Discover Account ResearchAESDR/BDRSales LeadershipFounder

Turn a Company URL Into a First-Call Brief

Turn a company URL into a sourced, opinionated first-call brief that maps your offer to your buyer's real pressures and earns the next meeting.

StageDiscover
Best forAE, SDR/BDR, Sales Leadership, Founder
Works withClaude, ChatGPT, Gemini, Perplexity
When to use it

When to use it

Run this in the 10 minutes before a discovery call or a high-effort outbound touch, when you have a URL but not the time to read their site, their last earnings note, three exec LinkedIn profiles, and a job board. It earns its keep when the call is worth winning: a named target account, a senior buyer, a deal you'd be annoyed to lose to under-preparation. It is the difference between 'I looked at your website' and 'I noticed you just opened a second fulfillment center and are hiring three dispatchers, how is route planning holding up?'

Do NOT use it as a substitute for thinking, and do not paste its output into an email as-is. The brief is a hypothesis generator, not a fact sheet. Every 'priority' it produces is a guess until the buyer confirms it on the call, which is exactly why the SMART QUESTIONS section matters more than the SNAPSHOT. Skip this prompt entirely for low-stakes, high-volume outbound where 90 seconds per account is 89 seconds too many, there, a templated trigger is fine.

The principle: relevance compounds. A buyer gives you attention in proportion to how clearly you understand their world before you ask for anything. This prompt front-loads that understanding and, critically, separates what's verified from what's inferred, so you sound informed without bluffing facts you can't defend when they push back.

The prompt

The prompt

Prompt, paste into Claude, ChatGPT, Gemini, Perplexity
You are a senior B2B sales researcher preparing a colleague for a high-stakes first call. You are precise, skeptical, and allergic to filler. You never invent facts, funding, headcount, revenue, customer logos, or quotes. You clearly separate what is verifiable from what is reasoned inference.

CONTEXT
Company to research: {{COMPANY_URL}}
What we sell (product + the outcome it drives): {{WHAT_WE_SELL}}
Who I'm meeting (name / title / anything I know): {{CONTACT}}
My goal for this call: {{CALL_GOAL}}
Our sharpest proof in their world (a relevant customer, result, or insight, leave blank if none): {{RELEVANT_PROOF}}

TASK
Build a first-call brief that helps me sound like I understand their business and ask questions worth a senior person's time.

METHOD (follow in order)
1. If you have live web access, read their homepage, product/pricing pages, an about/customers page, recent news, and the careers page. Cite the source domain inline in brackets, e.g. [pricing page], [careers]. If you do not have web access, say so in one line at the top and work only from what I pasted or what is broadly known, do not guess specifics.
2. Build a SNAPSHOT of what they actually do and who they sell to. Resist generic descriptions; be concrete.
3. Reason from the contact's ROLE to the 2-4 business pressures someone in that seat is likely measured on this year. Tie each to an observable signal where possible.
4. Map our offer honestly against those pressures. If the fit looks weak, say so plainly.
5. Surface only recent, referenceable signals (hiring, launches, leadership change, funding, expansion, site changes). No signal is better than a stale or invented one.
6. Write 5 discovery questions that are specific to THIS account and role, open-ended, non-leading, and impossible to ask of a random company.

OUTPUT FORMAT (use these exact headers)
SOURCING NOTE: one line on whether you used live web access.
1. SNAPSHOT (3 bullets): what they do, who they sell to, a size signal.
2. LIKELY PRIORITIES (3 bullets): pressures on {{CONTACT}}'s role this year; tag each (verified) or (inferred).
3. WHERE WE FIT (2-3 bullets): how {{WHAT_WE_SELL}} maps to those priorities; be honest about weak fit.
4. RECENT SIGNALS (up to 4 bullets, with sources): skip the section if you found nothing credible.
5. SMART QUESTIONS (5): tailored, open-ended, sequenced from context-setting to impact.
6. OPENER (1 sentence): a non-cheesy line that shows homework without flattery.
7. RISKS / WATCH-OUTS (2 bullets): what could make this a bad-fit or stalled deal.

CONSTRAINTS
- Whole brief under 350 words.
- Tag every non-verifiable claim (inferred). Never state funding, revenue, headcount, or customer names as fact unless you sourced them.
- No hype language. No padding. If a section has nothing real, write 'Nothing credible found' rather than filling space.
- If a critical input is missing (e.g., I gave no contact role), ask me one clarifying question before producing the brief.
Run it from the terminal

Run it from the terminal

zsh
$# research an account with a browsing-enabled model via Perplexity
$ls prompts/
account-brief.md cold-email.md icp.md sequence.md
$# fill the variables, then pipe the prompt to the llm CLI
$cat prompts/account-brief.md | llm -m sonar-pro
SOURCING NOTE: Used live web access (homepage, about, careers, blog). 1. SNAPSHOT - Regional last-mile delivery firm serving Midwest retail/grocery [about]. - Sells fleet delivery to mid-market retailers; case studies skew grocery [customers]. - ~20 open ops/dispatch roles, a 200-500 employee signal (inferred). 2. LIKELY PRIORITIES - Protecting margin as fuel and labor costs rise (inferred). - Absorbing a new grocery contract without SLA slippage (verified, blog). - Modernizing a manual dispatch process (inferred from job posts). 3. WHERE WE FIT - Route optimization hits Dana's two largest controllable costs, strong fit. - Weak spot: if they run a TMS with routing, it's rip-and-replace, qualify early. 5. SMART QUESTIONS - Walk me through how routes get planned today, from order intake to driver phone. - With the new grocery volume, where does that process start to crack? 6. OPENER "I saw you brought on the grocery contract and are staffing up dispatch..."
$# 82 words under budget; signals are current because browsing was on
$echo $?
0
$
Variables to swap

Variables to swap

VariableWhat to put there
{{COMPANY_URL}}The prospect's website URL or company name
{{WHAT_WE_SELL}}One line on your product/service and the concrete outcome it drives
{{CONTACT}}Name and/or title of the person you're meeting (role matters most)
{{CALL_GOAL}}What a good outcome from this call looks like (e.g. qualify fit and book a technical deep-dive)
{{RELEVANT_PROOF}}Your strongest proof point in their industry, a fictitious-friendly customer, a result, or a relevant insight (optional)
Worked example

Worked example

Input

Filled-in example
COMPANY_URL: northwind-logistics.com
WHAT_WE_SELL: AI-driven route optimization that cuts last-mile fuel spend and driver hours
CONTACT: Dana Ruiz, VP Operations
CALL_GOAL: qualify fit and book a 30-min ops deep-dive
RELEVANT_PROOF: a similar regional carrier trimmed driver hours ~12% in a quarter

Output

What the model returns
SOURCING NOTE: Used live web access (homepage, about, careers, blog).

1. SNAPSHOT
- Regional last-mile delivery firm serving retail and grocery clients across the Midwest [about].
- Sells fleet-based delivery to mid-market retailers; case studies skew grocery and convenience [customers].
- Careers page lists ~20 open roles weighted toward ops/dispatch, a ~200-500 employee signal (inferred).

2. LIKELY PRIORITIES
- Protecting margin as fuel and labor costs rise (inferred from sector + hiring pattern).
- Absorbing a new grocery contract without SLA slippage (verified, announced on blog).
- Modernizing what looks like a manual dispatch process (inferred from job descriptions referencing 'manual route building').

3. WHERE WE FIT
- Route optimization targets fuel + driver-hours, Dana's two largest controllable cost lines, strong fit.
- If dispatch is manual today, time-to-value is fast and easy to demonstrate live.
- Weak spot: if they already run a TMS with routing, we're a rip-and-replace conversation, not an easy add, qualify early.

4. RECENT SIGNALS
- New grocery client announced last quarter [blog], fresh volume, fresh routing pain.
- Hiring multiple dispatch + ops roles [careers], scaling under strain.

5. SMART QUESTIONS
- Walk me through how routes get planned today, from order intake to a driver's phone.
- With the new grocery volume coming on, where does that process start to crack?
- How do you measure a route-planning miss right now, fuel, overtime, SLA penalties?
- What would a 1% improvement in fuel or driver hours be worth to you annually?
- Who besides you would weigh in on a change to how routing works?

6. OPENER
"I saw you brought on the new grocery contract and are staffing up dispatch, I wanted to understand how route planning is holding up under the new volume before I show you anything."

7. RISKS / WATCH-OUTS
- May already own a TMS with native routing, confirm before positioning as net-new.
- VP Ops may not control budget; identify the economic buyer on this call.
Tips

Tips to get more out of it

  • Run it in a browsing-enabled model (Perplexity, or ChatGPT/Gemini/Claude with web on) so signals are current; if you can't browse, paste the homepage, pricing, an about page, and the careers page yourself and the brief stays sharp.
  • Treat the RISKS section as the most valuable output, it names the deal-killer your optimism wants to ignore. If the model returns soft risks, push it: 'What would make this a waste of my time?'
  • Everything tagged (inferred) is a question, not a fact. Never assert it on the call; ask it. Asserting an inferred headcount and being wrong torches your credibility in the first 60 seconds.
  • Fill RELEVANT_PROOF with a customer from THEIR industry, a smaller logo in their world beats a Fortune 500 logo in another. Leave it blank rather than letting the model invent one.
  • The OPENER is a draft, not a script. Rewrite it in your own cadence so it doesn't sound templated; the goal is for it to sound like a thought you had, not a line you generated.
  • If you're running this across many accounts, ask for the output as a compact table and feed a list, but accept shallower depth as the trade.
  • Cross-check any sourced number the model gives you against the actual page. Models occasionally 'source' a plausible figure that isn't on the cited page.

Get new prompts in your inbox

The best GTM prompts and AI use cases, every couple of weeks.

Subscribe →