Most articles on this subject are a list of tool names. This one is a list of sources, ranked by the thing that actually decides whether a lead is worth having: how likely it is that the business exists, matches what you sell, and can be reached.

The ranking criterion throughout is freshness first. That is not an aesthetic preference. A stale record costs you a bounce, bounces are charged to your sending domain rather than to the list, and enough of them will stop your mail reaching the good records too. Volume that you cannot contact is not an asset.

The ranking

#SourceFreshnessContact qualityEffort
1Live map and directory dataVery highHighLow with tooling
2Your own website visitorsImmediateHighMedium
3Association and chamber directoriesHighMediumLow
4Review platformsHighMediumMedium
5Referrals from customersImmediateVery highLow, but limited
6LinkedIn and professional networksHighMediumMedium
7Local business news and awardsVery highLowMedium
8Permit, licence and registry recordsHighLowHigh
9Job boardsVery highLowMedium
10Generic scraped databasesLowLowLow
11Purchased listsVery lowVery lowLow

1. Live map and directory data

For local businesses this is the strongest cold source available, and the reason is structural: the businesses maintain these listings themselves. A restaurant that moves updates its map entry within days, because customers cannot find it otherwise. No purchased database has that incentive behind it.

You get name, category, address, phone, website and often opening hours, which is enough to identify the business and enough context to write something specific. From the website you can usually reach a real email address.

Searching manually works and does not scale. This is the part worth automating, and it is what live discovery in Leads Ranger does: query by business type and location, get back businesses that exist right now, enriched with contact details and verified before anything is sent. Our full guide to finding local business leads covers the manual version too, because the method matters more than the tool.

Best for: any business selling to local, physical, service-based companies.

2. Your own website visitors

The highest-intent source almost every business already has and almost nobody works properly.

Someone reading your pricing page has done something no cold prospect has done: chosen to look. Even without a form submission, the traffic itself tells you which industries and regions find you, which is the single best input into your cold targeting.

Do the easy version first. Add a low-commitment way to identify themselves that is not "book a demo": a checklist, a tool, a template. Then treat form-fills as the priority queue they are, and use the demographic pattern to sharpen everything else.

Best for: everyone, and it is the first thing to fix if it is unworked.

3. Industry association and chamber directories

Trade associations, chambers of commerce, professional bodies and franchise networks publish member directories. They are small, current, and quietly pre-qualified: membership costs money, so a listed business is one that invests in its own growth.

They are also naturally segmented. "Members of the regional restaurant association" is a tighter, more defensible target than "restaurants".

Best for: B2B services selling into a defined trade or profession.

4. Review platforms

Wherever your target market collects reviews. These are current for the same reason maps are, and they carry a signal no directory has: how the business is doing and what its customers complain about.

That signal is directly usable. A business with strong reviews mentioning slow response times has told you what to open your email with, and it did not require guessing.

Best for: anyone whose product addresses a problem customers complain about publicly.

5. Referrals from existing customers

Highest conversion of anything on this list and the reason it is not ranked first is that it does not scale on demand. You cannot decide to have more referrals this quarter.

Ask specifically. "Do you know anyone who might need this" produces nothing. "You mentioned your supplier struggles with the same scheduling problem, would you introduce me" produces a meeting.

Best for: every business, as a habit rather than a channel.

6. LinkedIn and professional networks

Strong for reaching a person in a role rather than a business at an address, which makes it complementary to map data rather than competitive with it. Profiles are current because people maintain them for their own reasons.

The constraints are real: connection limits are enforced against your account rather than your campaign, and volume is punished. The limits and how to work within them.

Best for: selling to a specific job title, especially in larger organisations.

7. Local business news and awards

Openings, expansions, funding, awards, new premises, new hires. Low contact quality, because you usually get a business name and nothing else, but very high timing quality: a business that just expanded has budget and a new set of problems, and almost nobody is contacting them about either.

Use this as a trigger layer on top of another source rather than as a list in its own right.

Best for: anyone whose product matters most at a moment of change.

8. Permit, licence and registry records

Public records of new business registrations, building permits, liquor licences, contractor licences. High effort, often awkward formats, and worth it for a narrow set of sellers: if you sell to construction, hospitality or regulated trades, this tells you about a business before it appears anywhere else.

Best for: suppliers to regulated or project-based industries.

9. Job boards

A company hiring is a company with budget and a stated problem. A job advert is an unusually explicit statement of what a business is struggling with, written by the business.

Contact quality is poor and inference is required, but the targeting signal is excellent, particularly if what you sell is an alternative to the hire.

Best for: services that substitute for or support a role being hired.

10. Generic scraped databases

Large aggregated B2B databases, sold by the record or by subscription.

They have their place: for larger-company B2B, firmographic filtering at scale is genuinely hard to replicate. The problem is age. Data is scraped, aggregated, resold and aggregated again, and by the time it reaches you nobody can say when any given row was true. Contact-level accuracy decays fastest of all, because people change jobs constantly.

If you use one, verify every address before sending, without exception, and treat the stated data quality figure as marketing.

Best for: mid-market and enterprise B2B, with verification mandatory.

11. Purchased lists

Do not.

Not primarily for legal reasons, though those are real in the EU, the UK and the Gulf. The practical reasons are worse:

  • You inherit someone else's bounce rate. These lists have usually been sold repeatedly and mailed heavily. High bounce rates damage the sending reputation of your domain, which then affects your good campaigns too.
  • You cannot answer where a record came from. That question is the foundation of every compliance story you might need to tell.
  • Spam traps. Recycled lists frequently contain addresses specifically designed to identify senders who use recycled lists. Hitting one is not a bounce, it is a blocklisting.

There is no version of this that is worth the saved afternoon.

The pass that decides everything

A source produces candidates. Qualification produces leads, and skipping it makes every source on this list perform like the bottom of it.

Verify contactability. Check the email exists before sending anything. Check the phone connects. This is not optional and it is where most of the value of good sourcing is preserved or thrown away. The bounce rate guide explains the arithmetic.

Check fit against something written down. Not "small businesses" but the specific description you can build a list from: industry, size, location, and at least one signal that they need what you sell.

Score and order. You cannot contact everyone today, and sending capacity is deliberately limited. Something has to decide the order, and any scoring framework beats alphabetical.

Deduplicate. The same business will appear on a map, in a directory and on a review site. Contacting it three times because it arrived from three sources is the fastest way to look like a machine.

The test for any new source

Before committing to a source, take fifty records and check them by hand:

  1. Does the business still exist and still do what the record says?
  2. Is there a working website?
  3. Is there an email address you can verify?
  4. Does the phone number connect?

Score the percentage that pass all four. Above roughly seventy percent, the source is worth building on. Below about fifty, you are paying for the privilege of damaging your sending reputation, and the correct response is to stop rather than to verify harder.

An hour spent on this test has saved more outbound programmes than any amount of copy improvement.

Where to go next