Prospect Research Checklist: 8 Steps Before You Reach Out

James· 2026-11-17T08:37:33
Prospect Research Checklist: 8 Steps Before You Reach Out

Use this 8-step prospect research checklist to assess fit, spot buying signals, and write outreach that cites a real, recent reason to reach out.

A rep spends forty minutes on a prospect's LinkedIn profile, skims the homepage, and sends a note that opens with a compliment about the company's mission. The reply never comes. Nothing in that note explains why this company, and why this week.

The time cost is larger than one ignored email. Salesforce's 2024 State of Sales research surveyed 5,500 sales professionals across 27 countries and found that sellers spend 70% of their time on non-selling activities, including administration and meeting preparation. Research that never changes the first sentence is just more of that non-selling work. The same research reported that 67% of sales representatives did not expect to meet quota in 2024, which is exactly when teams start spraying more mail instead of tightening the brief.

A prospect research checklist fixes the sequence. Confirm fit. Confirm the account deserves a second look. Find a verifiable reason to reach out now. Record the evidence so the opener can cite it. Each step answers one question about fit, timing, or proof, and you stop when the note is sendable.

I keep the finished brief to five lines: ICP grade, one confirmed signal, source, date, and a first sentence a stranger could check. If you cannot point at the source, you do not have a reason. That standard is what defensible relevance in SaaS cold email requires.

Table of Contents

1. Define the Ideal Customer Profile and Firmographic Filters

Fit comes before curiosity. An ICP is a short set of company rules, not a paragraph about your "best customers." Industry, employee or revenue band, geography, business model, and the buyer role you can reach. Every later signal is judged against those rules. A hiring spike at a company you would never close is still a no.

A 2026 survey of 222 B2B sales professionals found that respondents estimated 32% of their CRM contact and account records were inaccurate, incomplete, or outdated. Scoring against a fuzzy ICP on dirty fields just ranks noise. The same survey found 75% would prefer perfect prospect data to an advanced AI sales assistant. I agree with the priority. Clean filters beat another enrichment vendor.

A professional businesswoman looking at her laptop screen displaying data while sitting at her office desk.

Say a B2B SaaS startup only sells to North American companies with 50 to 500 employees and a finance team already in place. A 40-person studio and a 2,000-person bank both fail before anyone opens LinkedIn. A RevOps team can run the same test on inbound signups and route only high-fit accounts to sales.

Write the rules so a new rep can grade an account without a meeting.

  • Must-haves stay few: Three to five testable criteria. Nice-to-haves live in a separate column so they never block a real fit.
  • Store the grade on the company record: One label, used by sales, marketing, and success. "Kind of fits" is not a grade.
  • Rewrite from closed deals each quarter: Closed-won and closed-lost both count. If losses cluster in one industry, that industry leaves the list.

A written ICP is the gate. The step-by-step ideal customer profile guide is a useful build sequence if your current definition is still a slide.

2. Technographic Research for the Prospect's Technology Stack

Stack research answers a narrower question than fit. Does this company already run the tool you integrate with, replace, or sit beside?

Say your product migrates teams off a competitor CRM. A careers page that mentions "Salesforce cleanup" plus a public script for that CRM is a second look. A marketing vendor selling a native app should filter for the commerce platform the app plugs into, not "companies that sell online." Website scripts, DNS records, job posts, and public product pages are the usual evidence. Vendor databases help, and a short comparison of technographic data providers is worth reading before you trust one feed.

What holds up in the first line

Mention a specific tool only after you can point at it. "Teams like yours often use X" is a guess, and buyers hear the guess. I flag anything a provider asserts without a page, a posting, or a detectable script as unconfirmed in the account note. Unconfirmed stack data never becomes the opener.

Tools change. A detection from last quarter can be a migration you missed. Re-check active accounts every few weeks, and treat a single job-post keyword as a hint until a second source agrees.

Practical rule: Two independent traces, or the tool stays out of the email.

The trade-off is coverage versus confidence. Broad technographic feeds find more logos. They also invent logos. I would rather mail fewer accounts with a stack I can defend than open with a product the prospect ripped out in March.

3. Hiring Signal Analysis from Job Postings

One open role is a job ad. A cluster is a budget decision.

Count postings in the same function over 30 to 60 days. Three RevOps roles plus a CRM admin in one month says the company is rebuilding how pipeline data moves. That is a fair reason to offer migration help. A local agency that sees a multi-location service business hiring a reputation manager has a similar opening, aimed at review operations rather than software.

A professional woman reviewing a resume while working at her desk with a laptop computer nearby.

Read the required skills, not just the title. "HubSpot admin, lifecycle stages, lead scoring" tells you the stack and the pain. "Outbound SDR, 80 calls a day" tells you they are adding volume, not fixing process. Those are different emails.

What does not work is quoting the posting back at them, or treating every engineer hire as a buying moment. Reference the role category and the problem it implies. Leave the job-ad prose alone. Cross-check the cluster against funding or a leadership change before you call it urgent. A single backfill after someone quit is maintenance, not a trigger.

Hiring shows where headcount budget is going. It does not prove they will buy your category.

If the posting is older than the window you set, drop it. Stale reqs are how "why now" turns into "why last spring."

4. Funding and Financial Event Monitoring

A round, an acquisition, or a filed expansion is public, dated, and easy to overuse. Everyone with a Google alert sends the same congratulations note on announcement day. That note does not get a meeting.

Wait a few weeks. The company is still doing press, and your email lands in a pile of identical openers. Verify the event in the original release or filing, not a rewritten blog. Then connect the capital to a problem you can name. A SaaS company that announces a Series B and hires a head of sales in the same quarter is staffing a new motion. Your line is about that motion, with the round as context, not the subject.

An acquisition is a different problem. Two customer bases, two tools, one leader who has to pick. Offering help with the merge is specific. "Congrats on the deal, want a demo?" is not.

Limit alerts to accounts already inside the ICP. Funding news across the whole market is noise, and noise trains reps to ignore the feed. I only keep a financial event in the brief when three things are true: the source is primary, the date is inside your outreach window, and you can say what the money or the deal forces them to change. If you cannot say the change, save the account for a later signal.

5. Leadership Change Tracking

The first 90 days, used carefully

A mid-market company posts that a new VP of Customer Success starts Monday. If the account already matches your ICP, the hire is a reason to look again, not a reason to mail every promotion on LinkedIn. Confirm the company still fits. Then confirm the seat is worth a second look: a book of customers, some budget, and a plan that is still open. A title change with no new scope fails that test.

Source the move from the company page, the press note, or the leader's own post. The citable fact is the role and the start date. Leave out school, family, and gossip about the last employer.

Relevance is highest in the first 30 to 60 days. The plan is still blank, and the inbox is already full of "congrats on the new role." A narrow offer for the first quarter works better than a connection request. For that VP, the line can be a retention read on the book they just inherited. For a new CMO, a 90-day look at current campaigns. Many B2B buyers want to hear from vendors while they are still evaluating options. They do not want your company biography.

Before you write, check whether the last leader's tools are already being replaced. A rip-and-replace means you are late, or you are the replacement. If nothing has moved, you may still shape the shortlist. Mail on the title alone and you sound like every other rep. Wait for a cleaner story and the window closes. I keep the signal only when the source, the date, and a specific first-quarter problem are all in the note.

A finished note the next sender can cite:

Fit: mid-market, already inside the ICP.
Signal: VP of Customer Success, start date, link to the company post.
Why now: first 45 days, stack not yet replaced.
Opening: "You inherited the book this month. Here is one retention gap teams in that seat usually find in month one."

Log the change, the date, the source, and whether you already mailed. The next rep should not rediscover the same announcement and send a second congratulations.

6. Review and Reputation Analysis

A dental practice has watched its Google rating fall for two months. The account already fits. The remaining question is whether the reviews give you a line you can send without scolding the owner.

Public reviews are a buying signal when you read the trend, not the trophy score. For local service businesses, Google Business Profile movement is often the cleanest "why now." For B2B software, G2 or Capterra clusters about onboarding, support, or a missing integration show which competitor's users are frustrated, and which complaint you can answer without insulting the reader.

BrightLocal's 2025 survey of 1,026 representative U.S. adults found that only 4% never read online business reviews. Owners feel that attention. A two-month rating slide is a concrete problem for an agency. A software vendor that finds repeated G2 complaints about a competitor's onboarding can talk to those companies about migration, not about "your bad reviews."

A professional woman in a blazer reviewing financial documents and taking notes at her office desk.

Run the check in order: fit, a second look, a reason you can cite, then the evidence on the record. Track four fields, not a single average.

  • Rating trend: Direction over 60 to 90 days, not today's star count. A decline in that window is the second look.
  • Recency and velocity: New reviews, or a stall after a previously steady pace.
  • Complaint theme: The repeated issue, paraphrased in the reviewer's words. That theme is the reason to reach out now.
  • Response latency: Unanswered negatives suggest the owner is not watching.

Frame the email as help with that theme. "Reviews this quarter keep mentioning onboarding delays after go-live" cites a pattern a buyer can check. "Your reviews are bad" does not. Verify the source and the date before you cite anything, and put both on the record next to the draft line.

Review count alone is a weak qualifier. Expectations about "how many reviews is enough" keep shifting, and a fat profile can still be declining. Because that consumer survey is U.S.-based, use the review sites common in your country. Fit, change, a citable line, then the link on the record.

7. Website Intent Signals and Account-Level Visit Tracking

A watched account returns to your pricing page three times in a week, then opens the integration docs. The old checklist stopped once you had a logo and a title. Much of B2B evaluation now happens anonymously on the vendor's site before anyone agrees to a call. A page visit is still not a purchase order.

Treat website intent as four stages in one workflow, not as a score you paste into a sequence.

Confirm fit first. Limit the read to accounts you already watch that are looking at a problem you solve. A pricing-page surge can be an analyst, a student, a competitor, or one curious employee. Return visits to pricing, demo, or integration pages beat a third-party in-market score that names no page.

Then confirm the account is worth a second look. Pair the visits with one other sign of urgency: an open role, a funding note, a leadership change, or a review slide. Visits alone start a nurture step. They do not justify a cold call.

The reason to reach out has to be something you can record and the opener can cite. Repeated pricing visits from a watched account, plus an open RevOps req, become a short note offering a pricing walkthrough. Name the topic, not the click path. "Teams evaluating pricing for multi-product setups" is fair. "I saw you on our pricing page at 4:12 p.m." is not. It reads as surveillance, and the buyer cannot check it.

Privacy sits inside the research. The GDPR began applying on May 25, 2018, and requires personal data to be processed lawfully, fairly, and transparently. Names, work emails, titles, and inferred contact details are personal data even when you started from a public company page. Keep organization facts apart from person-level fields. For a practical boundary on company-level visits without a person graph, see how to identify web visitors at company level.

8. Source Verification and CRM Record Enrichment

This is the step teams skip, and it is the step that makes the other seven usable. A signal without a source and a date is a rumor. Rumors get pasted into sequences, bounce, and train the team to distrust the CRM.

A finished note looks like this, and nothing longer:

Why now: Rating down over 60 days on the Google Business Profile, unanswered negatives about wait times. Source: profile reviews, checked 12 March. ICP: fit, 80 employees, two locations. Opener: "Your recent reviews keep mentioning wait times at the second location. We help multi-site clinics answer those within a day." Unconfirmed: claimed "new EHR," no posting or page.

A 2025 survey of 602 CRM users and administrators in the United States, the United Kingdom, and Australia found that 76% said less than half of their organization's CRM data was accurate and complete. Respondents also reported losing an average of 16 sales deals per quarter because of bad data. That is what happens when enrichment is a one-time append.

Require a source and a date on every material fact. Mark unverifiable claims unconfirmed and ban them from outreach. When two providers disagree on employee count or stack, keep both until a page resolves it. Do not let the last write win. Set freshness by volatility. Hiring, leadership, funding, reviews, and site changes get checked often. Industry and headquarters can wait. Before any automated CRM write, set three thresholds: auto-update, human review, and hold.

Map fields before you turn on sync. Firmographics, ICP grade, and the "why now" note should land on properties reps already open. A weekly refresh on active accounts beats a heroic quarterly cleanup. CRM data enrichment only helps if the workflow preserves provenance. If you later add model-assisted briefs, keep the same rule. AI business research workflows still need a human disposition for anything the model cannot cite.

How each prospect research step differs in purpose, best use, and main limitation

Item Core Question It Answers Best Use Main Limitation
Define the Ideal Customer Profile (ICP) and Firmographic Filters Which companies and buyer roles fit the target customer profile? Focus prospecting and align sales, marketing, and RevOps Narrow or outdated criteria can exclude viable buyers; firmographics do not show buying readiness
Technographic Research: Identifying the Prospect's Technology Stack What software and platforms does the company use? Find integration fit, competitor users, or relevant replacement opportunities Data may be inaccurate or outdated and does not reveal user satisfaction
Hiring Signal Analysis from Job Postings What priorities or capabilities is the company investing in? Identify timely outreach opportunities based on hiring activity Evergreen or stale postings can misrepresent real plans
Funding and Financial Event Monitoring Has a financial event created a reason or budget to act now? Time outreach around verified funding, acquisitions, or other financial changes Funding may not lead to immediate spending, and private-company data can be incomplete
Leadership Change Tracking Has a new leader arrived who may reassess vendors or priorities? Reach decision-makers during their early planning period A new leader may retain existing vendors; overly personal outreach can feel invasive
Review and Reputation Analysis What customer pain points or reputation changes are visible? Offer solutions tied to recent review trends or competitor dissatisfaction Reviews may be unrepresentative, temporary, or risky to reference directly
Website Intent Signals and Account-Level Visit Tracking Are employees at a watched account showing interest in specific topics? Prioritize accounts with repeated visits to high-value pages Visits may not indicate buying intent, and privacy rules and data coverage vary
Source Verification and CRM Record Enrichment Are research claims verified, current, and usable in the CRM? Maintain trustworthy records and support confident, traceable outreach Manual checks take time, while automated enrichment can introduce errors

Turn the Checklist Into a Routine You Run Every Week

Run the eight checks in order for every account you plan to contact. Confirm fit against written firmographic rules. Confirm the stack only when you can point at a tool. Then look for hiring clusters, financial events, leadership changes, review movement, and account-level site intent. Verify what you found, date it, and write one opener from the strongest sourced signal. Drop the rest.

Order matters because each stage kills work. Accounts outside the ICP never reach a job board. Accounts with fit but no second signal wait. Accounts with a signal you cannot cite stay in nurture. You are not trying to know everything about the company. You are trying to earn one sentence the buyer can check.

What fails in practice is the campaign-week scramble. Reps research 40 accounts on Monday, send on Tuesday, and never look again. Hiring reqs close, ratings recover, and the "why now" in the sequence is already wrong. A weekly pass on a watched list, with stale notes flagged, beats a deep dive you cannot repeat. High-volatility fields move. Your brief should move with them.

Keep the CRM honest while you do it. One ICP grade, one why-now field, source, date, and an explicit unconfirmed flag. If two facts conflict, leave both until a primary page settles them. Reps will trust a short record that admits uncertainty faster than a long record that pretends to be complete.

I would not open with the clever signal. I would open with the one you can defend this week. A role cluster, a filing, a rating slide, or a return to pricing from an account you already qualified. Cited, current, and specific is what gets the reply. Generic mission compliments do not, no matter how long you stared at the profile.

CapyScout can watch the companies already in your CRM, check the live web for those buying moments, and attach the evidence to a daily queue of who to contact and why now. The research stays in the same order as this checklist. Fit first, then a sourced reason, then a line you are willing to send.


CapyScout finds fitting companies, scores inbound signups for fit and risk, enriches the CRM, and monitors hiring, funding, leadership, reviews, and account-level website intent with the source attached. That is the prospect research checklist run on a schedule instead of rebuilt before every sequence. Visit CapyScout and turn the next watched account into a sourced reason to reach out this week.

prospect research checklist sales prospecting account research buying signals outbound sales
Turn every inbound email into a decision
See how CapyScout scores signups, forms, and replies in seconds.
Start free

Not ready to start? Get a free target list: 10 companies to sell to, picked from your website.