) outcomes, champion mapping, and trigger events. Each one is multi-line, ready to paste, and built to be re-run every quarter as your ICP shifts. I’ve grouped them by what you’re trying to learn, included example output, and added pro tips I use with early-stage B2B teams.

Pull quote: “Nearly all B2B buyers (94%) use answer engines during their search. Many of these buyers establish preference early in the process. If a solution isn’t showing up in that early research stage, it might never show up at all.” - Amy Bills & Karen Tran, Forrester, June 8, 2026.

Quick answer: what you’ll get from this prompt library

You’ll get a stackable system - 31 prompts, in 6 categories - that turns raw signal (reviews, call transcripts, support tickets, Reddit threads, churn notes) into a sharper ICP document. Run them on ChatGPT-4o, Claude 3.5 Sonnet, or your internal LLM. Treat every output as a hypothesis, not a fact. The win comes when three prompts point at the same pain from three different angles.

Here’s the 10-second mental model:

Persona reconstruction (1–5): rebuild the buyer, not the title.
Review & community mining (6–10): mine G2, TrustRadius, Reddit, PeerSpot, niche Slack groups.
Call & transcript mining (11–15): pull JTBD from Gong, Chorus, Grain, Fathom, Tactiq, Zoom.
JTBD & outcome prompts (16–20): translate features into jobs the buyer is hiring you to do.
Champion mapping & org-chart prompts (21–25): find the humans behind the logo.
Trigger-event & moment prompts (26–31): catch the buyer mid-trigger, not after the vendor’s already chosen.

The fastest path: pick the section that maps to your weakest ICP hypothesis, run all 5 prompts in that section on the same raw input (a week’s worth of customer calls, 50 G2 reviews, etc.), then meet for 45 minutes to synthesize.

Why most B2B ICPs are wrong in 2026

Most ICPs fail for the same reason: they describe who the buyer is, not who the buyer is becoming. The static persona - industry, headcount, ARR band, geography - was useful in 2018. In 2026, that shape no longer predicts pain.

Three forces have reshaped the buying committee, and your ICP needs to reflect all three.

1. The buying committee got bigger and more distributed. What used to be a VP and a champion is now 6–10 people across security, IT, legal, finance, RevOps, and end users. Forrester’s B2B buying network research (June 2026) puts the typical B2B decision-making unit at 6–10 named participants, with up to 3 veto holders. Your ICP can’t be one persona. It has to be a buying center map.

2. AI displaced the top of the funnel. Forrester (June 2026) reports 94% of B2B buyers use answer engines during search, and 73% of customers say companies treat them like individuals (up from 39% in 2023, per Salesforce’s State of the AI Connected Customer). The pain statements buyers are typing into ChatGPT are the new ICP. If your ICP doc doesn’t reflect those phrasings, you won’t be picked.

3. AI disrupted marketing’s day-to-day. HubSpot’s 2026 State of Marketing report found 61% of marketers say marketing is in its biggest disruption in 20 years due to AI, and 80% of marketers already use AI for content creation. Translation: your buyers are drowning in mediocre AI-generated content from your competitors. The only way to cut through is to know their pain so granularly that your messaging sounds like an internal monologue, not a brochure.

“A clear takeaway of the event was that building fast doesn’t matter if you’re not building the right thing. As AI makes creation easier, decisions become the differentiator.” - UserTesting Crafted 2026 event recap, Forrester, June 8, 2026.

The hidden pain - the one your buyers won’t say out loud in a sales call but will whisper to a peer in a Slack group, or type into a G2 review, or rant about in a Reddit thread - is the only pain worth building for. Surface-level pain gets you a comparison-shopped deal. Hidden pain gets you a champion.

The hidden-pain mining framework (HPM)

Before you paste a single prompt, you need a structure. The Hidden-Pain Mining framework has three layers, and every prompt below maps to at least one of them.

Layer 1 - Observed pain. What buyers say out loud in calls, reviews, support tickets, community posts. Easy to find, but it’s where everyone looks. Competitive noise lives here.

Layer 2 - Acknowledged pain. What buyers admit when pressed, but only after trust. This is where win/loss interviews, MEDDIC deep dives, and JTBD outcome interviews live. The Mom Test by Rob Fitzpatrick - used at Harvard, UCL, Seedcamp, and Shopify (momtestbook.com) - teaches you to surface this without leading the witness.

Layer 3 - Latent pain. The pain the buyer hasn’t articulated yet because they don’t have the vocabulary. Latent pain shows up in workarounds, in tools that shouldn’t coexist, in hand-rolled scripts, in the words they don’t say. This is the gold layer - and it’s where ChatGPT shines, because you can paste in thousands of words of messy signal and ask for the pattern nobody noticed.

Every prompt below is tagged with the layer it targets. If you only run prompts from one layer, you’ll get an ICP that’s correct and useless. Run at least two prompts per layer, per quarter.

Now, the prompts.

Section 1: ICP & persona reconstruction (prompts 1–5)

These five prompts help you rebuild the buyer from the inside out. Start here if your ICP feels like a job description, not a person.

Prompt 1 - Persona ghost-protocol (Layer 1 + 2)

Purpose/context: Reconstruct a real buyer persona from raw call notes, support tickets, or sales emails. Forces ChatGPT to behave like an investigative journalist, not a copywriter.

You are a customer research analyst. I will paste raw notes from sales calls,
support tickets, and onboarding sessions with buyers in [INDUSTRY /
VERTICAL]. Your job is to reconstruct the persona who is actually sitting
across from our sales team - not the persona on our marketing site.

Do the following:
1. Extract 5 direct quotes that signal frustration, anxiety, or relief.
2. Identify the buyer's primary job-to-be-done (in their own words).
3. List 3 tools they currently use that they openly complain about.
4. List 3 tools they use but never mention by name (workarounds).
5. List 3 emotional triggers (e.g., "I'm being evaluated on this").
6. Write a 4-sentence "day in the life" narrative from the buyer's POV,
   written in the first person, using their actual phrasing where possible.
7. Flag any pattern that contradicts the persona currently on our website.

Output as a structured report. Cite quote-to-source where I can.

Example output (abridged): Direct quote: “I’m spending more time defending my stack than using it.” JTBD: “Prove ROI on the CRM before QBR without pulling another all-nighter.” Workarounds: spreadsheet the team calls “the duct tape tab.” Emotional trigger: end-of-quarter board prep.

Pro tips:

Paste at least 3 different sources per persona (call, ticket, email). The contradictions are the signal.
Run this prompt per persona, not per segment. Don’t blend.
Save the output verbatim. Use it as the persona doc, not a starting point.

Prompt 2 - Anti-persona pressure test (Layer 2 + 3)

Purpose/context: Most ICPs describe who you want to sell to. This prompt surfaces who you shouldn’t sell to - and why their pain is a mirage.

Act as a sharp-eyed B2B positioning expert. Our current ICP is:

[PASTE YOUR ICP ONE-PAGER]

Generate a "do-not-sell" anti-persona by answering:
1. Which titles in our current ICP look great on paper but churn inside
   90 days? What signal in the first 30 days predicts the churn?
2. Which industries match our firmographics but have a different
   decision-making logic (e.g., they need a champion with 2 levels of
   authority we can't reach)?
3. Which ARR band looks like our sweet spot but actually has a
   different unit economics (e.g., they buy on price, not outcome)?
4. What is the "false-positive" buyer persona - looks like a fit, but the
   real decision is being made by someone we never talk to?
5. For each anti-persona, write the one question we should ask in the
   first discovery call to disqualify them in 5 minutes.

Be ruthless. If our ICP holds up, this should be a short answer.

Example output: Anti-persona A: VP of Sales at a 10–25 person startup. Looks like a fit. Reality: no procurement process, will churn at renewal when the founder takes the call back. Disqualifying question: “Who else needs to be in our first call besides you?”

Pro tips:

Run this quarterly. Your anti-persona list grows as you do.
Compare output against your last 10 lost deals. Patterns will line up.
Don’t soften the output. The whole point is to be cut.

Prompt 3 - “What keeps them up at night” extraction (Layer 1 + 3)

Purpose/context: Surfaces the latent anxiety that drives a buyer’s search, not the symptoms they put in a brief.

You are a B2B buyer psychology researcher. I will paste 10–20 raw text
samples - could be call snippets, G2 reviews, support tickets, or
LinkedIn DMs - from buyers in [PERSONA / SEGMENT].

Do not summarize. Instead, find the underlying anxieties. For each
anxiety, list:
- The anxiety in 6 words or fewer (a "headline")
- 2 direct quotes that hint at it
- The downstream business consequence the buyer fears
- The role the buyer is internally playing (e.g., "the fixer," "the
  visionary defender," "the person who'll get blamed")
- A "trigger phrase" - the literal words this buyer uses when the
  anxiety surfaces

Then rank the top 5 anxieties by frequency × emotional intensity.
Output as a table.

Example output:

Headline	Quote 1	Quote 2	Fear	Internal role	Trigger phrase
”Audit risk exposure"	"We had a near miss in Q3"	"I don’t want to be the email”	CFO scrutiny	The shield	”compliance gap"
"Tooling sprawl tax"	"Nobody knows what we pay for"	"We’re paying for ghosts”	Q3 budget cuts	The auditor	”shelfware”

Pro tips:

Aim for 20+ samples, not 5. Patterns need volume.
Use the trigger phrase list in your ad copy, sales scripts, and landing page H1s. It’s money.
The “internal role” column is the most undervalued. Personas act roles, not titles.

Prompt 4 - ICP document rebuilder (Layer 1 + 2)

Purpose/context: Use this when you have an existing ICP doc that feels stale, and you need to merge what you’ve learned in the last 90 days.

You are an ICP strategist. Our current ICP document is below. Our last
90 days of customer evidence is also below (churn reasons, win/loss
notes, top 3 support tickets, and 5 G2 reviews we didn't write).

[PASTE EXISTING ICP]
[PASTE RAW EVIDENCE]

Rewrite the ICP document with these sections:
1. One-paragraph ICP narrative (3 sentences max).
2. Firmographic filters (with confidence level: HIGH / MEDIUM / LOW).
3. 3 named buyer personas with a 1-sentence "pains they can't solve
   without us" line per persona.
4. 3 trigger events that put this buyer in-market.
5. 3 "anti-pain" hypotheses - things this buyer explicitly does NOT
   care about (so we can stop pitching them).
6. A "what would make us wrong" section. Be honest.

Output in clean markdown.

Example output: ICP narrative: Mid-market RevOps leaders (50–500 FTE) at B2B SaaS companies who have hit the “spreadsheet ceiling” on pipeline reporting and need a board-ready source of truth within 30 days. Trigger event: a new VP of Sales or CRO joining within the last 6 months.

Pro tips:

The “anti-pain” section saves real money in ads. Most B2B waste comes from pitching pains the buyer doesn’t feel.
Re-run this every quarter. The 90-day window matters.
Don’t edit the output for tone. Edit for accuracy only.

Prompt 5 - Hidden org-chart decoder (Layer 2 + 3)

Purpose/context: Buying committees are no longer linear. This prompt maps the real power map, not the org chart.

You are an enterprise sales strategist. Below is a list of people we
interacted with during a closed-won deal in [SEGMENT]:

[PASTE LIST OF NAMES, TITLES, AND INTERACTION NOTES]

Map the real decision-making network. For each person, label them:
- Champion (advocates internally)
- Coach (gives you intel, no advocacy)
- User (cares about outcomes, no authority)
- Blocker (can kill the deal)
- Skeptic (silent veto holder)
- Ignored (talks a lot, no power)

Then:
1. Identify the "economic buyer" - the one person who signs because of
   business outcome, not feature.
2. Identify the "trust anchor" - the person other stakeholders defer to.
3. List the 2 people we should NEVER have skipped, even if our CRM says
   we talked to them.
4. Suggest the 1 question to ask the Champion in the next call to map
   the rest of the committee.

Example output: Economic buyer: VP RevOps. Trust anchor: Director of Sales Enablement. Skipped: Senior Manager of Procurement (silently killed renewal last year). Champion question: “When you imagine this rolling out internally, who do you call first to validate the decision?”

Pro tips:

Run this on the last 5 closed-won and the last 5 closed-lost deals. The lost deals are where the org chart lies the most.
The “ignored” label is the most underused. Watch for the loud-but-powerless buyer.
Save the output to your CRM as a custom field per deal.

Section 2: Review & community mining (prompts 6–10)

Your buyers write the most honest descriptions of your pain in places you don’t control. These prompts turn G2, TrustRadius, PeerSpot, Capterra, Reddit, Slack communities, and niche Discord servers into ICP fuel.

“B2B review content from G2, TrustRadius, PeerSpot, and, yes, the robust discussions on Reddit substantially informs the LLMs that feed results on ChatGPT, Claude, Perplexity, Microsoft Copilot, and other AI-driven answer engines.” - Forrester, June 8, 2026.

Prompt 6 - Competitor 1-star miner (Layer 1)

Purpose/context: Your competitor’s worst reviews are your roadmap and your sales pitch. Don’t read them like a customer - read them like a researcher.

Act as a competitive intelligence analyst. I will paste 15–30 one- and
two-star reviews of [COMPETITOR NAME] from G2, TrustRadius, and
Capterra. I want the pain gold, not the rants.

For each cluster of complaints, do the following:
1. Group the complaints into 3–5 "pain clusters" with descriptive names.
2. For each cluster, write a 1-sentence "buyer quote" that captures
   the emotional core.
3. List the 3 product features the reviewer wishes existed.
4. Identify which persona (by title) wrote the most emotional reviews
   in each cluster.
5. Write a 2-sentence "headline" we could A/B test on a landing page.
6. Flag any cluster that aligns with a feature we already ship - these
   are the easiest wins to copy into our messaging.

Example output: Pain cluster: “Reporting that doesn’t lie.” Buyer quote: “I’d rather pull a report in Excel than trust their dashboard.” Wished-for feature: scheduled CSV exports with version history. Persona: Director of RevOps. Landing page headline: “Reports your CFO will believe. Built in 30 seconds.”

Pro tips:

Mine at least 3 competitors per quarter. The clusters you can ignore are the clusters you should be selling against.
Pull reviews older than 6 months too. Newer reviews often hide a competitor’s old pain.
Don’t paste competitor names into the prompt if you don’t have to - frame the work as “the category,” not “them.”

Prompt 7 - Review-to-JTBD translator (Layer 1 + 2)

Purpose/context: Reviews describe symptoms. JTBD describes the job the buyer is hiring a product to do. This prompt bridges the gap.

I will paste 20+ reviews of [CATEGORY / TOOL] from G2 and Reddit.
For each review, extract:
1. The job the buyer was "hiring" the product to do (in 8 words or fewer).
2. The "metric" the buyer used to decide if the product succeeded.
3. The "anti-job" - what the buyer explicitly didn't want the product
   to do.
4. The "context" - when, in their day, they reach for the product.
5. The "emotional baseline" - anxious, hopeful, relieved, proud.

Then synthesize:
- Top 5 jobs across all reviews.
- Top 3 metrics that correlate with 5-star reviews.
- The 1 anti-job that shows up in 1-star reviews (a deal-breaker).
- A "job statement" in the format: When [situation], I want to [job],
  so I can [outcome].

Output as a clean table.

Example output:

Job	Metric	Anti-job	Context	Emotional baseline
”Stop the meeting about the meeting”	Hours saved per week	”Be a CRM”	Before every QBR	Anxious

Job statement: When I have to defend pipeline in a QBR, I want to generate a clean forecast in 15 minutes, so I can spend the meeting talking about deals, not data.

Pro tips:

5-star and 1-star reviews often share the same job but disagree on the metric. That’s where positioning lives.
Save the job statements. Use them as your homepage H1 candidates.
Re-run on reviews of adjacent tools (Excel, Notion, a competitor) to see if your job is broader than your category.

Prompt 8 - Reddit thread archeology (Layer 2 + 3)

Purpose/context: Reddit and niche Slack/Discord communities are where buyers say what they really think. The phrasings are raw, the emotional cues are high, and the LLMs that feed answer engines (Forrester, June 2026) are already indexing them.

I will paste 5–10 Reddit threads (titles + top comments + 2 replies each)
from r/[SUBREDDIT] or r/[NICHE COMMUNITY] where buyers discuss
[CATEGORY / TOOL / PROBLEM].

For each thread, do the following:
1. Write the "true question" the original poster actually wanted
   answered (often different from the title).
2. Identify the 2 most-upvoted comments. Why did they resonate?
3. Extract any "war stories" - first-person accounts with a specific
   dollar amount, time cost, or failure described.
4. Find the "vocabulary gap" - words the buyer uses that our
   marketing site does NOT use. List them.
5. Identify the "silent majority" - the implicit assumption everyone
   in the thread shares but nobody says.

Then, across all threads, output:
- A "voice memo" - 5 sentences in the buyer's actual voice that we
  could use verbatim in ad copy.
- A list of 3 product objections we didn't know existed.
- A list of 3 features the buyer is clearly hand-rolling.

Example output: True question: “Is there a tool that doesn’t make me look like I don’t know what I’m doing in front of my CEO?” War story: “I spent 6 hours rebuilding a dashboard the night before a board meeting.” Vocabulary gap: “CFO-grade,” “board-safe,” “no-surprise.” Silent majority: every poster is one bad quarter away from being replaced.

Pro tips:

Sort by controversial, not by top. The upvoted comments are consensus. The controversial ones are the ICP.
Always include the original post body, not just titles. The phrasing matters.
Build a rolling doc of “voice memos.” They’re your best ad copy, ever.

Prompt 9 - Community pain harvest (Layer 1 + 2)

Purpose/context: Slack communities, Discord servers, and niche forums (RevOps Coop, DemandGen Slack, Pavilion, OnDeck) are gold. This prompt organizes raw signal.

I'm going to paste 30+ messages from a private Slack community
([COMMUNITY NAME]) discussing [TOPIC / TOOL / CATEGORY].

Treat each message as a data point. Then:
1. Cluster the messages into 5–7 "pains" with clear names.
2. For each pain, quote the single most vivid 1-sentence message
   verbatim.
3. Identify the "10xer" - the 1 message that, if we solved it,
   would eliminate 10 other pains. (Usually a workflow or
   integration complaint.)
4. List the tools mentioned (positive AND negative).
5. Find the "silent trend" - a pain nobody's named but 3+ messages
   point to obliquely.
6. Recommend 1 conversation we should start, not join, in this
   community next week.

Example output: Cluster: “Tooling that requires 2 admins to maintain.” Vivid quote: “We have a part-time contractor whose only job is keeping our RevOps stack from falling over.” 10xer: a real-time data sync between CRM and the BI tool. Silent trend: nobody’s said the word “audit” but 4 messages mention compliance, SOC 2, and “the email no one wants to send.”

Pro tips:

Don’t paste anything from communities that prohibit sharing (paid Slacks, gated Discords). Use summaries, not raw exports.
Run this monthly. Communities have seasons. The pain in January is rarely the pain in July.
The “10xer” output is often the seed of a new product line, not a feature.

Prompt 10 - Review-site AEO audit (Layer 3)

Purpose/context: In 2026, your G2 / TrustRadius / PeerSpot presence feeds the answer engines your buyers use to choose vendors. This prompt audits your footprint.

You are an Answer Engine Optimization (AEO) strategist for B2B SaaS.
I will paste our company's reviews, ratings, and category presence on
G2, TrustRadius, PeerSpot, and Capterra. I will also paste 3–5
competitor profiles on the same platforms.

Do the following:
1. Identify the 5 review "themes" most associated with our 5-star reviews.
2. Identify the 5 review "themes" most associated with our 1- and
   2-star reviews.
3. Cross-reference with competitors. Which of our 5-star themes do
   competitors *also* claim? (Table stakes.) Which are unique to us?
4. Flag any negative theme that, if surfaced in a ChatGPT or
   Perplexity answer, would actively hurt our win rate. Suggest a
   counter-message.
5. Propose 3 "review-site content plays" - specific things we should
   ask happy customers to mention in their next review to feed
   answer engines (per Forrester, June 2026: 94% of B2B buyers
   use answer engines during search).
6. Suggest 1 "best of" list (e.g., "Best for X") we should target on
   G2 in the next 90 days and why.

Example output: Table-stakes 5-star theme: “Easy to use.” Unique theme: “Forecasts I can defend in a board meeting.” Hurt-rated theme: “Slow support” (counter: a public SLA dashboard). Review prompt: ask champions to mention “board meeting” or “QBR” by name. Target list: “Best Revenue Intelligence Tools for Mid-Market.”

Pro tips:

AEO is the new SEO. Forrester (June 2026) makes this explicit. Treat your review profiles as landing pages for LLMs.
Ask customers to use specific phrasings in reviews. It’s allowed, ethical, and high-leverage.
Re-run this quarterly. Answer engines update their source weighting often.

Section 3: Call & transcript mining (prompts 11–15)

Calls are the most underused ICP source. Every Gong, Chorus, Grain, Fathom, Tactiq, or Zoom transcript is a structured document hiding inside an unstructured mess. These prompts turn call libraries into ICP fuel.

Prompt 11 - Discovery call JTBD extractor (Layer 1 + 2)

Purpose/context: Pull jobs-to-be-done from a single discovery call transcript. The output is a per-deal job statement you can act on.

Act as a JTBD interview analyst. I will paste the transcript of a
discovery call. Do not summarize the call. Instead, extract:

1. The "functional job" - the concrete task the buyer is trying to
   accomplish.
2. The "emotional job" - how the buyer wants to feel when the job
   is done.
3. The "social job" - how the buyer wants to be perceived by peers,
   boss, or board.
4. The "consumption chain" - the 4–6 steps the buyer takes today
   to get the job done, with the worst step called out.
5. The "hired" tools - the products the buyer mentioned using
   today (positive or negative).
6. The "fired" tools - the products the buyer mentioned quitting.
7. The "constraints" - the 2–3 things the buyer said they cannot
   do (budget, time, politics).
8. A "big-fish-small-pond" score (1–10): how badly does the buyer
   need a vendor like us, and how many alternatives have they
   already tried?

Output as a structured report. Quote verbatim where possible.

Example output: Functional: “Build a board-deck-ready pipeline report in under an hour.” Emotional: “Stop dreading the QBR.” Social: “Look like the operator who fixed the data mess.” Worst step in the chain: data cleaning. Hired: Salesforce + a homegrown ETL. Fired: a previous BI tool. Big-fish score: 9.

Pro tips:

Run this on your lost deals. JTBD shows up loudest when the buyer chose someone else.
Save the output as a deal-level artifact, not a persona-level one. JTBDs are personal.
The “big-fish” score is the most underused call signal. Track it.

Prompt 12 - Win/loss pattern finder (Layer 2 + 3)

Purpose/context: Aggregate win/loss patterns across 20+ deals. Identifies the systemic pains your ICP either rewards you for or punishes you with.

I will paste transcripts (or written summaries) of 10 closed-won and
10 closed-lost deals from the last 90 days. For each deal, I have
noted: industry, segment, deal size, persona, and outcome.

Do the following:
1. Cluster the wins and losses by "buyer's primary pain." Give each
   cluster a name.
2. Identify the 3 pains that appear in 80%+ of the wins.
3. Identify the 3 pains that appear in 80%+ of the losses.
4. Find the "switching trigger" - the moment in the call where the
   buyer either leaned in or leaned out. Quote both.
5. Identify the 1 competitor or workaround that came up the most in
   losses.
6. List the 3 "we're not for you" signals that, if heard early,
   would have saved the deal cycle.
7. Recommend 1 ICP hypothesis to test in the next 30 days based on
   the pattern.

Example output: Win pain #1: “Can’t defend pipeline numbers in a QBR.” Loss pain #1: “Don’t have time for an implementation.” Switching trigger: when the buyer asked, “How fast can my CFO see value?” - wins said “10 days,” losses said “10 weeks.” Workaround that killed us: an in-house spreadsheet maintained by a hero employee. ICP hypothesis: target post-funding B2B SaaS companies 6–12 months after a CRO hire.

Pro tips:

Win/loss is the highest-ROI research a startup can do. Don’t outsource it to a panel - do it on real calls.
The “we’re not for you” signal is gold. Save it as a disqualification question in your discovery script.
Re-run quarterly. The patterns shift, especially post-PMF.

Prompt 13 - Competitor mention radar (Layer 1 + 3)

Purpose/context: Every call transcript mentions competitors. The way buyers mention them is a positioning goldmine.

I will paste 20+ call snippets where buyers mentioned [COMPETITOR
NAME]. For each mention, classify it as one of:
- ACTIVE EVAL (they're already in a trial)
- COMPARISON SHOP (tire-kicker)
- REPLACEMENT (switching away)
- DEFAULT (they think of you and the competitor as the same)
- ANTI-CHOICE (they've ruled them out)

Then:
1. For each category, find the exact phrase the buyer used.
2. Identify the 1 phrase that's the most damaging to the competitor
   in our favor.
3. Identify the 1 phrase we should NEVER say about the competitor
   (because the buyer uses it as a positive).
4. Suggest 3 battlecard updates based on the pattern.
5. Suggest 1 contrarian ad angle we could take based on what
   buyers *don't* say about the competitor.

Example output: Most damaging phrase: “We used them, but the dashboards never matched the CRM.” Default: buyers say “your tool or [Competitor]” interchangeably. Never say: “enterprise-grade” - buyers use that as a positive. Battlecard update: lead with “real-time CRM sync,” not “dashboards.” Ad angle: target the buyer’s hidden fear of the competitor (e.g., implementation cost, not features).

Pro tips:

“Default” mentions are the most dangerous. If buyers think you’re interchangeable, price becomes the only differentiator.
Active evals are the highest-priority sales pipeline. The phrasings tell you what the buyer’s already been told.
Re-run every 6 weeks. Competitive narratives shift fast in 2026.

Prompt 14 - Pricing objection autopsy (Layer 1 + 2)

Purpose/context: Pricing objections are rarely about price. They’re about ROI clarity, procurement friction, or fear of internal justification.

I will paste 15 call snippets where the buyer pushed back on price.
For each, identify:
1. The literal objection (verbatim).
2. The real objection underneath (usually 1 of 5: trust, ROI clarity,
   procurement friction, internal politics, or timing).
3. The 1 question the rep could have asked to surface the real
   objection earlier.
4. The 1 piece of evidence (case study, calc, peer quote) that, if
   shared earlier, would have unblocked the deal.
5. The persona most likely to raise this objection (and why).

Then, synthesize:
- Top 3 real objections across all 15 calls.
- 3 new sales assets we should build this quarter.
- 1 ICP refinement: is there a buyer segment for whom our price is
  a feature, not a bug?

Example output: Literal: “It’s 30% more than we budgeted.” Real: ROI clarity - they don’t have a defensible model. Earlier question: “Walk me through how you’ll measure ROI to your CFO in the first 90 days.” Evidence needed: a CFO-ready ROI calculator. Persona: VP-level, risk-averse, first-time buyer of the category.

Pro tips:

If “real objection” is procurement friction, your ICP may be too high in the org. Drop a tier.
If “real objection” is internal politics, your champion may be the wrong champion. Map the org again.
Pricing-objection autopsies often surface your pricing page problem, not your price problem.

Prompt 15 - Buying signal classifier (Layer 3)

Purpose/context: Not every signal is buying intent. This prompt distinguishes curiosity from commitment using call transcript data.

I will paste 30 call snippets. Classify each one as one of:
- CURIOSITY (information gathering, no timeline)
- EVALUATION (comparing options actively)
- DECISION (vendor selected, negotiating terms)
- INERTIA (status quo, no movement)
- POLITICS (internal alignment, not product fit)

For each, output:
1. The 1 phrase that signals the category.
2. The next-best action the rep should take in 24 hours.
3. The 1 question to ask to move the buyer to the next stage.
4. A red flag (if any) that suggests the deal is in trouble.

Then, across all snippets:
- Distribution (% of each category).
- Top 3 "evaluation" phrases we should listen for in real time.
- 1 ICP signal that we keep misclassifying.

Example output: Distribution: 40% curiosity, 30% evaluation, 10% decision, 15% inertia, 5% politics. Phrase for evaluation: “We’re looking at 3 vendors by end of month.” Next-best action: send a vendor comparison doc tailored to the 3. ICP misclassified: we keep calling “Q4 budget freeze” a real objection, but it’s actually evaluation timing.

Pro tips:

Politics is the most underweighted category. Buyers stuck in internal alignment are a time bomb.
Inertia is a feature, not a bug, sometimes. Some “leads” should never have been leads.
Run this weekly. The distribution shifts with your messaging.

Section 4: JTBD & outcome prompts (prompts 16–20)

JTBD is the most underrated ICP tool in 2026. These prompts turn feature lists into jobs the buyer is hiring you to do, with measurable outcomes.

Prompt 16 - Outcome statement architect (Layer 2)

Purpose/context: Every B2B pitch should end with a measurable outcome the buyer can defend internally. This prompt builds those statements.

You are a JTBD + outcomes strategist. Our product does [1-SENTENCE
DESCRIPTION]. Our buyers are [ICP]. Their primary pains are [LIST 3
PAINS].

For each pain, generate 5 "outcome statements" using this format:
- Outcome: [What changes for the buyer]
- Metric: [How the buyer measures it]
- Timeframe: [When they expect to see it]
- Defensible to: [Which internal stakeholder the buyer would
  quote this outcome to]
- Probability of being true: [HIGH / MEDIUM / LOW based on our
  current data]

Then pick the top 1 outcome per pain. Rewrite it as a single
sentence a buyer could say to their boss:

"By [timeframe], we will [outcome] so that we can [business
impact]."

Example output: Pain: “Can’t defend pipeline numbers in a QBR.” Outcome: “A board-ready pipeline report in 15 minutes.” Metric: hours saved per QBR. Timeframe: within 30 days. Defensible to: CFO, CRO. Defensible sentence: “By end of Q3, we will cut QBR prep time from 8 hours to 1 hour, so that I can stop defending tools and start defending pipeline.”

Pro tips:

The “defensible to” column is the secret weapon. If the buyer can’t say it to that person, it’s not a real outcome.
Aim for outcomes that map to a board metric (revenue, churn, CAC). Operational outcomes are table stakes.
Re-run quarterly. Outcomes change as your product matures.

Prompt 17 - “Forced trade-off” pain finder (Layer 3)

Purpose/context: The most painful jobs are the ones the buyer is forced to do badly. This prompt surfaces the trade-offs the buyer hates.

I will paste 10+ call snippets, support tickets, or G2 reviews where
buyers describe their current workflow.

For each, identify the "forced trade-off" the buyer is making. A
forced trade-off looks like:
"They do X, but they hate that it costs them Y."

Then:
1. Cluster the trade-offs into 3–5 themes.
2. For each theme, write the buyer's internal monologue in 3
   sentences.
3. Identify the 1 trade-off that, if we removed it, would
   eliminate 3+ downstream complaints.
4. Suggest 1 product positioning angle that names the trade-off
   directly in the headline.
5. Suggest 1 case study angle that proves we can remove the
   trade-off in 30 days.

Example output: Theme: “Accurate data, but it takes all week.” Monologue: “I trust the data, I just can’t get to it. By Friday, I give up and pull numbers from memory.” Trade-off to remove: the time cost of trustworthy data. Headline: “Pipeline data you can defend, in 15 minutes.” Case study angle: a customer who cut QBR prep from 6 hours to 45 minutes.

Pro tips:

Forced trade-offs are where AI pitches win. “We removed the trade-off with AI” is the 2026 buyer’s dream.
The monologue format is your most underused copy asset. Use it on landing pages.
The 1 trade-off that removes 3 complaints is your wedge feature.

Prompt 18 - Feature-to-JTBD translator (Layer 1 + 2)

Purpose/context: Most B2B websites list features. Few explain the job. This prompt turns your feature list into JTBD statements buyers can repeat internally.

Here is our product feature list:

[PASTE FEATURE LIST]

For each feature, generate:
1. The functional job it does (in the buyer's voice).
2. The emotional job it does.
3. The social job it does.
4. A "day-in-the-life" sentence showing the feature in use.
5. A JTBD statement: When [situation], I want to [job], so I can
   [outcome].
6. A "kill-the-feature" verdict: if we removed this feature,
   which persona would notice first? (If nobody notices, kill it.)

Then synthesize the top 3 features that map to the deepest job
(usually emotional + social, not just functional).

Example output: Feature: “Slack alerts for deal stage changes.” Functional job: “Know the moment a deal slips.” Emotional job: “Stop being surprised in pipeline reviews.” Social job: “Look like I have my finger on the pulse.” Day-in-the-life: “I’m in a customer call when Slack buzzes - I save a deal that would have been a ‘no’ by Friday.” Kill verdict: 1% of customers would notice. The persona who would: RevOps leaders who run weekly forecast calls.

Pro tips:

The “kill-the-feature” verdict is brutal and necessary. Run it every 6 months.
Social jobs are the most underused in B2B. Use them in ad copy.
The day-in-the-life sentence is your best sales-asset format. Make 20 of them.

Prompt 19 - JTBD interview script generator (Layer 2)

Purpose/context: Use this to generate a Mom-Test-compliant interview script for live customer calls. The Mom Test by Rob Fitzpatrick (momtestbook.com) is the gold standard.

Act as a customer discovery expert trained on The Mom Test (Rob
Fitzpatrick) and the JTBD framework (Tony Ulwick, Bob Moesta).

I want to interview 5 customers in [PERSONA / SEGMENT] about
[CATEGORY / PROBLEM]. Generate a 30-minute interview script with
exactly:
- 5 context questions (about their world, not your product).
- 5 problem-validating questions (about the last time the problem
  happened).
- 5 workarounds questions (what they did before they had a tool).
- 5 "would you pay" validation questions (without asking "would
  you pay").
- 5 "what would have to be true" closing questions.

Constraints:
- No "do you like our product" questions.
- No hypotheticals.
- No leading questions.
- Every question must be answerable in a single sentence.

Output as a clean script with follow-up probes for each question.

Example output: Context: “Walk me through the last time you had to defend pipeline to your CEO.” Probe: “What was the first thing you did?” Problem: “What did the prep cost you, in hours?” Probe: “How did you feel halfway through?”

Pro tips:

Don’t pay for interviews. Pay in attention, not incentives.
Record (with consent). Transcribe with Grain, Fathom, or Tactiq. Then run prompts 11–15 on the transcript.
Run 5 interviews per segment. Patterns surface at 5, harden at 10.

Prompt 20 - Outcome metric mapper (Layer 2 + 3)

Purpose/context: The buyer doesn’t buy features. They buy outcomes they can measure. This prompt identifies the exact metrics your ICP uses to judge success.

You are a B2B positioning strategist. I will paste 10+ call snippets
and 5+ G2 reviews for [CATEGORY / PRODUCT].

Identify:
1. The 5 metrics buyers explicitly mention when describing success.
2. The 3 metrics buyers hint at but never say by name.
3. The 1 metric buyers would defend to their CFO (the "CFO metric").
4. The 1 metric buyers would defend to their team (the "team
   metric").
5. The 1 metric buyers would defend to themselves (the "personal
   metric").
6. The "lie metric" - a metric buyers say they care about but
   don't actually measure.
7. The "anti-metric" - what buyers track to know they're failing.

Then:
- Map each metric to a feature in [YOUR PRODUCT].
- Map each metric to a buyer title.
- Recommend 1 new metric to *introduce* into the category (the
  metric nobody tracks yet, but everyone will in 12 months).

Example output: CFO metric: pipeline coverage ratio. Team metric: hours saved per QBR. Personal metric: not getting “the email.” Anti-metric: tool adoption rate (high adoption can mean high bloat). New metric to introduce: “forecast confidence score” (0–100).

Pro tips:

The “lie metric” is the most important output. Buyers say they care about adoption. They actually care about not getting fired.
Personal metrics drive emotional urgency. Use them in ad copy.
The new metric is your category-creation wedge. Own it.

Section 5: Champion mapping & org-chart prompts (prompts 21–25)

Your ICP isn’t a title. It’s a network. These prompts help you find the humans behind the logo - the champions, the coaches, the silent veto holders, and the people you’ll never meet who can kill your deal.

Prompt 21 - Champion profile builder (Layer 1 + 2)

Purpose/context: Build a profile of the buyer most likely to advocate for your product internally - not just buy it.

Act as a B2B sales enablement strategist. I will paste notes from
5 closed-won deals in [SEGMENT]. For each, the deal notes mention
the person who was the internal advocate.

For each champion, identify:
1. Their title and 1-line job description.
2. The metric they're personally evaluated on.
3. The internal "enemy" they fight (could be a competing
   initiative, a boss, or a status quo).
4. The 1 thing they would say to a peer in a private Slack channel
   to convince them to buy.
5. Their "political cover" - what they need from a vendor to look
   smart internally.
6. Their "career risk" - what they're personally afraid of if
   this goes wrong.

Then synthesize:
- A composite champion profile (titles, metrics, fears).
- The 3 things a champion needs to hear in the first call to
  become an advocate.
- The 1 question that disqualifies a buyer from being a champion
  (i.e., they're a buyer, not a champion).

Example output: Champion profile: RevOps leader, evaluated on forecast accuracy, fights a status-quo CFO, needs a vendor who makes them look proactive. Political cover: “I brought in a tool that’s already paying for itself.” Career risk: “If this doesn’t work, I own the spend.” Disqualifying question: “Will you present this to the CFO yourself, or do you need us to do it?”

Pro tips:

A champion without political cover is a buyer. Buyers don’t sell for you.
The “internal enemy” is the most underweighted data point. Champions need a villain.
Save the disqualification question. It saves cycles.

Prompt 22 - Buying center radar (Layer 2 + 3)

Purpose/context: Map the full buying center, not just the title on the email signature.

You are a B2B sales strategist. I will paste a deal summary that
includes the people we interacted with (titles, interactions,
emails, meeting notes). Map the buying center.

For each person, label:
- Champion (will advocate internally)
- Coach (gives intel, no advocacy)
- User (cares about outcomes, no authority)
- Approver (signs off, doesn't care about product)
- Blocker (can kill, won't say so)
- Skeptic (silent veto holder)
- Ignored (talks a lot, no power)

Then:
1. Identify the "economic buyer" (signs because of business
   outcome, not features).
2. Identify the "trust anchor" (others defer to them).
3. Identify the "shadow buyer" - the person not in the deal notes
   but inferable from the org context.
4. List 2 people we should NOT skip, even if CRM says we spoke
   to them.
5. Suggest the 1 question to ask the Champion in the next call
   to map the rest.

Example output: Economic buyer: VP RevOps. Trust anchor: Director of Sales Enablement. Shadow buyer: Senior Manager of Procurement (signed last year, never showed up on calls). Skip-list: VP of IT (silent veto) and a Senior Director of Finance (will block if forecast looks off). Next question: “When you imagine this rolling out internally, who do you call first to validate?”

Pro tips:

The “shadow buyer” is the most underdiagnosed deal-killer in 2026. Procurement, legal, and IT are usual suspects.
Don’t ask the champion to map the org for you. Ask open questions and map it yourself.
Re-run this on every deal in your pipeline monthly.

Prompt 23 - Power-vs-interest grid (Layer 2)

Purpose/context: The classic 2x2, but in 2026, “interest” is not the same as “engagement.” This prompt updates the model.

Act as a change-management strategist. I will paste a list of
people involved in a deal, with their role, title, and a 1-line
description of their current engagement.

Build a 2x2 grid:
- X-axis: Power (low → high)
- Y-axis: Interest in *outcome*, not in our product (low → high)

Quadrants:
- "Move with us" (high power, high interest) → Champions.
- "Keep satisfied" (high power, low interest) → Brief regularly.
- "Keep informed" (low power, high interest) → Users, future
  champions.
- "Monitor" (low power, low interest) → Likely not worth effort.

For each quadrant, output:
- Names of people in our deal.
- 1-line strategy for each.
- 1 risk if we mis-classify.

Then, identify the 1 person we should move from "monitor" to
"keep informed" by raising their interest.

Example output: Move with us: VP RevOps (champion). Keep satisfied: VP IT (signs off on security, doesn’t care about product). Keep informed: 3 end users (loving it). Monitor: 1 procurement manager (low power, low interest). Risk: ignoring VP IT will kill the deal. Move: send procurement a 1-pager tailored to their language (compliance, not features).

Pro tips:

“Power” includes informal power. A 10-year admin can outrank a 6-month VP.
Interest in outcome ≠ interest in your product. Disambiguate.
The 1 person to “move” is usually the deal-saver.

Prompt 24 - Champion enablement brief (Layer 1)

Purpose/context: Build a 1-page brief your champion can forward internally without you in the room.

You are a sales enablement copywriter. Our champion (name, title,
segment) needs to defend our solution internally to a CFO, a VP
IT, and a Director of Sales Enablement.

Create a 1-page "champion brief" with:
1. A 1-paragraph executive summary written in the champion's
   voice (not our brand voice).
2. The 3 business outcomes the CFO will care about, with
   defensible numbers.
3. The 3 security / IT facts the VP IT will ask about, with the
   one-liner answer.
4. The 3 reasons the Director of Sales Enablement will love
   this.
5. The 3 objections each stakeholder will raise, with the
   champion's pre-loaded response.
6. A "if you only have 5 minutes" version of the brief.
7. A "if you have 30 minutes" version (the full thing).

Constraints:
- Plain text only (no logos, no design).
- Defensible numbers only (cite source).
- Voice: peer-to-peer, not vendor-to-buyer.

Example output: CFO ask #1: “Cut QBR prep from 8 hours to 1 hour.” Source: Customer case study, [Company], Q1 2026. VP IT one-liner: “SOC 2 Type II, SSO via Okta, single-tenant data isolation.” Pre-loaded objection (CFO): “We bought a tool like this 2 years ago and it failed.” Response: “Different. This one takes 10 days to first value, not 10 weeks.”

Pro tips:

The champion’s voice is non-negotiable. If your brand voice bleeds in, the brief gets ignored.
“Defensible numbers” only. Soft claims (“we save time”) lose in internal rooms.
Update the brief every quarter with fresh case studies.

Prompt 25 - Org-chart evolution tracker (Layer 3)

Purpose/context: Buyers change roles. Your ICP from 6 months ago may have moved companies. This prompt keeps your ICP alive.

You are a B2B data strategist. I will paste a list of 50+ contacts
from our CRM, with title, company, last interaction date, and a
deal stage.

For each, predict:
1. Probability they have changed roles in the last 6 months
   (HIGH / MEDIUM / LOW).
2. Probability their company has changed ICP fit (e.g., headcount
   band, segment).
3. The single data point (LinkedIn, hiring spree, funding round)
   most likely to invalidate our current ICP assumption.
4. Recommended next action (re-engage, archive, re-qualify).

Then, synthesize:
- The 10 contacts most likely to have changed roles.
- The 5 companies most likely to have changed ICP fit.
- The 1 macro ICP trend (e.g., "VP Sales title is dying") implied
  by the data.

Example output: HIGH change probability: 12 contacts. Macro trend: “VP of RevOps” is replacing “VP of Sales Ops” as the buying title in mid-market SaaS. Recommended action: re-engage 8 contacts with a “what’s changed” email.

Pro tips:

LinkedIn is your best org-chart oracle. Use it.
Title changes often signal ICP changes. Track them.
Run this quarterly. Your ICP is a living document.

Section 6: Trigger-event & moment prompts (prompts 26–31)

The best time to reach a buyer is mid-trigger, not mid-funnel. These prompts help you detect and act on the moments that put your ICP in-market.

Prompt 26 - Trigger event library builder (Layer 1 + 2)

Purpose/context: Build a comprehensive list of the trigger events that put your ICP in-market. Trigger events are the moments between “fine” and “fix this now.”

You are a B2B demand generation strategist. Our ICP is [ICP]. Our
product helps with [OUTCOME].

List 30 trigger events that put this buyer in-market. For each:
1. The trigger in 6 words or fewer.
2. The observable signal (LinkedIn post, job posting, press release,
   earnings call phrase, etc.).
3. The persona most likely to surface the trigger.
4. The time window in which the buyer is most receptive
   (typically 30–90 days after the trigger).
5. The 1 message we should send in that window.

Group triggers by category:
- Organizational (new hire, reorg, promotion)
- Financial (funding round, budget cycle, missed quarter)
- Operational (system outage, audit miss, compliance flag)
- Strategic (new market, M&A, board mandate)
- Personal (career move, performance review, peer pressure)

Example output: Trigger: “New VP of Sales hired.” Signal: LinkedIn announcement in last 30 days. Persona: VP Sales + their chief of staff. Window: 60–90 days. Message: “Most new VPs of Sales rewrite the playbook in their first quarter. Here’s the playbook 12 of them used.”

Pro tips:

Trigger events are 5–10x more predictive than firmographics.
Run a weekly scan for trigger events on your top 100 accounts.
The message in the window matters more than the trigger itself.

Prompt 27 - “Switching cost moment” detector (Layer 2 + 3)

Purpose/context: The buyer’s current tool isn’t broken - it’s just painful enough that a switch is now justifiable. Find the moment.

Act as a B2B behavioral strategist. I will paste 20+ call snippets
or support tickets where buyers described their current tool.

For each, identify the "switching cost moment" - the moment in
the conversation where the buyer admitted that staying with the
current tool is *more painful* than switching.

For each moment, output:
1. The literal phrase the buyer used.
2. The "tipping point" - the threshold that, if crossed, would
   force a switch.
3. The internal story the buyer tells themselves to justify
   *not* switching yet.
4. The external event that would break that story.
5. The 1 thing our sales team can do to make the switch story
   easier to tell.

Then, synthesize the top 3 "switching cost moments" across all
samples and the 1 product change that would lower the cost for
all 3.

Example output: Phrase: “It’s not great, but it’s not worth the implementation pain.” Tipping point: “When my new boss asks why we’re paying for this and not using it.” External event: a leadership change. Story to break: “We’ll deal with it next quarter.” Sales action: a 10-day implementation guarantee, not a 10-week one.

Pro tips:

The “story the buyer tells themselves” is the most underused concept in B2B. Break the story, and the deal is yours.
Switching cost moments are your highest-converting trigger events.
Re-run monthly. The story shifts with every market change.

Prompt 28 - Funding-round trigger responder (Layer 1)

Purpose/context: New funding means new budgets. This prompt generates outreach for companies that just raised.

You are a B2B outbound strategist. I will paste a list of 10
companies that just raised funding rounds in [SEGMENT]. For each,
output:
1. The funding round details (amount, stage, lead investor).
2. The 2 most likely *first hires* the company will make in
   the next 90 days (based on the round and stage).
3. The 2 *first tool purchases* they'll make in the next 90
   days.
4. The 1 person at the company most likely to be the buying
   decision-maker for our category.
5. A 4-sentence cold email from the CEO/founder of [YOUR
   COMPANY] that:
   - References the round specifically (not generically).
   - Offers a peer story (a customer who did the same thing
     12 months ago).
   - Asks for a 15-minute call, not a demo.
   - Has a P.S. with a 1-line "if not now, when?" nudge.
6. The 1 follow-up message for 14 days later if no reply.

Example output: Company: Acme SaaS, $12M Series A, led by [VC]. First hires: VP Sales, RevOps Lead. First tool purchase: forecasting software. Decision-maker: VP Sales. Email: “Congrats on the Series A. We work with 3 portfolio companies at [VC] who had the same go-to-market gap you probably have by Q3 - happy to share what they bought first.”

Pro tips:

Personalize the round, not just the company. “Congrats on the round” is table stakes. “Congrats on doubling down on RevOps” is signal.
Reference the lead investor’s portfolio. Buyers respect pattern recognition.
The 14-day follow-up should be a 1-line “no pressure” note, not a second pitch.

Prompt 29 - “Career-moment” trigger responder (Layer 1 + 2)

Purpose/context: A buyer’s personal career moment (new job, promotion, board seat) is the highest-intent trigger in B2B. This prompt personalizes for it.

You are a B2B lifecycle marketing strategist. I will paste a list
of 10 buyers who recently changed jobs or got promoted in
[INDUSTRY].

For each, output:
1. Their new role and the company.
2. The 3 "career moments" they're likely experiencing in the
   first 90 days (e.g., "build a quick win," "blaze a new
   initiative," "evaluate the inherited stack").
3. The 1 most likely first-priority initiative for our
   category.
4. A 3-touch outreach sequence (Day 1, Day 7, Day 21):
   - Day 1: a peer-congrats note, no pitch.
   - Day 7: a "the first 90 days playbook" asset.
   - Day 21: a 1-question email asking which of 3 common
     challenges is top of mind.
5. The 1 thing to *not* say (e.g., "I noticed you just
   started...").

Constraints:
- No flattery.
- No product pitch in touch 1.
- Voice: peer, not vendor.

Example output: Day 1: “Saw the news on LinkedIn. We work with 4 VPs of Sales who started in the same quarter. No agenda, but I’d love to send you the playbook they all used.” Day 7: a 1-page “first 90 days” doc with 3 customer stories. Day 21: “Of the 3 challenges below, which is the loudest right now?”

Pro tips:

Touch 1 with no pitch is the highest-converting cold outreach in 2026. Don’t ruin it.
The “first 90 days playbook” is the most underused offer in B2B.
“Noticed you just started” is creepy. Don’t say it.

Prompt 30 - Disqualification trigger finder (Layer 3)

Purpose/context: The best deals to disqualify are the ones you almost took. This prompt finds the disqualifying triggers before you waste a quarter on a bad fit.

You are a B2B sales strategist. I will paste 10 closed-lost deals
from the last 90 days. For each, identify the "disqualification
trigger" - the moment in the deal where we should have walked
away, but didn't.

For each, output:
1. The literal signal (verbatim if possible).
2. Why we missed it.
3. The opportunity cost (in time, focus, cycle) of not
   disqualifying earlier.
4. The 1 question we should have asked in the first call to
   surface the disqualifier.
5. The "soft disqualifier" - the polite, non-confrontational
   way to walk away.

Then, synthesize:
- Top 3 disqualification triggers across all 10 deals.
- 1 ICP refinement (a segment we should explicitly de-prioritize).
- 1 sales script change to surface the disqualifier earlier.

Example output: Trigger: “We’re looking at 3 vendors and you’re the most expensive.” Missed: we assumed price was a tactic, not a fact. Question we should have asked: “Have you already allocated budget, or is this a stretch use of an existing line?” Soft disqualifier: “Sounds like price is your top criterion. We may not be the best fit - happy to refer you to 2 vendors who are.”

Pro tips:

Disqualifying well is a competitive advantage. Most reps can’t do it.
Track opportunity cost. The cost of a bad deal is more than lost revenue - it’s lost focus.
Refine your ICP every quarter based on the disqualifiers.

Prompt 31 - “Why now” diagnostic (Layer 2 + 3)

Purpose/context: Every in-market buyer has a “why now.” If you can’t articulate it, you can’t prioritize the deal.

Act as a B2B sales coaching expert. I will paste notes from 10
discovery calls. For each, identify the "why now" - the specific,
time-bound reason the buyer is in-market *now*, not last quarter
or next quarter.

For each, output:
1. The "why now" in 1 sentence.
2. The evidence in the call (verbatim if possible).
3. The consequence of waiting (what happens to the buyer if
   they don't act in 90 days).
4. The internal stakeholder pushing the "why now."
5. The 1 question to ask the buyer to make the "why now" explicit
   to themselves (often the same question that reveals urgency).

Then, synthesize:
- The 3 most common "why now" patterns across all 10 calls.
- 1 ICP refinement: are we attracting buyers with the *right*
  why now?
- 1 messaging angle that names the most common "why now"
  directly in the headline.

Example output: Why now: “Board meeting in 8 weeks; need a defensible pipeline story.” Consequence: “Publicly miss the number, lose credibility with the board.” Stakeholder: CFO. Question: “What happens to you personally if you walk into that meeting with the same data as last quarter?”

Pro tips:

“Why now” is the most important question in B2B sales. If you can’t answer it, the deal will stall.
The consequence-of-waiting question is the deal-accelerator. Ask it early.
A “why now” tied to a calendar (QBR, board meeting, audit, funding) is 3x stronger than a vague one.

Comparison table: prompt categories vs. ICP layer vs. output

This is the cheat sheet. Use it to pick the right prompt, fast.

Prompt #	Category	Best for	ICP layer	Primary output	Time to run
1	Persona reconstruction	Rebuilding a stale persona	1 + 2	4-sentence day-in-the-life	20 min
2	Anti-persona pressure test	Cutting bad-fit pipeline	2 + 3	Do-not-sell list	15 min
3	Anxiety extraction	Surfacing latent fears	1 + 3	Top 5 anxieties table	25 min
4	ICP document rebuilder	Quarterly ICP refresh	1 + 2	Updated ICP doc	30 min
5	Hidden org-chart decoder	Mapping buying center	2 + 3	Power map	20 min
6	Competitor 1-star miner	Positioning vs. competitor	1	Pain clusters	20 min
7	Review-to-JTBD translator	Turning reviews into jobs	1 + 2	Job statements	25 min
8	Reddit thread archeology	Voice-of-customer research	2 + 3	Voice memo	20 min
9	Community pain harvest	Slack/Discord signal mining	1 + 2	Pain clusters + 10xer	30 min
10	Review-site AEO audit	Answer engine optimization	3	5 review themes + counter-messages	30 min
11	Discovery call JTBD	Per-deal job statement	1 + 2	JTBD report	15 min
12	Win/loss pattern finder	Pipeline retrospective	2 + 3	Win/loss patterns	45 min
13	Competitor mention radar	Battlecard updates	1 + 3	5 mention categories	25 min
14	Pricing objection autopsy	Pricing + packaging	1 + 2	Real objection list	25 min
15	Buying signal classifier	Pipeline prioritization	3	5-category distribution	20 min
16	Outcome statement architect	Sales enablement	2	Outcome sentences	20 min
17	Forced trade-off finder	Positioning angle	3	Trade-off + headline	20 min
18	Feature-to-JTBD translator	Marketing copy	1 + 2	Day-in-the-life + JTBD	30 min
19	JTBD interview script	Customer research	2	30-min interview script	20 min
20	Outcome metric mapper	New metric introduction	2 + 3	Metric map + new metric	30 min
21	Champion profile builder	Sales enablement	1 + 2	Champion profile	25 min
22	Buying center radar	Deal strategy	2 + 3	Power map + shadow buyer	30 min
23	Power-vs-interest grid	Stakeholder strategy	2	2x2 grid + risk flags	20 min
24	Champion enablement brief	Internal champion support	1	1-page champion brief	25 min
25	Org-chart evolution tracker	CRM hygiene + ICP refresh	3	10 contacts to re-engage	30 min
26	Trigger event library	Demand generation	1 + 2	30 trigger events	30 min
27	Switching cost moment	Outbound messaging	2 + 3	3 tipping points	25 min
28	Funding-round responder	Outbound at scale	1	10 personalized emails	45 min
29	Career-moment responder	Lifecycle marketing	1 + 2	3-touch sequence	30 min
30	Disqualification trigger	Sales discipline	3	3 disqualifiers + ICP refinement	30 min
31	”Why now” diagnostic	Deal acceleration	2 + 3	Why-now pattern	20 min

The 14-day ICP stress-test sprint

Here’s the playbook I’d run in your seat. Two weeks, part-time, with a team of 2–3.

Day 1–2 - Set the baseline. Run prompt 4 (ICP document rebuilder) on your current ICP. Save the output as icp-baseline-2026Q2.md. Goal: a clean starting point.

Day 3–4 - Mine the public web. Run prompts 6 (competitor 1-star miner), 7 (review-to-JTBD translator), and 8 (Reddit thread archeology) on 3 competitors and 3 review sites. Save each output as a separate file. Goal: external pain map.

Day 5–6 - Mine your own calls. Pull 10 discovery call transcripts from Gong, Chorus, Grain, or Fathom. Run prompts 11 (JTBD extractor) and 12 (win/loss pattern finder). Goal: internal pain map.

Day 7 - Synthesize. Run prompt 3 (anxiety extraction) on the union of all the above. Goal: top 5 latent anxieties.

Day 8–9 - Champion and org work. Run prompts 5 (hidden org-chart decoder), 22 (buying center radar), and 21 (champion profile builder) on your 5 most recent closed-won deals. Goal: power map of the buying center.

Day 10 - Disqualification audit. Run prompt 30 (disqualification trigger finder) on the 5 most recent closed-lost deals. Goal: the deals you should never have taken.

Day 11–12 - Trigger events. Run prompt 26 (trigger event library builder) for your top 100 accounts. Pick the 5 highest-priority trigger events. Run prompt 28 (funding-round responder) on those that just raised.

Day 13 - Refine the ICP doc. Run prompt 4 again, this time with all the new evidence. Save as icp-v2-2026Q2.md. Diff against the baseline. The diff is your ICP insight.

Day 14 - Ship it. Update your homepage H1, your ad copy, your discovery script, and your disqualification questions. Send a 1-page “what changed” memo to the team. Set a reminder to run the sprint again in 60 days.

The sprint takes ~3–5 hours per person per day. Output: a refreshed ICP, 31 prompt outputs saved as evidence, a buying-center map, a trigger-event list, and 3 new sales assets (champion brief, day-in-the-life, why-now question).

Common mistakes to avoid

Mistake 1: Treating ChatGPT output as ground truth. It’s a hypothesis engine, not an oracle. Always validate the top 3 patterns with a live customer call. If the buyer flinches, the pattern is wrong.

Mistake 2: Skipping the anti-persona. Most teams write ICPs they want to win. The real ICP insight is the anti-persona - the buyer you should disqualify. Run prompt 2 every quarter.

Mistake 3: Confusing observed pain with latent pain. Observed pain is in every review. Latent pain is in the workarounds, the silent trends, the vocabulary gaps. Latent pain is where category creation happens.

Mistake 4: Forgetting the trust anchor. In a 6–10 person buying committee, the economic buyer is rarely the trust anchor. Map both, separately. The trust anchor decides if you survive the second meeting.

Mistake 5: Running prompts once. An ICP is a living document. Run the library every 90 days. The prompt that didn’t surface anything in Q1 will surface gold in Q3, after 3 more months of call transcripts.

Mistake 6: Not using the Mom Test. ChatGPT can extract signal. It can’t build trust. Pair prompt 19 (JTBD interview script) with live, Mom-Test-compliant calls. The script is the scaffolding. The trust is the data.

Mistake 7: Confusing AEO with SEO. Answer engine optimization (Forrester, June 2026) is not a replacement for SEO. It’s an addition. Treat your G2, TrustRadius, PeerSpot, and Reddit presence as AEO surface area. Treat your blog and docs as SEO surface area. Both matter.

Mistake 8: Letting “AI-generated” be your differentiator. HubSpot’s 2026 State of Marketing found 80% of marketers use AI for content. The bar is quality, not AI usage. Use AI to find pain. Use humans to write the copy that lands.

Final word

The best ICP isn’t a doc. It’s a discipline. The 31 prompts above are a starting system, not a finish line.

Run them on the signal you already have. Reviews you’re ignoring. Calls you’re not transcribing. Reddit threads you’re not reading. Slack communities you’re not in. Every one of them is a place a buyer is telling you exactly what they need - in their own words, in their own moment.

Your job is to listen. ChatGPT’s job is to help you hear the pattern. The prompt library above is the bridge.

Pick three prompts. Run them this week. Update your ICP. Send the diff to your team. Repeat in 30 days.

The hidden pain is hiding in plain sight. The 2026 buyers are writing it for you. All you have to do is show up, mine it, and ship the answer.

Reader disclosure & educational-purpose notice

This page is published by SuperFreshAI for general informational and educational purposes only. By reading it, you agree to the points below.

Editorial independence. All reviews, guides, and recommendations are written by our editorial team based on hands-on use. Some links on this site are affiliate links, and some articles are produced as partner content — both are always clearly labeled. Our editorial conclusions are never shaped by partners or affiliates.
Not professional advice. Nothing on this page constitutes legal, financial, medical, tax, or other professional advice. AI tools, pricing, and capabilities change quickly — always verify current information with the tool's official documentation before making a decision.
Educational purpose only. The content here is intended to help you learn about AI tools and workflows. It is not a guarantee of results, performance, fitness for a particular purpose, or suitability for your specific situation. Your results may vary.
No warranties. The site and its content are provided on an "as is" and "as available" basis. We make no warranties, express or implied, about accuracy, completeness, reliability, or availability. See our Terms and Privacy for the full legal terms.
Your responsibility. You are responsible for how you use the information on this page, including any decisions you make based on it. Always do your own research and consult a qualified professional when appropriate.
Affiliate & partner disclosure. When you click certain outbound links, we may earn a commission at no extra cost to you. When a piece of content is produced as partner content, it is labeled at the top of the page. See our Editorial Policy for the full standards we follow.

By continuing to read, you acknowledge that you have read and understood this notice.

31 ChatGPT prompts for B2B startup founders to uncover hidden ICP pain points

Quick answer: what you’ll get from this prompt library

Why most B2B ICPs are wrong in 2026

The hidden-pain mining framework (HPM)

Section 1: ICP & persona reconstruction (prompts 1–5)

Prompt 1 - Persona ghost-protocol (Layer 1 + 2)

Prompt 2 - Anti-persona pressure test (Layer 2 + 3)

Prompt 3 - “What keeps them up at night” extraction (Layer 1 + 3)

Prompt 4 - ICP document rebuilder (Layer 1 + 2)

Prompt 5 - Hidden org-chart decoder (Layer 2 + 3)

Section 2: Review & community mining (prompts 6–10)

Prompt 6 - Competitor 1-star miner (Layer 1)

Prompt 7 - Review-to-JTBD translator (Layer 1 + 2)

Prompt 8 - Reddit thread archeology (Layer 2 + 3)

Prompt 9 - Community pain harvest (Layer 1 + 2)

Prompt 10 - Review-site AEO audit (Layer 3)

Section 3: Call & transcript mining (prompts 11–15)

Prompt 11 - Discovery call JTBD extractor (Layer 1 + 2)

Prompt 12 - Win/loss pattern finder (Layer 2 + 3)

Prompt 13 - Competitor mention radar (Layer 1 + 3)

Prompt 14 - Pricing objection autopsy (Layer 1 + 2)

Prompt 15 - Buying signal classifier (Layer 3)

Section 4: JTBD & outcome prompts (prompts 16–20)

Prompt 16 - Outcome statement architect (Layer 2)

Prompt 17 - “Forced trade-off” pain finder (Layer 3)

Prompt 18 - Feature-to-JTBD translator (Layer 1 + 2)

Prompt 19 - JTBD interview script generator (Layer 2)

Prompt 20 - Outcome metric mapper (Layer 2 + 3)

Section 5: Champion mapping & org-chart prompts (prompts 21–25)

Prompt 21 - Champion profile builder (Layer 1 + 2)

Prompt 22 - Buying center radar (Layer 2 + 3)

Prompt 23 - Power-vs-interest grid (Layer 2)

Prompt 24 - Champion enablement brief (Layer 1)

Prompt 25 - Org-chart evolution tracker (Layer 3)

Section 6: Trigger-event & moment prompts (prompts 26–31)

Prompt 26 - Trigger event library builder (Layer 1 + 2)

Prompt 27 - “Switching cost moment” detector (Layer 2 + 3)

Prompt 28 - Funding-round trigger responder (Layer 1)

Prompt 29 - “Career-moment” trigger responder (Layer 1 + 2)

Prompt 30 - Disqualification trigger finder (Layer 3)

Prompt 31 - “Why now” diagnostic (Layer 2 + 3)

Comparison table: prompt categories vs. ICP layer vs. output

People Also Ask: 10 questions, answer-first

1. What are ChatGPT prompts for B2B ICP pain points?

2. How do you identify hidden pain points in B2B customers?

3. What’s the best framework for ICP pain-point research?

4. How is B2B buying behavior different in 2026?

5. Can ChatGPT really find ICP pain points from call transcripts?

6. How many customer calls do I need to find real ICP pain?

7. What’s the difference between BANT, MEDDIC, and SPICED?

8. How do I mine G2 reviews for ICP pain points?

9. How do I turn a Gong or Chorus transcript into an ICP insight?

10. How often should I refresh my ICP?

The 14-day ICP stress-test sprint

Common mistakes to avoid

Final word

SuperFresh AI

43 ChatGPT prompts for non-native English speakers to polish interview answers

41 ChatGPT prompts for SaaS founders in San Francisco to map local partnership opportunities

How to Detect AI-Generated Content

What Is the Best AI Tool for Writing?

AI Newsletter Writing Guide

Get practical AI insights in your inbox