How to Get Contacts for a Cold Email Campaign Safely

Contacts for cold email come from five places, and each one carries a different level of risk. From lowest risk to highest:

1. Public sources you search by hand: company websites, team and leadership pages, press releases, author bios and LinkedIn profiles.
2. B2B data platforms, where you filter by title, industry, country and company size.
3. LinkedIn Sales Navigator to find the people, plus an email finder for their addresses.
4. Scraping websites, then enriching the results with a second tool.
5. Bought lists.

Three rules apply to every route. Verify every address before you send to it. Put the people who decide on the list, by name and by the title they really hold. Make the first list big enough to give you an answer. A few hundred hand-picked names rarely can, unless your whole market is that small.

When a founder tells me nobody replies to their cold emails, I ask for the spreadsheet before I read a word of the email. Sharp copy sent to the wrong people from a shaky domain still earns silence. The list often breaks the campaign more than the copy does.

Size matters as much as quality. You need quantity AND quality. A campaign that runs for months eats new rows every week, and someone has to source and check each one before it goes out.

Which contact source is safest, fastest or cheapest?

By risk I mean addresses that bounce, people who cannot buy, and data you cannot trace back to a source.

Free public sources rank first for trust. You can see the person's real title on the company's own pages, so you know why every row exists. The cost is time, and the gaps are real: many companies publish no addresses, and small firms often have no team page.

Paid platforms trade a bit of that certainty for speed. Fair deal, as long as you check what they hand you. My team builds lists from Sales Navigator combined with data platforms such as Apollo, UpLead and ZoomInfo, and that mix beats a scraped file. Platform data still goes stale and still carries wrong titles, so it goes through the same checks as everything else.

Scraping is quick to run and slow to clean. You get addresses with no context, many of them generic, and, in the UK and EU, a duty to tell people where you found them. Bought lists have all of those problems at once, plus a seller you cannot question.

What the paid routes cost to start, at the time of writing, from each vendor's own pages:

Add a verifier on top of any of these (prices in the verification section). Scraping has no published price; its cost is the cleanup time. Manual sourcing costs nothing but hours, which is why the test near the end of this article times it.

If LinkedIn will run next to email, source so that the person you connect with on LinkedIn is the same person you email. Then the two channels back each other up instead of reaching two different crowds. Take that person's title from the deals you have already won, meaning whoever really signed off. A title dropdown or a bare "manager" label will mislead you.

When both channels run in our setups, the LinkedIn touch comes first and the warm email follows it.

Use these decision rules to pick a route:

• If your whole market is a few hundred named companies, use routes 1 and 3 and skip bulk tools.
• If you need thousands of contacts in a few countries and one or two industries, start with route 2, then check titles by hand on a sample.
• If LinkedIn will run alongside email, use route 3, so both touches reach the same person.
• If a seller or tool cannot tell you where a contact came from, do not send to that contact.
• If a list costs far less than what the platforms charge per contact, treat it as route 5, whatever the seller calls it.

Whatever you pick, audit the lead list row by row before anything goes live.

What should you search on company websites and LinkedIn?

Find the person first, then the address. Most weak lists are built the other way around.

Start with the deals you have already won. For each one, note who had the final say on the purchase. Then search for the titles that niche really uses. Leave out generic titles such as Assistant, Manager and Associate on purpose.

Take a supplier of site safety equipment selling to UK contractors. The titles to keep are health and safety manager, procurement manager, contracts manager and project director. A specific title that happens to contain "manager" stays. A bare "Manager" goes, along with assistants, associates and office coordinators.

Where to look on the company side:

• Leadership and team pages, for names and titles.
• Press releases, which often end with a named contact and a real address.
• Author bios on the company blog.
• Project and case study pages that name the people who ran the work.
• LinkedIn, to confirm the person still holds the role.

Then use the pattern guess to find the address:

1. Find one published address on the same domain. A search such as "@contractorname.co.uk" site:contractorname.co.uk often turns one up.
2. Read the pattern from it: first.last, first initial plus surname, or first name only.
3. Apply that pattern to your person, for example jane.doe@contractorname.co.uk (a made-up address).
4. Run the guess through a verifier before it goes near a campaign.

Hand-built rows are slow, so spend them where they pay off: your most important accounts, or a market small enough to cover completely. Hand-checked rows are one stage of a step-by-step B2B lead list build.

Is it legal to scrape emails for cold email?

Collecting business addresses from public pages can be lawful. Finding them gives you no permission to email them. In the UK you can email corporate subscribers (limited companies, LLPs and government bodies), provided you meet UK GDPR. In the EU you need a lawful basis, a notice naming your source, and the recipient country's own email rules.

In the US every message must meet CAN-SPAM. This is not legal advice.

In the UK, the ICO explains in its guidance on business-to-business marketing that PECR's email consent rules do not cover corporate subscribers. That page names Companies House, press coverage, social media and a company's own website as public sources. Using them still means meeting PECR and UK GDPR wherever personal data is involved.

If you collect someone's details in their work role to market to them, you have to tell them, and you need a lawful basis. The ICO's electronic mail marketing page adds that the soft opt-in (the exception for emailing your own past customers) does not reach new contacts taken from bought-in lists.

In the EU, Article 14 of the GDPR applies whenever the data did not come from the person. You must say which source it came from and whether that source was publicly accessible. If you use the data to contact the person, that notice is owed no later than your first message to them. Scraped and enriched contacts fall right under this rule.

Article 14 covers the notice only. Article 6 of the same regulation still requires a lawful basis for using the data, and national email rules differ by country, so check each recipient's country before you send.

US law treats a sourced B2B list like any other. The FTC's CAN-SPAM compliance guide is plain that B2B messages get no exemption from the Act. Whichever route filled your list, each message carries honest header and subject information plus your postal address, and every opt-out request must be honored within 10 business days.

Gmail adds a platform rule on top: Google's email sender guidelines tell senders not to purchase email addresses from other companies.

My rule: add a source column to every row, with where you found the contact and on what date. Without it you cannot write an Article 14 notice, and you cannot answer a prospect who asks how you got their address. If you cannot name the source of a row, delete the row.

Location changes my answer too. A sourced row located in Germany, Austria or Czechia stays out of the email send unless a compliant basis exists, and LinkedIn handles that contact.

How do you verify email addresses before cold email?

My team verifies every address right before it enters a campaign, and only verified addresses get sent. We keep the bounce rate below 0.5%, using NeverBounce or DeBounce.

Bounces cost you more than the wasted send. Much more. Enough of them and mailbox providers start distrusting your domain, and that distrust drags down every campaign that comes after it.

Sending platforms police this too. Infobip's deliverability documentation says a healthy bounce rate should stay below 2%. It warns senders above 3% and disables the sender above 5%. If your sending side is not ready yet, sort out the domains, inboxes and warm-up before you worry about the list.

Two common tools, at the time of writing:

• Hunter. Its Email Verifier checks syntax, domain information and the server response, and looks every address up in Hunter's own B2B database. By the vendor's own figures, the free plan verifies up to 100 addresses a month.
• NeverBounce. It sells pay-as-you-go list cleaning at $8 per 1,000 credits, and credits expire 12 months after purchase. It also has a Growth plan at $49 a month that syncs with your CRM. Both prices are the vendor's own.

Each tool labels its results slightly differently. These five cover the usual outcomes and what to do with each:

The catch-all row is where self-built lists often leak. A domain that accepts everything looks fine in a spreadsheet, and the bounce only shows up after you send.

BulkEmailChecker, a verification vendor, cites industry data putting 8.6% to 15.25% of addresses in an average list on catch-all domains, with B2B lists often at 20-30%. Treat those figures as one vendor's rough range. If your list comes back well above it, check the source before you check the copy.

How many contacts do you need for a first cold email campaign?

Plan for at least 20,000 contacts on the first list, and count your target market first to see whether it can supply them. For planning, I count on needing at least 10,000 people contacted every month to land 30 to 35 B2B sales calls in that month. A first list of 20,000 covers about two months at that pace, which gives the market time to answer.

Here is one anonymized client, an SEO agency in Singapore: 78 booked sales calls in its first 30 days. The funnel behind them: about 16,000 people contacted, about 800 replies, 78 booked calls. That is roughly 200 people contacted for every booked call.

Treat the 10,000 figure as a planning number, not a promise. The offer, the industry and what prospects find when they look you up all move it.

Why so many? I treat 50 to 100 "perfect fit" leads a month as a coin flip. In any given month most of them are busy, months away from buying or simply uninterested, and a list that small runs out before you learn anything. Volume evens out luck.

Worked example for the site safety supplier from earlier:

1. Decide how many roles per company you will keep, say 3: health and safety, procurement and contracts.
2. If each company gives you 3 sendable roles, 20,000 contacts means about 6,700 companies to source. Fewer roles per company means more companies.
3. Check that your target countries and company sizes actually hold that many contractors.
4. Replace the rows that fail verification before launch, so the list still holds 20,000 sendable contacts.

If your real market holds fewer decision-makers than that, the rule changes shape. Use everyone who fits and do not pad the list with weak rows just to hit a number.

Also plan past the first list. Outreach that runs all year needs a steady supply of new, verified rows. More volume means more chances to win. A one-off scrape or a cheap file of a few thousand rows cannot keep that going.

What should you do with free lists, role emails and stale data?

Free-only lists. Free sources and a free verifier tier are fine for two jobs: proving your titles are right, and building rows for your top accounts. They cannot fill a list of the size above in any reasonable time. Use them to test your targeting, then move the volume to a paid route.

Role addresses. Addresses like info@, sales@ and office@ reach whoever reads a shared inbox, which is rarely the person who signs. Keep one only when a small company publishes no named contact at all. Never let role addresses become your main source.

Stale data. People change jobs all the time, so last quarter's file is already out of date. B2B data decay means a list loses accuracy every month it sits. Before you reuse an old export, recheck the titles and run it through the verifier again.

Cheap freelancer lists. In my experience, cheap freelancer files bring weak opens and a pile of bounces. Every legal problem above applies to these files too, and they usually arrive with no source column.

Check what kind of addresses you are buying, too. Salesforce's Account Engagement documentation says role-based addresses such as info@ or sales@ often reach multiple recipients, so they are more likely to generate spam complaints. Higher Logic rejects them on bulk upload because they cause high bounce rates and spam complaints. Avoid these moves:

• Buying a file because the price per row looks low.
• Accepting "it's clean" without the name of the verification tool and the date it was run.
• Keeping role addresses such as info@ and sales@ in the file.
• Mixing a bought file into a verified list, where its bounces hit the same domain.

Do this before you buy anything. Pick your ten most important target accounts and find the decision-maker at each one by hand, with a verified address and the source written down. Time it. If ten rows took an afternoon, you now know manual work belongs to your top accounts, and a platform has to carry the rest of the list.

Fix the list first, then scale.

Trying to decide which route to take for your first list? Book a call, tell me your market and target titles, and I will say which source I would start with.