Before cold emailing or calling a company, confirm what the business does, whether it fits your offer, what has changed recently, what observable problem or opportunity connects to what you sell, which sources support that conclusion, who is likely to own the issue, and whether an appropriate contact route exists. Stop when you have enough evidence to decide whether to proceed, hold or reject the account.
Before you open the browser, define what would make the company relevant
The most common research mistake happens before research starts: the seller has not defined what they are looking for. If your offer is “AI automation,” almost any business can look interesting. If your offer is “automate missed-call follow-up and lead-response workflows for multi-location service businesses,” you can look for much more concrete evidence.
Write three lines first:
- Who can we realistically serve? Geography, industry, size, business model and operational constraints.
- What observable conditions make our offer relevant? Examples: phone-only booking, repeated response-time complaints, multiple locations, high-volume inquiry forms, hiring for manual coordination roles.
- What would disqualify the account? Examples: recently replaced the relevant system, outside your service area, no appropriate business contact route, or a business model you do not support.
This turns research from browsing into a test.
Research sources in an order that protects you from weak assumptions
A practical order is to start with sources closest to the company, then widen out:
| Source | Use it to learn | Watch for |
|---|---|---|
| Company website | Offer, locations, audience, positioning, booking/contact process, leadership, careers | Marketing language is first-party but still promotional. |
| Company newsroom / official social | Launches, expansion, leadership, events, priorities | Announcement date and whether the change is actually relevant. |
| Official filings / registries | Entity identity, formal disclosures, public-company information | Only use sources applicable to the company and jurisdiction. |
| Roles, tenure, job changes, company headcount/hiring context, recent activity | Titles can be ambiguous; do not assume authority from title alone. | |
| Independent news / industry sources | External context, acquisitions, funding, market changes | Check date and original source. |
| Reviews / forums | Repeated customer-experience themes and hypotheses | A review is an individual's report, not proof of an internal system problem. |
HubSpot's current prospect-research guidance recommends understanding both the company and the person before first contact. LinkedIn's Sales Navigator Account Pages similarly consolidate company size, people, news, headcount and senior-management changes to help sellers determine relevance. The exact tool is optional; the research questions are not.
A step-by-step company research process before outreach
- Confirm you have the right company.
Match the website, location, trading name and basic activity. This sounds trivial until two companies share a name, a directory links to an old website, or a branch is mistaken for headquarters.
- Read the homepage for the business model, not the slogan.
Identify what the company actually sells, who appears to buy it, whether it is B2B or B2C, how customers engage, and where the company operates. Capture a one-sentence description in your own words.
- Check the pages closest to your offer.
If you sell conversion work, inspect booking, forms, pricing and contact flows. If you sell recruiting, inspect careers and team pages. If you sell cybersecurity, look for technology/security disclosures and job roles. Avoid a generic checklist that ignores the offer.
- Look for change.
Newsrooms, social posts, leadership pages and hiring activity can reveal whether the company is expanding, restructuring or investing. LinkedIn currently highlights leadership changes, hiring spikes, funding and expansion as account-prioritization signals. A change is a reason to investigate, not proof of demand.
- Find one observable business condition connected to your service.
Write it as a fact first: “The website lists four locations but each contact page uses the same general phone number.” Then separately write what it may mean: “Centralized phone handling may create routing or after-hours response pressure.”
- Find at least one supporting or contradicting source.
If one review complains about response time, look for a pattern. If a careers page suggests rapid hiring, check whether the roles are current. If an article says the company raised money, find the company announcement or reputable original coverage.
- Identify the likely function that owns the issue.
Do this before searching for a name. A marketing-site problem may belong to marketing, growth or the owner. Scheduling may belong to operations. Security may belong to IT/security. The organizational context determines the person.
- Check the contact route and channel.
Is there a professional business email, company contact path, telephone number, LinkedIn route or other appropriate channel? Keep legal and platform requirements separate from research quality.
- Decide what the first message needs to validate.
Do not present an outside inference as fact. The outreach angle should connect an observed signal to a question or relevant value proposition.
Signals that deserve a second look
HubSpot defines buying signals as actions and data points that can indicate readiness for a sales conversation, and its current Sales Hub documentation draws from intent, recent news and engagement. A small business or consultant may not have access to enterprise intent feeds, but many useful public signals are available.
- Hiring: a growing function can indicate investment, capacity pressure or strategic importance.
- New leadership: a new executive or functional head may be reviewing priorities and vendors.
- Expansion: new locations, markets or service lines can create operational and marketing needs.
- Product or pricing change: can create new go-to-market, enablement or customer-experience requirements.
- Technology change: migration, new integrations or changed customer flows can make adjacent services relevant.
- Repeated public friction: patterns across recent customer reviews can suggest an experience issue worth validating.
- Content and leadership activity: what senior people publicly discuss can reveal priorities, but should be interpreted carefully.
How to use reviews and job postings without overclaiming
Reviews
Customer reviews are useful for identifying repeated themes: slow callbacks, confusing booking, inconsistent service, shipping delays, poor handoffs. They are weak evidence for diagnosing internal causes. Five reviews saying “nobody called me back” can support a hypothesis that response handling deserves investigation; they do not prove which software or employee caused the problem.
Job postings
Job postings can show where a company is investing and what capabilities it values. A posting for three implementation specialists suggests growth in delivery capacity; a marketing-operations role may reveal the tools and processes that team owns. Again, avoid turning a posting into a buying claim. Hiring says “this function matters,” not “they want our service.”
Find the likely problem owner after the company makes sense
Once the company has passed the fit and evidence tests, find the person. LinkedIn's current Sales Navigator guidance recommends titles and seniority filters to target decision-makers, while its larger-account guidance warns that buying groups may include several stakeholders.
For a small business, the owner may be appropriate. For a 50-person agency, a Head of Growth, Operations Director or Managing Director may own different problems. For a large enterprise, the operational owner, technical evaluator, procurement stakeholder and economic buyer may all differ.
Record why the role is relevant. “VP” is not reasoning. “Owns revenue operations; company is hiring SDRs; our offer concerns lead-routing and handoff” is.
Turn company research into a short pre-outreach brief
A good brief should be readable in under a minute. Include:
- Company and one-sentence description
- Fit summary
- Observed opportunity or problem
- Source URLs and dates
- What is inferred rather than observed
- Likely owner and role reasoning
- Contact route
- Why now
- Unknowns
- First-conversation question
The brief is not a script. Its job is to give the seller enough context to sound informed and enough uncertainty to ask a real question.
Fast research versus deep research
| Mode | Use when | Research depth | Output |
|---|---|---|---|
| Fast triage | Lower-value or high-volume accounts | Entity, fit, one relevant signal, likely role, contactability | Proceed / hold / reject |
| Standard research | Normal B2B service prospecting | Company, problem, 2+ sources, role mapping, timing | Opportunity brief |
| Deep research | High-value strategic accounts | Multiple stakeholders, filings/news, initiatives, technology, contradictions, account history | Account plan + stakeholder map |
Do not measure quality by time spent. The correct depth is the cheapest amount of research that reduces enough uncertainty for the next decision.
Research quality does not replace outreach compliance
If your research leads to cold email, telephone outreach or other direct marketing, follow the laws that apply to the recipient, channel and jurisdiction. Rules differ. For example, the UK Information Commissioner's Office notes that B2B marketing rules under PECR vary by method and by the type of business being contacted; corporate subscribers and sole traders can be treated differently. That is one reason a global “B2B cold email is always allowed” rule is unsafe.
Noustiq's editorial advice is therefore simple: research public business information responsibly, use appropriate professional channels, maintain do-not-contact preferences, and obtain jurisdiction-specific legal advice where needed.
Worked example: researching a multi-location dental practice for an automation consultant
Suppose you sell lead-response and booking automation to dental practices.
| Research step | What you observe | What you can reasonably infer |
|---|---|---|
| Fit | Four locations, new-patient services, centralized contact number. | Multi-location coordination may matter. |
| Customer journey | Website asks visitors to call for some appointment types. | Phone handling is part of conversion flow. |
| Reviews | Several recent reviews mention difficulty getting callbacks. | Response experience may deserve validation. |
| Hiring | Open front-desk roles at two locations. | The practice may be adding service capacity; no assumption about cause. |
| Owner | Website lists an Operations Manager; LinkedIn shows current tenure. | Operations is a plausible first role for a workflow conversation. |
A weak cold message would say, “Your office is missing leads and needs automation.” The research does not prove that. A better opening would be closer to: “I noticed some appointment types still route through the phone and you are operating across four locations. We help multi-location practices improve what happens after an inquiry arrives. Is lead-response consistency something operations is currently working on?”
The difference is not just personalization. The second version preserves the boundary between what you observed and what you still need to learn.
Turn research into a reusable brief
Noustiq Pro keeps the company, evidence, likely decision-maker, unknowns and first-conversation context together before outreach.
See the product workflowCommon mistakes
- Reading only the homepage. The useful evidence is often in careers, contact flows, locations, reviews or leadership changes.
- Researching forever. Use a stop rule tied to the decision.
- Copying AI summaries without checking sources. Keep source URLs for important findings.
- Writing outreach while researching. Decide whether the company is worth contact before optimizing copy.
- Assuming titles equal authority. Map the problem to the function and role.
- Ignoring contradictory evidence. A new website may invalidate your “outdated website” pitch even if an old directory screenshot suggests otherwise.
How to take research notes that remain useful tomorrow
Research notes often fail because they preserve conclusions but not the path to them. A line such as “good prospect, growing fast” is hard to audit a week later. A better note keeps four parts: observation, source, interpretation and implication.
| Field | Example | Why it matters |
|---|---|---|
| Observation | Careers page lists two customer-service coordinators and a scheduling lead. | States what is visible without interpretation. |
| Source | Careers URL, checked 26 Aug 2026. | Lets another person verify freshness. |
| Interpretation | The company may be investing in inquiry handling and scheduling capacity. | Labels the researcher's judgement. |
| Implication | Worth checking whether lead-response workflow is an operations priority. | Connects research to the next sales question. |
This structure is especially useful when research is handed to another person. It prevents the salesperson from seeing a confident conclusion with no way to tell how strong the evidence was.
Actively search for contradictions before you contact
Confirmation bias is expensive in prospecting. Once a researcher likes an account, every new fact can start to look supportive. Add a deliberate contradiction step:
- If you think the website is outdated, check recent launch announcements or the site footer/change history.
- If you think the business is expanding, verify that the new locations or jobs are current.
- If reviews suggest poor response time, check whether the company recently introduced a new booking or support system.
- If a role appears to own the problem, look for an org/team page that may show responsibility sits elsewhere.
The objective is not to disprove every opportunity. It is to prevent a weak sales story from becoming stronger merely because nobody looked for counter-evidence.
Research differently for email, phone and LinkedIn
The core company research stays the same, but the output should match the channel.
Before a cold email
You usually need a concise reason for relevance, a current role, a professional contact route and one or two evidence points that can support a short message. Avoid turning the email into a research report.
Before a cold call
Prepare a sharper verbal hypothesis and two or three questions. The caller needs enough context to explain why the call is relevant, then shift quickly into discovery. Research should reduce awkwardness, not eliminate conversation.
Before LinkedIn outreach
Role context, professional activity and shared context can matter more. Keep the message tied to the person's work rather than personal details. LinkedIn's own Sales Navigator guidance emphasizes recent posts, job changes and common connections as useful professional context.
Use a research stop rule
Research can expand indefinitely because the internet always has another page. A stop rule protects time. Stop when one of these conditions is true:
- Reject: a non-negotiable fit condition fails.
- Proceed: fit is clear, one material reason is supported, likely ownership is defensible and an appropriate contact route exists.
- Hold: the company fits but one specific unknown blocks a confident next step.
For a strategic account, you may intentionally go deeper after “proceed” to map stakeholders and initiatives. For a $1,000 service prospect, another hour of research may be irrational. Research depth should reflect expected account value.
Freshness matters because companies change
A prospect record is not permanent truth. Websites are replaced, people move, hiring closes, locations open and priorities change. Record the date you checked each important source and consider a recheck policy for held accounts. LinkedIn's emphasis on job-change and account alerts exists for the same operational reason: account context is dynamic.
A practical freshness rule is not “everything older than 90 days is invalid.” Different facts decay at different speeds. A legal entity name may be stable; a job title or open role can change within weeks. Mark recheck urgency according to the type of evidence.
A company-research worksheet that forces clarity
If you research manually, use a structured worksheet rather than free-form notes:
Company:
Website:
What they sell:
Who they serve:
Why they fit us:
Observed condition:
Source URL:
Source date / checked date:
What it may mean:
Contradicting evidence checked:
Likely owner function:
Likely role / person:
Contact route:
Why now:
Unknowns:
Proceed / Hold / Reject:
First question to validate:The format is intentionally short. A completed worksheet should tell another person why the account deserves attention without requiring them to repeat the research.
How to convert research into outreach without sounding invasive
Use business-relevant observations that the recipient would reasonably recognize as public. Avoid demonstrating that you collected every available detail about them.
A useful structure is:
- Observation: a current company fact.
- Relevance: why that fact connects to the area you work in.
- Value: what kind of problem you help solve.
- Question: a low-assumption way to validate whether the issue exists.
For example: “I noticed you have added two locations and your appointment flow still routes several services through a central phone number. We work on lead-response workflows for multi-location practices. Is call routing and after-hours follow-up something your operations team is reviewing?”
The observation is visible. The interpretation is modest. The question lets the buyer correct you.
When public records add real value
Do not use official registries simply because they look authoritative. Use them when they answer a material question. SEC filings can help with public-company strategy, risk, business segments, executives and formal disclosures. A local company registry may help confirm legal identity. These sources are especially useful when the account is valuable enough to justify deeper work or when first-party marketing copy is ambiguous.
For most small local businesses, the company website, professional profiles, reviews and current business listings will be more practical than a regulatory filing. Source sophistication should follow the question, not become a badge of research quality.
Adjust the source mix to the type of company
A local service business and a public software company leave different evidence trails. For a local business, the useful sources may be the company website, booking flow, Google Business Profile, public reviews, local licensing information and LinkedIn. For a public company, investor relations, SEC filings, earnings material, executive appointments and detailed careers pages can add much more context.
Do not force enterprise research methods onto a ten-person company, and do not rely on a homepage alone for a high-value strategic account. The research question determines the source depth.
Likewise, a sole consultant may have little formal company information but a rich professional profile and public content history. Confirm that the person actually operates the business before assuming an individual profile represents the company you intend to approach.
Sources and further reading
These sources were reviewed on 26 Aug 2026. They support the external facts and platform-specific guidance referenced in this article. Noustiq's frameworks, examples and decision rules are editorial synthesis, not claims made by these sources.
- HubSpot: A 3-Step Guide to Performing Prospect Research
- LinkedIn Sales Navigator: How to Use Sales Navigator
- LinkedIn Sales Navigator: How to Prioritize the Right Accounts
- HubSpot: Use Buying Signals in the Sales Workspace
- LinkedIn Sales Navigator: How to Find and Target the Right Buyers Within Accounts
- U.S. SEC: Search Filings / EDGAR
- UK ICO: Business-to-business marketing guidance