What makes a B2B lead worth contacting?
A lead is not a row in a spreadsheet. For outbound, a useful lead has four parts: a company that fits your ideal customer profile, a person there who owns the problem you solve, a verified way to reach them, and ideally a reason to reach out now.
Most lists have the first three and miss the fourth. That is why the same list can produce very different results depending on when and how it is used. The sources below differ mainly in how well they cover each of the four parts.
Where can you find B2B leads? The main sources compared
Each source has a job. Use one for coverage, one for timing, and keep the trust-based ones (partners, inbound, referrals) running in the background.
| Source | Examples | Pros | Cons |
|---|---|---|---|
| Contact databases | Apollo, ZoomInfo and similar | Wide coverage; fast filters by industry, size and title | No timing; the same contacts are sold to every competitor; records age as people change jobs |
| Startup and industry directories | Product Hunt, app marketplaces, association member lists | Public and niche; often show what a company just shipped | Little contact data; quality varies a lot by directory |
| Online communities | Reddit, X, Slack and Discord groups, niche forums | People describe their problems in their own words, as they happen | Many accounts are anonymous; open promotion gets you banned |
| Search results | Queries such as "Shopify agency Chicago" or "HR software for clinics" | Finds small companies that databases cover poorly | Slow by hand; every result needs to be classified |
| Events | Conferences, meetups, webinars, speaker and sponsor lists | Shared context and a natural reason to talk | Seasonal, costly to attend, attendee lists often restricted |
| Partners | Agencies, consultants and integration partners serving your ICP | High trust and warm introductions | Slow to build; you have to give value back |
| Inbound | Demo requests, signups, content downloads, site visitors | The highest intent of any source | You do not control the volume; slow to build from zero |
Which public sources show who needs you now, and what rules apply?
Some of the most useful B2B lead sources are public and free, and they surface young and small companies that paid databases often miss. Each comes with rules, set by law, the platform or the community, that decide how you may use what you find.
| Source | What it tells you | Rules and limits |
|---|---|---|
| SEC EDGAR | Form D notices from companies raising money under Regulation D, generally filed within 15 days after the first sale, with the amount and the executive officers | Public record. Not every round produces a Form D, and the filing is a notice, not an announcement; the SEC explains it on its What is Form D? page |
| State business registries | Newly formed LLCs and corporations, with filing date and registered agent | Free search on each Secretary of State website. Many filings show only a registered agent, so you still need the website and founder; the SBA registration guide explains what gets filed |
| USAspending.gov | Federal contract and grant awards by recipient, with amounts and dates | Official public data. An award means a new project, often with hiring and subcontracting |
| Y Combinator startup directory and other accelerator portfolios | Young companies with a description, founders and batch date | Public pages. Each batch is a dated list of companies choosing their first vendors at the same time |
| Product Hunt | Daily launches with maker profiles and comments | Documents an official API. Launch comments are for feedback, not pitches |
| Tool requests, complaints and questions in topic communities | Software access is governed by the Reddit Data API Terms. Each subreddit sets its own self-promotion rules; never pitch in threads or private messages | |
| X | Founder posts, launch threads, hiring and funding news | Access and use are set by the X Developer Policy |
| Career pages and job boards | Open roles, which reveal budget and priorities | Widely indexed by search engines, often through JobPosting structured data |
| GitHub | Open-source projects and the companies behind them | Its Acceptable Use Policies forbid using site information for spam, including unsolicited email. Research only; never harvest addresses |
| Current titles and team structure | Its User Agreement prohibits scraping and automated access. Confirm roles by hand |
How do you search for B2B leads by hand?
A search engine is the cheapest lead database there is, if you know how to ask. Four operators do most of the work: quotes for an exact phrase, site: to stay on one domain, a minus sign to exclude a word or site, and OR to accept either of two terms. Four example searches, each for a different seller:
- A RevOps consultant looking for companies building a sales team: site:jobs.lever.co "revenue operations" "HubSpot". Many startups host their job boards on Lever, so each result is a company with an open RevOps role and a known CRM. Check the posting date before you add it.
- An agency that builds wholesale portals for Shopify brands: "powered by Shopify" "wholesale inquiries" -jobs. The results are stores that handle wholesale through a contact form today, which is the manual process the agency replaces. More angles in lead generation for ecommerce agencies.
- An SEO agency that serves law firms in one city: "personal injury lawyer" Phoenix -site:yelp.com -site:avvo.com. Skip page one and look at pages two and three: firms that exist but do not rank are the ones with a reason to talk. See lead generation for SEO agencies.
- A bookkeeping firm that works with agencies: "recommend a bookkeeper" agency site:reddit.com. Each result is a public question from someone with the problem. Read the subreddit rules before answering, and email only when the person's company is public.
What is signal-based prospecting?
Signal-based prospecting starts from events instead of filters. You watch for public moments that suggest a company needs something right now, then check whether it fits your profile, instead of the other way round. Signals do not replace the profile; they sort it. A company that fits and just showed a signal goes to the top. One that showed a signal but does not fit is skipped, however tempting the post.
Startories automates this part: it watches Reddit, X, Product Hunt, directories and search results for your niche, matches each signal to a real company, checks it against your ICP and keeps a link to the original post. See how it works in AI lead generation, or read the buying signals guide for more examples. The signals that tend to matter most:
- A request for a tool recommendation: the person is choosing now and the shortlist is still open.
- A complaint about a competitor: there is pain, an existing budget line and a reason to switch.
- A product launch: a team that needs users, feedback and often outside help.
- A job post for a specific role: budget has been approved for the problem that role handles.
- A tech stack change: neighboring tools and services get reconsidered at the same time.
- A new company: banking, payroll, software and marketing are all being chosen for the first time.
What should every lead record contain?
Write down more than a name and an email. The extra fields take seconds when you capture the lead and save hours later, when you need to explain a score, write a first line or answer someone who asks where you got their address. The example below is fictional.
| Field | Example | Why you need it |
|---|---|---|
| Company and domain | Northwind Clinics, northwindclinics.com | The domain is the key you deduplicate on across sources |
| Fit reason | Nine locations in Texas, online booking, no in-house IT | Shows why the account is in your ICP, so you can check the rule later |
| Signal and link | Job post for a practice operations manager, with its URL | Gives the email its first line and lets anyone check the fact |
| Signal date | September 30 | Signals age fast; the date lets you sort by freshness |
| Contact and role | Dana Ruiz, chief operating officer | The person who owns the problem, not just the most senior title |
| Email and verification date | Verified on October 6 | Tells you when to re-verify before the next send |
| Source | Company careers page | Answers "where did you find me?" and meets EU transparency rules |
| Region | United States, Texas | Decides which rules apply: CAN-SPAM, GDPR, state privacy laws |
| Status | New, contacted, replied or suppressed | Stops two people or two campaigns from writing to the same person |
Reach companies with a reason to buy this week
Startories finds the buying signal, verifies the decision-maker and runs the outreach until they book a call.
How do you verify B2B leads before contacting them?
Verification protects your sending domains and your time. A bad address bounces, and enough bounces push your emails into spam for everyone else on the list.
Check the company is real and active
A live website, a current product or service, and recent activity such as posts, releases or job openings. Directories are full of companies that closed quietly.
Check the person still works there and owns the problem
Compare the role with a recent source such as the team page or a recent post. Titles vary: at a 20-person company, the "head of growth" may be the founder.
Verify the email address
Run a verification step that checks the domain's mail server and the mailbox without sending a message. Treat "catch-all" results as risky.
Apply your suppression list
Remove anyone who unsubscribed, bounced or said no in any earlier campaign.
Re-verify close to the send date
A list verified three months ago is not verified today. Run the check again right before each batch goes out.
How many leads do you need? A worked example
Start from the outcome and work backwards. The numbers below are assumptions for illustration, not benchmarks: replace them with your own after the first few hundred sends.
| Step | Assumption | Result |
|---|---|---|
| Meetings needed | Goal set by the team | 10 per month |
| Positive replies needed | 1 in 2 positive replies books a meeting | 20 positive replies |
| Contacts to reach | 2% of contacts reply positively | 1,000 contacts |
| Contacts to source | 75% of found contacts pass verification | About 1,330 contacts |
| Companies to find | 1.3 contacts per company on average | About 1,025 companies |
| Emails to send | 3 emails per contact on average, as some reply early | About 3,000 emails a month |
| Inboxes needed | 30 emails per inbox per day, 21 sending days | 5 inboxes |
What does a week of prospecting look like? A worked example
A fictional week, with assumptions you can replace. A three-person agency sets up HubSpot for US B2B SaaS companies with 20 to 100 employees. One person spends about an hour a day on prospecting and aims to reach around 35 researched contacts a week.
Monday: build the account list
Pull 120 companies that match the profile from a database export and an accelerator directory. Remove anyone already in the CRM, in a sequence or suppressed. About 100 remain.
Tuesday: look for timing
Check each company for a signal from the last 30 days: a marketing operations job post, a recent Form D, a launch, a public complaint about its CRM. In this example, 22 have one and become the priority group.
Wednesday: find and verify people
For those 22, find whoever owns the CRM (head of marketing, RevOps lead or founder) and verify the email; assume 18 pass. Do the same for 22 accounts without a signal as a comparison group; assume 17 pass.
Thursday: write and send
Open the signal group's emails with the event and the comparison group's with fit alone. Send from a warmed-up secondary domain, within the daily cap.
Friday: log and review
Log replies by group. After four weeks, about 140 contacts, compare positive replies per contact. If the signal group wins clearly, move Tuesday's hour to more signal sources and spend less on database pulls.
Which lead sources fit your type of business?
The right mix depends on who you sell to and how they buy. Start with the closest match below, then test one new source per month. Whatever the source, follow the rules for using contact data: in the US, commercial email must meet the CAN-SPAM Act; in the EU, prospecting needs a lawful basis such as legitimate interest, and people can object at any time. Respect each platform's terms and each community's rules on promotion as well.
- SaaS selling to startups: launch sites, funding news and communities give you timing that databases cannot. See lead generation for SaaS.
- Agencies: search results and directories find the small and mid-sized businesses that databases cover poorly, and launches or hiring posts show who needs help now. See lead generation for marketing agencies.
- Consultants and high-ticket services: partners and referrals carry the most trust; outbound opens the doors partners cannot. See lead generation for consultants.
- Products with a strong community around the category: discussion threads where people ask for help are the richest source. See Reddit lead generation.
Which prospecting mistakes waste the most time?
- Filtering by title alone. A "VP of Marketing" at a 15-person company may not exist; the buyer is the founder. Filter by company first, then pick the role that owns the problem at that size.
- Exporting more than you can contact. Leads that sit unused for two months go stale: people change jobs and signals expire. Pull what you can reach in the next two weeks.
- Writing to role addresses. Inboxes such as info@ or sales@ are read by whoever is on duty, and verification tools often flag them as risky. Find the person who owns the problem.
- Not deduplicating across sources. The same company appears in a directory, a database and a public thread. Without a shared key, usually the domain, two sequences can reach the same person in the same week.
- Counting leads instead of conversations. A list of 5,000 is not progress. Judge every source by positive replies and booked meetings.
How to start finding leads this week
Pick one source for coverage and one for timing. Pull 200 companies that fit your profile, flag those with a recent signal, verify the contacts and reach the flagged ones first. After two weeks, compare positive replies between the two groups and move your effort to the one that wins.
Startories does the sourcing, matching and verification and includes the sending, so you do not need a separate database or email finder. The Starter plan covers one ICP, one intent source and about 250 qualified, enriched leads a month. See plans and pricing.
Frequently asked questions
What is the best way to find B2B leads?
Combine a narrow ideal customer profile with one source for coverage, such as a database or niche search, and one source for timing, such as communities, launches, hiring or funding news. Verify every contact and reach the companies with a recent signal first.
Are free B2B lead sources any good?
Yes, especially for timing. Communities, launch sites, job boards and search results are free and often more current than paid databases. They take more work, because you still need to identify the company, find the decision-maker and verify the email.
How many leads do I need each month?
Work backwards from meetings. With example assumptions of a 2% positive reply rate and half of positive replies booking a call, 10 meetings need about 1,000 verified contacts. Replace those assumptions with your own numbers after a few hundred sends.
Should I buy a lead list?
Be careful. Bought lists are often out of date, shared with many other senders and carry no timing, which raises bounces and spam reports. If you buy one, verify every address before sending and test a small batch first.
What is signal-based prospecting?
It means choosing who to contact based on a recent public event, such as a tool request, a complaint about a competitor, a launch, a job post or a funding round, then checking that the company fits your profile before you reach out.
Is it legal to email B2B leads found in public sources?
In the US, CAN-SPAM allows commercial email without prior consent if you follow its rules: honest headers and subject, a postal address and a working opt-out. In the EU you need a lawful basis such as legitimate interest and must say where you got the data. Platform terms apply on top.
Sources
- SEC: EDGAR full-text search
- SEC: What is Form D?
- SBA: Register your business
- USAspending.gov
- Y Combinator: Startup directory
- Product Hunt
- Product Hunt API 2.0 documentation
- Reddit: Data API Terms
- X Developer Policy
- Google Search Central: Job posting structured data
- EUR-Lex: General Data Protection Regulation (EU) 2016/679
- California Attorney General: CCPA
- FTC: CAN-SPAM Act compliance guide for business
- European Commission: What does "legitimate interest" mean?