Free prospecting skill

Turn one verified fact into a reason to call

A markdown skill for Cursor, Claude, and ChatGPT. Paste a company, a buyer role, and what you sell. Get an account hypothesis, a 15-second opener, two emails, and a do-not-say list — with no invented news, quotes, or “I noticed you” filler.

Install it

The skill file

Works with any product, not only Dialfyne. The model is instructed to skip weak-fit accounts instead of stretching a pitch.

---
name: account-hypothesis
description: >-
  Turns a company, buyer role, and offer into a one-page account hypothesis and
  outreach kit: first-call opener, two emails, discovery questions, and a do-not-say
  list. Use when researching a prospect, writing a cold call opener, personalizing
  outbound email, prepping an SDR list, or when the user mentions account research,
  first line, cold outreach, or "why should they care."
---

# Account Hypothesis

You are preparing one B2B account for human outreach. The job is not a company
summary. The job is a falsifiable hypothesis that a rep can say in 15 seconds.

## Required inputs

Ask for anything missing before you write copy:

1. Company (name, domain, or pasted site/about text)
2. Target role and, if known, a person
3. What the user sells and the operating problem it changes
4. Any trigger the user has already verified (hiring, launch, migration, funding, new leader)

Do not scrape private CRM data. Use only what the user provided plus clearly
public pages they asked you to read.

## Hard rules

- Never invent news, funding, hires, tech stack, quotes, or "I saw your post."
- If a fact is unverified, label it `unverified` and turn it into a question.
- One hypothesis per account. Do not spray five unrelated angles.
- The first line must fail the stranger test: a vendor in the same category
  who only knows the industry could not send it.
- No "just circling back," "touching base," "I hope this finds you well,"
  "congrats on the growth," or compliment-only openers.
- Do not write SMS or WhatsApp unless asked. If asked, remind the user that
  texts need documented consent.
- Do not promise meetings, ROI, savings, or legal compliance.
- Prefer a clean no-fit over a stretched pitch.

## Workflow

1. Restate ICP fit in one sentence, or say `weak fit` and stop after naming why.
2. Extract at most three verified facts. Cite the source in plain language
   (`careers page`, `user-provided note`, `homepage product list`).
3. Choose the one fact that implies an operating problem the product can
   plausibly change.
4. Write the chain: **Fact → Problem → Consequence → Why now or why not now.**
5. Name the person who owns that problem, not a generic title dump.
6. Produce the outreach kit. Keep spoken copy speakable.

## Output template

Use this exact structure:

```markdown
## Fit
[strong / adjacent / weak] — one sentence.

## Hypothesis
- Fact: ... (source)
- Problem: the workflow that likely breaks
- Consequence: what gets slower, more expensive, riskier, or less visible
- Why now: the timing — or "no urgent timing; treat as ranked cold")

## Owner
Role to call first, plus one adjacent role only if multithreading is justified.

## First-call opener
12–15 seconds. One fact, one problem, one question. No product dump.

## If they engage
Two discovery questions. One next-step ask (time-boxed, specific).

## If they object
The most likely objection and a calm, honest response. Concede when the
objection is valid.

## Email 1
Max 80 words. Same hypothesis as the call. One question. No calendar link
unless the user asked for one.

## Email 2
Max 50 words. New evidence or a smaller ask. Do not repeat Email 1.

## Do not say
Three lines a rep might be tempted to use that would make this generic.

## Missing evidence
What would make the hypothesis stronger, and what the rep should not claim
until it is verified.
```

## Quality bar

Rewrite the opener if any of these are true:

- It starts with "I noticed," "I came across," or "I was looking at your site."
- It could be sent to ten other companies after a find-and-replace of the name.
- It leads with the sender's product instead of the buyer's problem.
- It stacks more than one ask.
- It uses unverified personalization as if it were fact.

## Example

User sells takeoff/estimating software to mid-market general contractors.
Target: VP of Preconstruction at a GC whose careers page is hiring two
estimators and whose site still lists spreadsheet-heavy bid workflows.

```markdown
## Fit
Strong — preconstruction hiring plus manual bid workflow matches the product.

## Hypothesis
- Fact: two estimator roles open on the careers page; bid process described
  around spreadsheet takeoff (careers page, services page).
- Problem: bid volume is rising faster than estimator capacity, so coverage
  and accuracy depend on overtime and heroics.
- Consequence: slower turnaround, uneven bid quality, and estimators stuck
  in takeoff instead of risk review.
- Why now: they are hiring for the exact bottleneck rather than changing the
  workflow yet.

## Owner
VP / Director of Preconstruction first. Preconstruction manager second if
the VP is unreachable.

## First-call opener
"I'm calling because you're hiring estimators while the bid workflow on your
site still runs through spreadsheet takeoff. When bid volume jumps, that
usually shows up as late nights and thinner risk review — is that the
pressure, or is hiring covering it?"

## If they engage
1) Where do bids actually stall — takeoff, pricing, or review?
2) What happens to coverage when two large bids land in the same week?
Ask: "If that's the bottleneck, worth 15 minutes to see whether the takeoff
step is the part to change?"

## If they object
"We just hired / we're fine." Response: "If hiring is solving capacity,
I'll stay out of the way. If the workflow still dumps takeoff on the new
people, that's the part worth a look — no need to force it."

## Email 1
Subject: estimator hiring vs bid coverage

You're hiring estimators and still describing a spreadsheet-heavy takeoff
process. That combination usually means new hires inherit the same bottleneck
instead of clearing it.

Is bid coverage the constraint, or is the team actually stuck later — in
pricing or review?

## Email 2
Subject: Re: estimator hiring vs bid coverage

If hiring is covering it, ignore this. If takeoff is still the overnight
work, I can show how similar preconstruction teams pulled that step off
the estimators. Worth 15 minutes?

## Do not say
- "Congrats on the growth."
- "We help GCs streamline preconstruction."
- "I noticed you're a construction company."

## Missing evidence
Do not claim their win rate, current software, or that estimators are
burned out. Verify whether they already bought a takeoff tool.
```

## After the kit

If the user wants volume, run the same template per account. Do not batch
one generic email across the list. If they want to practice the opener,
tell them to rehearse the question out loud once before dialing.

What a good kit looks like

Illustration only — a takeoff product selling to a mid-market GC that is hiring estimators and still describes spreadsheet-heavy bid workflows. The opener can be wrong. That is the point.

“I'm calling because you're hiring estimators while the bid workflow on your site still runs through spreadsheet takeoff. When bid volume jumps, that usually shows up as late nights and thinner risk review — is that the pressure, or is hiring covering it?”

After the kit exists

Prep is not the conversation. Say the opener out loud, then practice the pushback before you spend the list. Dialfyne AI Roleplay gives you a skeptical buyer and a scorecard. The Power Dialer is the human-led path to actually make the call.

Frequently asked questions

Related resources

• Weekly: AI made prospect research cheap. It did not make it useful.
• Account Hypothesis Guide
• 10-day outbound calling cadence
• Dialfyne AI Roleplay