How to Identify a Target Buyer's ICP From Their Company Website (A 15-Minute Drill)

Every sales manager says "do your research before the call." Almost nobody says what that means, so reps do research theater: ten open tabs, a skim of the About page, a glance at LinkedIn, and a first call that still opens with "so, tell me about your business." The prospect can smell it. Meanwhile the company's own website — the most information-dense, freely available document about how they make money — goes unread in any structured way. This post is the structure: a 15-minute drill that walks six sections of a prospect's website in a specific order, tells you what signal each one actually carries, runs the whole thing on a real company so you can see it work, and ends with the thing research is for — better discovery questions.

What you're actually looking for (it's two questions, not twenty)

Pre-call website research answers exactly two questions. Everything that doesn't feed one of them is trivia.

1. Do they fit our ICP? Your ideal customer profile is the set of firmographic and situational traits your best customers share — size, motion, market, growth stage. (If you haven't defined yours rigorously, start with our ICP scoring criteria guide — this drill assumes you know what "good fit" means for you.) The website either confirms the fit signals, contradicts them, or reveals the account was mis-segmented before you ever dialed.

2. Who is the buyer, and what do they wake up caring about? Not their job title — their scoreboard. The website tells you how the company makes money, what it brags about, and what it's investing in. The person you're calling is measured on some piece of that. Your job is to figure out which piece before they answer the phone.

Hold those two questions in your head and the website stops being a brochure and becomes a deposition: every page is the company testifying, under no pressure at all, about what it values.

The 15-minute drill

Set a timer. The cap is the point — research past fifteen minutes has sharply diminishing returns, and unbounded research is usually procrastination wearing a productivity costume. Six stops, in this order, because each one narrows what you look for in the next.

Minutes 0–2: the homepage hero

Read the headline, the subheadline, and the logos. Three signals:

• Who they say they serve. "For enterprise revenue teams" and "for freelancers who hate invoicing" are different companies with different budgets and different buyers, and they told you in eight words.
• The category they claim. The noun they use for themselves ("CRM," "revenue intelligence platform," "scheduling infrastructure") tells you who they think their competitors are — and what analyst category their budget comes from.
• Their words for their customer's pain. Write down the exact phrases. Mirroring a prospect's public language in your call ("you talk about eliminating scheduling friction — is that still the internal framing?") signals homework without you ever saying "I did my homework."

Minutes 2–4: the pricing page

The single highest-signal page on any B2B site, because pricing is strategy wearing numbers:

• Self-serve card checkout → product-led motion, decisions made low in the org, small initial deals, fast cycles.
• "Contact sales" on the top tier → they run a sales team, which means they have quota-carrying reps, sales leadership, and (if you sell anything touching revenue) a genuine buyer.
• The tier names and gates. A tier gated on SSO, audit logs, and SAML means they sell upmarket and know what enterprise buyers demand. Per-seat pricing means adoption breadth matters to them; usage pricing means consumption growth does. Whatever they charge on is what their own customers' finance teams scrutinize — and often what their internal metrics orbit.
• No pricing page at all → enterprise-only motion: long cycles, procurement, committees. Calibrate your expectations (and your deal math) accordingly.

Minutes 4–6: customers and case studies

Companies brag selectively, and the selection is the signal:

• Which segment stars in the case studies. If the homepage says "for everyone" but every case study is a 2,000-person logistics company, believe the case studies. That's who they monetize.
• Which metric the case studies celebrate. "Reduced onboarding time 40%" vs. "grew pipeline 3x" vs. "cut costs $2M" — that's the company's own value language, and it's very likely the language your buyer uses internally to justify purchases.
• Logo overlap with your customers. If their case-study logos look like your customer list, you've found a lookalike account — and possibly a shared reference you can name on the call.

Minutes 6–8: the careers page

The most underrated stop in the drill, because job listings are paid, current, and honest — nobody posts aspirational job ads for fun:

• What they're hiring tells you what they're investing in. Six SDR openings = scaling outbound (they feel pipeline pressure). A RevOps lead = systematizing the sales machine (tooling budget exists). Three ML engineers = product bet in flight.
• Job descriptions leak the tech stack. "Experience with Salesforce and Outreach required" just told you two tools they pay for — instant integration talking points, instant competitive intel.
• Hiring velocity is a budget proxy. Twenty open roles reads very differently from two, and a hiring freeze you can verify (an empty careers page at a company that was hiring last quarter) is a trigger event in itself.

Minutes 8–10: about, newsroom, and blog

Skim for trigger events and org shape, not history:

• Funding announcements — a raise in the last 12 months means budget and growth mandates; a Series B company is supposed to be buying tools.
• Leadership page — find your likely buyer's boss. Knowing your champion's VP is ex-Oracle tells you what buying culture they learned. New executive in your buyer's chain (last 6 months) is the single best trigger event in B2B: new leaders change vendors.
• What the blog writes about — companies write content about what they want to be known for next, which is usually a quarter ahead of what they're doing now.

Minutes 10–12: the product/solutions navigation

Open the "Solutions" dropdown. If it's organized by role ("for Sales," "for Marketing," "for Finance"), the company sells into departmental budgets and has already mapped the org chart you're about to navigate. If it's organized by industry, verticals drive their strategy, and the vertical page matching your prospect's flagship customers tells you where they're doubling down. If it's organized by company size, they've formalized segments — and you can read exactly where they draw their own SMB/mid-market/enterprise lines.

Minutes 12–15: synthesize onto one card

Stop browsing. Write five lines — if you can't, you weren't extracting, you were skimming:

1. Fit verdict: ICP yes / no / partial, with the one signal that decided it.
2. Likely buyer: role, and the scoreboard they're measured on.
3. Their money engine: one sentence on how this company grows.
4. Two hypotheses: informed guesses about their pain that your discovery call will test.
5. Three questions: discovery questions this research earned you (more on these below).

The drill, run for real: Calendly through a seller's eyes

Abstract frameworks are cheap, so here's the drill run out loud. Pretend we sell AI-powered sales coaching software, and Calendly's sales org just surfaced in our territory. Fifteen minutes on calendly.com (as of this writing — run it fresh yourself; sites drift):

• Hero (0–2): Scheduling automation, positioned well beyond "a link for meetings" — the messaging leans toward teams and revenue use cases, not just individuals. Category language: scheduling platform. So they've moved upmarket from their PLG roots — which means they've had to build a sales motion on top of a self-serve product, one of the hardest GTM transitions there is. Note the phrase; it's a pain hypothesis already forming.
• Pricing (2–4): Free and standard self-serve tiers, then team tiers, then an enterprise tier with the classic gates — SSO/SAML, admin controls, "talk to sales." Per-seat pricing. Translation: a real sales team exists, it sells seats into large organizations, and reps there run discovery and demos — the exact people a sales-coaching tool trains. Fit signal: strong.
• Case studies (4–6): Customer stories skew toward sales and recruiting teams at recognizable mid-market and enterprise names, celebrating metrics like faster speed-to-lead and more meetings booked. Their value language is conversion velocity. Our buyer almost certainly reports on pipeline metrics in that same dialect.
• Careers (6–8): Look for open AE, SDR, and sales-enablement roles. Every quota-carrying opening is literally a seat our product would coach; an enablement opening is better still — it means someone's job is about to be "make the reps better," and our tool is their first line item. Job posts mentioning Salesforce or Gong confirm the stack we'd integrate with.
• About/news (8–10): Established growth-stage company, sales leadership on the leadership page (a CRO or VP Sales — that's the economic buyer's chain). Any recent revenue-leadership hire would be our trigger event: new sales leaders audit their team's call quality in their first quarter.
• Solutions nav (10–12): Solutions organized by team/use case — sales prominently among them. They map buyers by department, our buyer's department included. They think in exactly the org-chart terms we'll navigate.
• The card (12–15): Fit: yes — seat-based sales org selling upmarket. Buyer: VP Sales / Head of Enablement, measured on ramp time and quota attainment. Engine: converting a self-serve base into enterprise contracts, which demands consultative reps. Hypotheses: (1) reps who grew up closing inbound PLG signups now have to run enterprise discovery, and the skill gap shows; (2) coaching is bottlenecked on managers reviewing call recordings by hand. Questions: below.

Fifteen minutes, no LinkedIn, no data vendor — and we walk into the call with a thesis instead of a blank page.

Turning signals into discovery questions

Research that doesn't change what you ask was entertainment. The translation rule: never present the signal as a conclusion; present it as a question the prospect gets to correct. You're not showing off the homework — you're using it to skip the useless opening lap of discovery.

Website signal	Weak use (sounds like a stalker's book report)	Strong use (a question it earned)
Enterprise tier appeared on pricing page	"I see you moved upmarket."	"When a self-serve customer graduates to the enterprise tier, who runs that sales conversation — and how were they trained on it?"
Six SDR job openings	"Looks like you're hiring a lot of SDRs."	"With the outbound team growing that fast, how are you keeping ramp time down while the managers' review load doubles?"
Case studies celebrate speed-to-lead	"Great case studies!"	"Your customer stories all speak in conversion velocity — is that the metric your own team gets measured on internally too?"
New VP Sales announced	"Congrats on the new VP."	"New sales leadership usually means the playbook's getting rewritten — what's changing about how your reps run calls this year?"

Notice each strong version does three things at once: it proves the research silently, it tests a hypothesis from your card, and it opens a pain thread a discovery call can pull. If you need the broader question bank to slot these into, the 20 discovery-call questions post pairs with this drill — and if you're selling to the revenue leader the drill usually points at, the VP of Sales persona page covers what their scoreboard looks like from the inside.

Never present a website signal as a conclusion — present it as a question the prospect gets to correct. Research isn't for showing off. It's for skipping the lap of discovery that everyone else wastes the first ten minutes on.

Common questions about identifying ICP from a website

How do you identify a company's ICP from their website? Read the pricing page, case studies, and solutions navigation — the three places a company reveals who it actually monetizes. The pricing structure shows the sales motion and deal size, the case studies show the segment and metric they brag about, and the solutions nav shows how they map their own buyers. Together those define the ICP more honestly than any About page.

What's the fastest signal on a B2B website? The pricing page. Self-serve vs. "contact sales," what the top tier gates (SSO, audit logs), and what the pricing unit is (seats, usage, revenue) tell you the motion, the buyer's altitude, and what their finance team scrutinizes — in under two minutes.

How long should pre-call research take? About fifteen minutes for a standard discovery call. Less and you're guessing; much more and you're procrastinating — the marginal signal past fifteen minutes rarely changes what you'll ask. Enterprise pursuits with committees justify deeper account research, but the first-call drill stays short.

Should I mention my research on the call? Never as a conclusion, always as a question. "I noticed X — how is that playing out?" proves the homework and hands the prospect the floor. Reciting facts about their own company at them proves nothing except that you can read.

What if the website contradicts my ICP data? Trust the website — it's more current than most data vendors. If your enrichment tool says 200 employees but the careers page shows 40 open roles across three new offices, the company outgrew the database. Re-segment the account before the call, not after.

You've read their website. Now practice the call it earned you.

SalesArmor closes the loop on this drill: paste the prospect's website (or their LinkedIn profile) and it builds the buyer — role, company, temperament, the objections someone in their seat would raise — then puts you on a live voice call with them. Test your two hypotheses and your three questions out loud before the real buyer picks up, and get scored on discovery depth when you hang up.

Turn a website into a practice call →

A note on sources

This drill distills the practitioner literature on pre-call account research: Aaron Ross's account-research discipline in Predictable Revenue, Mark Roberge's ICP-construction work, and the pre-call preparation frameworks published by SDR practitioners and the major sales-intelligence vendors (Clearbit, 6sense, Cognism, Apollo) on reading firmographic and intent signals from company web presence. The section-by-section signal readings — pricing as strategy, careers pages as investment disclosures, case-study selection as monetization evidence — are the standard tools of competitive and account intelligence, arranged here into a timed sequence a rep can actually run before a call. The Calendly walkthrough reflects their public website at the time of writing and is illustrative; run the drill fresh on any account before you trust it.