SalesAccount Intelligence

AI Prompts for Account Research

Account research is the difference between outreach that references something true and outreach that could have been sent to anyone. The discipline is not volume of information but relevance: what is this company trying to do, what is in the way, and why would our offer matter to that — with each claim traced to a source and each inference marked as such.

These prompts build the brief, form pain hypotheses, and audit the result for the assumptions that would embarrass you on a call. Give the model the material you have gathered — filings, job posts, press, earnings calls, the website — and treat anything it adds as a hypothesis to verify.

Before you use these

Have these ready to replace the highlighted [variables]:

Sources you have gathered: annual report or investor materials, recent press, job postings, leadership changes, product announcements, the company website
Your product's value drivers and the problems it solves
What you already know from prior contact, if any

The prompts

1. Build the account brief
2. Form pain hypotheses tied to their priorities
3. Audit the research for unverified assumptions
4. Run a ten-minute pre-call research sprint

1. Build the account brief

Best forA one-page brief that a rep can read in five minutes before any contact.

Inputs needed

Gathered sources
Product context

How to use itPaste the source material. Require a source reference next to every fact; the brief is only useful if the rep can trust it.

Expected outputStructured brief: company snapshot, stated strategy and priorities, recent events, organization and leaders, technology and vendors, and open questions — each fact sourced.

Act as a sales research analyst preparing an account brief on [company].

Sources: [paste or summarize: investor materials, press, job postings, leadership announcements, product news, website copy, prior notes]
Our product: [what it does, value drivers, typical buyer]

Build the brief:
1. Snapshot: what they do, size, segments, geographies, ownership, financial trajectory — one paragraph, each fact with a source reference.
2. Stated strategy and priorities: what leadership has said they are trying to achieve, in their words where possible, with source and date.
3. Recent events (last 12 months): funding, acquisitions, leadership changes, launches, restructuring, regulatory events — dated and sourced.
4. Organization: relevant functions, named leaders, recent hires or open roles that indicate investment or pain.
5. Technology and vendors: known systems, integrations, competitors' products in use — with the evidence quality (job post mention vs confirmed).
6. Open questions: what we would need to learn on a first call.

Rules: mark every statement as sourced (with the reference) or inferred. Do not add facts from memory about this company; if you know something not in the sources, list it separately as 'unverified background' rather than in the brief.

2. Form pain hypotheses tied to their priorities

Best forTurning research into two or three specific, testable reasons this account might need what you sell.

Inputs needed

Account brief
Product value drivers

How to use itAsk for hypotheses linked to their stated priorities, not to your product features. Each should be phrased as something you could ask about, not assert.

Expected outputRanked hypotheses with the evidence chain, the question that tests each, the value driver it maps to, and the trap to avoid.

You are forming pain hypotheses for [company] ahead of outreach.

Account brief: [paste]
Our value drivers: [problems we solve and the outcomes customers get]

1. Generate 4–6 hypotheses of the form: 'Because [stated priority or event], they probably face [specific problem], which shows up as [observable consequence].' Each must cite the brief evidence it rests on.
2. Map each hypothesis to the value driver it connects to. Discard any hypothesis that exists only because we have a feature for it.
3. Rank by evidence strength and by value to them if true.
4. For the top three: the discovery question that tests it without leading, the answer that would confirm it, and the answer that would kill it.
5. Traps: assumptions in the hypotheses that a knowledgeable insider would reject, and the alternative explanation for the evidence.
6. Outreach implication: which hypothesis is safe to reference in a first message (well-evidenced, public) and which should wait for a conversation.

Phrase hypotheses as questions to explore, not claims to make. Do not reuse the same hypothesis for every account by changing the company name.

3. Audit the research for unverified assumptions

Best forCatching the confident-sounding claims that would fall apart on a call.

Inputs needed

Brief and hypotheses
Source list

How to use itRun before the first call or message. The model rates each claim by evidence and flags what needs verification or softer phrasing.

Expected outputClaim-by-claim audit with evidence rating, the risk if wrong, the verification step, and rewritten phrasing for weak claims.

Act as a skeptical sales manager auditing an account brief before it is used.

Brief and hypotheses: [paste]
Sources available: [list]

1. Extract every factual claim and every inference. For each: evidence rating — verified (primary source), reported (secondary source), inferred (our reasoning), assumed (no basis given).
2. For each inferred or assumed item: what happens on a call if it is wrong, and how to phrase it as a question instead of a statement.
3. Stale information: anything dated more than [n] months that may have changed (leadership, strategy, vendors), and how to check.
4. Contradictions between sources, and which to trust.
5. The three claims most likely to be used in outreach and the specific verification step for each before sending.
6. A rewritten version of any hypothesis that currently reads as an assertion.

Be specific about what is unverified. Do not soften the audit because the brief reads well.

4. Run a ten-minute pre-call research sprint

Best forThe minimum useful research before a call that is starting soon.

Inputs needed

Company name and the person's role
Ten minutes
What you already know

How to use itTell the model what you have. It sets the sprint: what to look at in what order, what to extract, and the two hypotheses to go in with.

Expected outputOrdered checklist with time per step, the facts to capture, the two hypotheses, and the one question to open with.

Act as a sales coach designing a ten-minute research sprint before a call with [role] at [company] in [n] minutes.

What I already know: [notes]
Our product: [one line]

1. Ordered sprint: source → what to look for → what to capture, with minutes per step (e.g. 2 min company site 'about' and latest news; 2 min the person's profile and recent posts; 2 min job postings in the relevant function; 2 min last earnings/press for priorities; 2 min our CRM for prior contact). Adapt to what I already know.
2. The capture template: five fields to fill (their stated priority, a recent event, the person's focus, a likely pain, prior contact).
3. Two hypotheses to go in with, each phrased as a question.
4. The opening question for the call, built from the strongest fact found.
5. What not to do in ten minutes (deep financials, competitor analysis) — save for after the call if it progresses.

Keep it executable under time pressure.

Related prompts

Logical next step

After this, most sales teams move on to Buyer and Buying Committee Mapping.

B2B Sales Prospecting System ($39) — goes further: a repeatable outbound engine: research → sequence → objections → follow-up. See what's inside →

Get the free B2B Sales AI Starter Kit → Nine of these prompts as a target → engage → qualify workflow with an intake worksheet, delivered by email. See what's inside

All Sales prompts · Search the full library