How to build a prospect list that books more meetings from fewer names.

The clearest case of how to build a prospect list that books meetings came from a commercial cleaning company that bought a list of 3,000 property managers and sent them a four-email sequence.

Nine people replied, and only one turned into a meeting. Several replies said some version of "we don't manage that building anymore" or "we're not in your area."

The owner, one salesperson and I spent two days building a list of 250 instead, from buildings within a 20-mile drive of their crews and over 50,000 square feet, the size they clean best. The new list named each building's property manager and noted one detail per building, like a recent renovation or a new tenant. The same sequence went out with that detail in the first line, and fourteen people replied and five booked a walk-through.

The bought list cost less per contact and more per meeting. The method below works in HubSpot, Salesforce or a spreadsheet.

The list of 250 got more replies and five times the meetings.

Same four-email sequence, two lists, replies and meetings booked

The list of 250 got more replies and five times the meetings.

Item	Replies	Meetings booked
Bought list of 3,000	9	1
Built list of 250	14	5

Why the list decides more than the copy.

Hopkins named the problem in 1923.

"But the salesman wastes much of his time on prospects whom he never can hope to interest. He cannot pick them out." Claude C. Hopkins, Scientific Advertising (1923), chapter 5, "Headlines"

Hopkins's answer was the headline, which lets interested readers pick themselves out. A cold email arrives unasked, and the prospect list has to do that picking first.

In randomized field experiments published in Marketing Science in 2018, Navdeep Sahni, Christian Wheeler and Pradeep Chintagunta emailed millions of people for three companies. In the main experiment, adding the recipient's name to the subject line raised sales leads by 31%, from 0.39% to 0.51% of recipients. In the cleaning company's case, meetings per contact went from 1 in 3,000 to 1 in 50.

That isn't a controlled comparison (consumer email against B2B outreach), and it suggests the order of work: the copy change moved results by tens of percent and the list change by a multiple. The researchers' tests suggest personal details make people read the rest more closely, which only helps when the detail is true.

A wrong list also costs you the inbox.

Email to people who can't buy or left last year gets ignored or marked as spam. Google asks senders to keep Gmail spam reports under 0.1% and never reach 0.3%, and Yahoo also draws its line at 0.3%. Both cover consumer inboxes (Google's FAQ excludes Google Workspace accounts), and mail to a company domain still goes through Google's, Microsoft's or the company's own filters.

Permission rules reach further. Yahoo's guidance says, "Don't purchase mailing lists," and HubSpot's anti-spam policy bans purchased and third-party lists and asks for bounces below 5% and complaints below 0.1%. On October 21, 2026, Spamhaus wrote that "lack of permission is one of the two most common reasons we make a listing." When cold email goes out from the same domain as your client mail, every contact who can't buy puts that mail at risk.

Every list starts going stale the day it's built.

In January 2026, 20.6% of U.S. wage and salary workers had been with their employer a year or less, according to the Bureau of Labor Statistics (a supplement to the Current Population Survey, about 60,000 households).

Half of workers aged 35 to 44 had been with their employer under five years.

Median years with current employer, U.S. wage and salary workers, January 2026

Half of workers aged 35 to 44 had been with their employer under five years.

Item	years
Ages 25 to 34	3.0
Ages 35 to 44	4.7
Ages 45 to 54	7.0
Ages 55 to 64	9.6

That counts moves between employers only, and people also change buildings and roles inside a company. Re-check a list before every reuse.

How to build a prospect list in five steps.

1. Start from the clients you'd want more of.

Pick by profit and fit, like your 10 best clients at a small company. Note what they share (industry, size, location, who signed, the problem that brought them in) and write the group in one sentence: "property managers of buildings over 50,000 square feet within a 20-mile drive of our crews."

At a larger company, run a closed-won report for the last 24 months by industry and size, with win rate and margin per group, and record the result in HubSpot's default Ideal Customer Profile Tier property (Tier 1 is the closest match) or a custom Salesforce Account field.

2. Set filters you can check by hand.

The second list used four filters (place, size, named person, current detail), each one checkable from a public source. A filter you can't check, like "growth-minded companies," lets the wrong accounts in.

Name each list after its filters and date, like "2026-11-property-managers-50k-sqft-20mi". At a larger company, keep one list per territory and product line with the filters in its description.

3. Find the person by name, and a reason now.

Every row needs a named decision maker, a verified email and one detail that shows a reason to need you now (a renovation, a new location, recent job openings). If you can't find a detail, the account usually isn't ready for an email written for it.

Clay's waterfall enrichment checks several data providers in turn, and for a few hundred accounts, company websites, LinkedIn and local news are enough.

Verification confirms a mailbox exists, not that the person still runs the building. Treat "catch-all" results (Kickbox defines them as domains that accept mail for any address, valid or not) as risky and send to them in small batches, if at all.

4. Score every account from 0 to 10.

Before anything is sent, add the fields that decide fit (size, industry, location and a signal like recent job openings) and score each account from 0 to 10 on how closely it matches your best clients, weighting the fields that matter most. An enrichment tool can fill in most fields, and the template below has the spreadsheet formulas.

One mistake to avoid is scoring missing data as zero. Small companies with little public data land at the bottom, including some that look like your best clients. Count missing data as neutral and give those rows a manual check.

In Salesforce formula fields, the "Treat blank fields as blanks" option and the BLANKVALUE function keep an empty field from becoming a zero. In a spreadsheet, use IF(ISBLANK()).

At a larger company, put the score on the account record, refresh it monthly and compare it with closed-won deals after a quarter. If 8-plus accounts don't close better than 5-to-7 accounts, change the weights.

5. Match the effort to the score.

Accounts scoring 8 or higher get a first line written from the details you found, and a person reads each one before it goes out, even when an AI tool drafts it. Accounts at 5 to 7 get an email written for their industry. Accounts under 5 wait, which protects your domain for the accounts that fit.

The cutoff is a cost decision. Jan Roelf Bult and Tom Wansbeek (Marketing Science, 1995) picked direct mail households by setting marginal cost against expected return, and an account earns a researched email when the expected value of a reply covers the research time and the risk to your sending reputation.

A worked example: cost per meeting.

Cost per meeting = (data cost + research hours × hourly cost + sending cost) ÷ meetings booked.

Example numbers, with assumed costs and the case's results:

• Bought list: 3,000 contacts at $0.50 is $1,500, plus $200 of sending tools, or $1,700 for one meeting.
• Built list: 250 contacts take two people on your team three days, 48 hours. At $40 an hour that's $1,920, plus $150 of data and $50 of sending, or $2,120 for five meetings, $424 each.
• Break-even: the built list needed only two meetings to win ($1,060 each), and its filters carry over to the next 250.

Convert benchmarks before you compare. Instantly's 2026 report, from billions of cold email interactions on its platform in 2025 (Instantly sells sending and lead data), puts the average reply rate at 3.43% of emails sent. Per person, the bought list got replies from 0.3% and the built list from 5.6%, and over four emails each, both sit under that average, which counts every reply. Let meetings per contact decide.

The templates.

Common mistakes, and the pushback a specialist would raise.

"The test wasn't clean, because the first line changed too." True, and that first line only existed because the list was researched. The replies to the bought list said the list was wrong ("we don't manage that building anymore," "we're not in your area"), and no first line fixes that.

"A built list is bought data too." Often it is, because enrichment tools pull from the same third-party data providers that sell lists. The difference is that each row passed the four filters and a person's check and carries a reason to write now, which usually keeps complaints low.

Neither list has permission, and that decides where it can go out. Mailchimp bans "publicly available data" lists by name, and HubSpot bars using any HubSpot service to contact third-party lists. Keep a cold list out of Mailchimp entirely, score and track it in your CRM, and send it one to one from a cold outreach tool, with an opt-out in every email.

"Send more email from more inboxes, because volume is the strategy." High-volume outbound can fit a huge market with a simple offer, and every unwanted email adds to the complaints that get senders listed. When the market is a few thousand accounts, a list you can defend usually costs less per meeting.

Skipping the law because it's B2B. The FTC's CAN-SPAM guide says the law "makes no exception for business-to-business email," and the FAQ below lists some of its rules.

A test you can run this month.

1. Pull 150 contacts at random from the list you'd normally use (cell A, logged with the month, like "2026-11-list-test-cell-a"), and build 150 with the four filters and a score of 8 or higher (cell B).
2. Send both the same sequence, with no added first line in either cell, on the same days from the same inboxes, alternating sends.
3. Wait three weeks for the follow-ups to finish and the late replies to arrive.
4. Measure right-person and wrong-person replies per contact, bounces, meetings per contact and cost per meeting.
5. Set the rule in advance. At 150 a cell, one or two meetings either way is noise, and replies from the right person decide it. If cell B gets at least twice cell A's rate and at least five more of those replies, confirm with a second batch, then move the list budget.

Answer the wrong-person replies the test turns up (the cold outreach guide covers how), and for a list you already own, start with the list-first rule.

Frequently asked questions.

What should a prospect list include?

The account, a named decision maker, a verified email with its status, the source and the date found. For outreach, add one current detail with a link, a 0 to 10 fit score and the date each person last heard from you.

How many contacts should be on a prospect list?

As many as you can verify and write to well. In the case above, 250 researched names beat 3,000 bought ones. Start with what two people can build in a few days and grow once you know which filters produce meetings.

Is it legal to buy an email list for B2B outreach?

In the U.S., CAN-SPAM doesn't ban bought lists, and it covers B2B email and requires an opt-out honored within 10 business days and a valid postal address. Mailchimp and HubSpot bar sending to purchased lists through their tools. The EU, UK and Canada have their own rules, and a lawyer can check yours.

How often should you clean a prospect list?

Before every send, and with a 20-contact audit when replies drop. About one in five U.S. wage and salary workers had been with their employer a year or less in January 2026 (BLS). Remove hard bounces right away, recheck catch-all addresses and re-verify anyone not emailed in three months.

Sources.

• Claude C. Hopkins, Scientific Advertising, 1923, chapter 5; text as reprinted by Carl Galletti, p. 27. Internet Archive text
• Navdeep S. Sahni, S. Christian Wheeler and Pradeep Chintagunta, Personalization in Email Marketing: The Role of Noninformative Advertising Content, Marketing Science 37(2), 2018. DOI
• Jan Roelf Bult and Tom Wansbeek, Optimal Selection for Direct Mail, Marketing Science 14(4), 1995. DOI
• U.S. Bureau of Labor Statistics, Employee Tenure in 2026, September 24, 2026. BLS
• Google, Email sender guidelines and Email sender guidelines FAQ, checked October 22, 2026. Guidelines and FAQ
• Google Workspace Admin Help, Add custom spam filters to Gmail, checked October 22, 2026. Google
• Microsoft Learn, Anti-spam protection, updated July 24, 2026. Microsoft
• Yahoo, Sender Best Practices, checked October 22, 2026. Yahoo Senders
• The Spamhaus Team, One clear set of rules: the Spamhaus Sender Guidelines, October 21, 2026. Spamhaus
• Mailchimp, Acceptable Use Policy, checked October 22, 2026. Mailchimp
• HubSpot, Anti-Spam Policy, last modified February 2, 2026. HubSpot
• HubSpot Knowledge Base, HubSpot's default company properties, updated August 7, 2026. HubSpot
• Salesforce, Examples of Advanced Formula Fields, updated July 22, 2026. Salesforce PDF
• Kickbox, Terminology (Accept-All or Catch-All Email Address; Deliverable), checked October 22, 2026. Kickbox Docs
• Clay, Waterfall enrichment, checked October 22, 2026. Clay
• Instantly, Cold Email Benchmark Report 2026, updated January 12, 2026. Instantly
• Federal Trade Commission, CAN-SPAM Act: A Compliance Guide for Business, checked October 22, 2026. FTC