AI prompts for sales and marketing that don't read like spam

Generate 12 landing page headline candidates.

Product: product
Audience: audience
The single most painful problem this audience has: pain
The unique mechanism we use to solve it: mechanism
Banned words (overused in this space): banned words

Headline rules:
- Maximum 9 words.
- No words from the banned list.
- No "Unlock", "Empower", "Transform", "Revolutionize", "Cutting-edge", "Game-changer", "Seamless".
- At least 4 should be questions or statements written in the audience's own voice.
- Mix: pain-led, outcome-led, mechanism-led, contrarian.

For each headline, label its style. After the 12, pick your top 3 and explain why.

Write a cold outreach message.

To: person name, their role at their company
From me: my role at my company
What I want from them: specific ask
One specific thing I know about them (not from their bio): specific signal

Rules:
- ≤ 60 words total.
- First line must reference the specific_signal. No generic compliments.
- No "I came across your profile", "I hope this finds you well", "saw your post".
- The ask must be the last line, phrased as one question.
- No links unless the message would be useless without one.

Output the message only.

Help me draft a positioning statement using April Dunford's framework.

Product: product
What it does in one sentence: what it does
Closest alternative the customer compares us to: alternative
Customers we serve best: best customer
Customers we serve poorly (be honest): worst customer

Produce:
1. **The positioning statement** (1–2 sentences) in the form:
 "For [best_customer], [product] is the [category] that [unique_value], unlike [alternative]."
2. **Three reasons to believe** — the proof points that back the unique_value claim.
3. **The 'who this is not for' list** — the worst_customer described in their own language so prospects can self-select out.
4. **One belief shift** the prospect must make to choose us over the alternative.

A prospect raised this objection: "objection".

Context:
- Product: product
- Where in the sale this came up: sale stage
- What they actually want (your best guess at the underlying concern): real concern

Produce:
1. **Name the real concern in one sentence** — what they actually mean.
2. **Validate it briefly** — one line that shows you heard it (no over-apologizing).
3. **The reframe** — one sentence that re-anchors the conversation on a different axis where you win.
4. **The proof** — one specific data point, case study reference, or guarantee.
5. **The next-step question** — one question that moves the deal forward without pressuring.

Output as a single 4–6 sentence reply that flows naturally. Then list the structural beats you used.

Below is raw text from source type (interview notes, reviews, support tickets, etc).

Extract:
1. **Pain phrases** — the literal words customers use to describe what hurts. 5–10 verbatim quotes.
2. **Desire phrases** — the words they use to describe what they want instead. 5–10 verbatim quotes.
3. **Banned words** — language THEY do not use that we should never use either (jargon, marketing-speak we caught ourselves using).
4. **One outsider-vocabulary insight** — a word or phrase they use that our category doesn't.

Rules:
- Quote them verbatim. Do not paraphrase.
- If you cannot find 5 in any category, say so.

Source:
"""
paste feedback here
"""

Build a discovery call plan for this prospect.

What we sell: product
What they said when they got in touch: inbound context
What I know about their company: company context
What I need to establish to qualify them: qualification criteria

Produce:
1. The hypothesis: what I think their real problem is, and how confident I am
2. Opening question that gets them describing their current process, without mentioning our product
3. Questions that establish: what they do today, what it costs them, what they have already tried, why it failed, and what happens if they do nothing
4. For each, the follow-up when the answer is vague — the one that gets a number or an example
5. The question that would disqualify them, and my willingness to ask it early
6. What I must not do on this call: the points where I would be tempted to start pitching, and what to say instead
7. What a good call ends with — a specific next step. "Send some information" is not one

Rules:
- No question answerable with yes or no, except the disqualifier.
- Do not ask about budget before establishing cost of the status quo. The order matters.

Write a case study from this customer interview.

Transcript:
"""
paste transcript here
"""

Who reads this and what they are deciding: audience decision
What we can and cannot claim publicly: approval constraints

Structure:
1. The situation before — in their words, including what was frustrating about it
2. What they tried first and why it did not work. If they mention a competitor, handle it factually.
3. Why they chose us — their stated reason, in their words
4. What actually changed, with their numbers. Attribute every number to the transcript.
5. What is still not solved, or what took longer than expected. Include this — it is what makes the rest believable.
6. A pull quote, verbatim

Rules:
- Every claim traces to something in the transcript. Where you need a claim that is not there, mark it [NEEDS CONFIRMATION]. Do not write it yourself.
- Do not smooth their language into marketing copy. If they said "a nightmare", do not write "challenging".
- List separately: what to check with them before publishing, and what needs legal or their comms approval.

Write subject line options for this email.

What the email says: email content
Who receives it and their relationship to us: audience
What I want them to do: desired action
Subject lines that have worked before: past winners
Subject lines that flopped: past losers

Produce eight options, each using a different mechanism. Label the mechanism:
- The specific detail — one concrete fact from the email
- The direct statement — what it is, no framing
- The question the reader already has
- The number
- The consequence of not opening
- The named thing — a person, product or place they recognize
- The unexpected juxtaposition
- The plainest possible description

For each: the mechanism, the line, and the audience it works worst for.

Then: which two to test against each other and why those two, given what has worked before.

Rules:
- Maximum char limit characters.
- No curiosity gaps that the email does not immediately pay off.
- No "quick question", no false "re:", no fake urgency, no emoji unless the past winners show they work here.

Draft a comparison page: us versus competitor.

What we do well, with evidence: our strengths
What they do better, honestly: their strengths
Who we are genuinely wrong for: bad fit
Who they are genuinely wrong for: their bad fit
Verified facts about their product, and where each came from: competitor facts
Date of verification: verified date

Produce:
1. A one-line answer to "which should I choose", conditional on the reader's situation
2. A comparison table of the dimensions that actually differ. Skip dimensions where both are equivalent — a table of ticks in every row destroys credibility.
3. Where they are better, stated plainly and without a spin sentence afterwards
4. Who should choose them. Be specific enough that a real buyer would self-identify.
5. Where we are better, with evidence. Adjectives do not count
6. What we could not verify, marked as such, with the date the rest was checked

Rules:
- Every factual claim about the competitor comes from the verified facts above. Do not infer their behavior or pricing.
- No claim about their roadmap or their intentions.
- If a claim would embarrass us in six months when they ship something, do not make it.

Analyze why this deal was lost.

What we were selling and to whom: deal context
Stated reason for the loss: stated reason
How the process went, stage by stage: process timeline
Who was involved on their side, and when each appeared: their stakeholders
Where things slowed down or went quiet: friction points
What we did in response: our actions

Work through:
1. The stated reason — is it plausible given the timeline, or is it the polite reason?
2. Where the deal was actually lost. Usually earlier than where it ended. Point to the stage and the evidence.
3. Which of these it was: wrong fit that we should have disqualified, a real product gap, a process failure on our side, a champion who could not carry it internally, or a decision that was already made before we arrived
4. What signal was present at the time that we could have read, and when
5. Whether this was winnable at all. If not, say so — the lesson is about qualification.
6. The one change that would matter across similar deals, beyond this one
7. What to ask the buyer in a loss review, phrased so they answer honestly

Do not conclude "price" unless the timeline supports it. Price is the most common polite reason.