Lead Generation and Prospecting
How to Build a Prospect List From Scratch, in Any Industry
A six-step method for a prospect list you can work: describe the buyer, filter, find candidates, record evidence, score them and find the right person.

In this article
- What a prospect list is for
- Step 1: Describe the buyer in a few sentences
- Step 2: Turn the description into filters
- Step 3: Find candidates
- Step 4: Record the evidence along with the names
- Step 5: Score the candidates
- Step 6: Find the right person
- Keep the list small and keep it alive
- Contacting the people on your list: two sets of rules
- A week to a first working list
- Where software fits
- Sources
Most prospect lists fail the same way. They are long, they are made of names, and nothing on them says why anyone should hear from you this month. The owner sends a first batch of messages, gets little back, and decides that outreach does not work. The list was the problem.
A list you can work has three properties. Every row is a business that plausibly needs what you sell. Every row carries a reason to get in touch now, with a note of where that reason came from. And the whole list is small enough that you will actually reach the end of it. This article is a method for building one from nothing, for a business in any industry.
The method has six steps: describe the buyer, turn that description into filters, find candidates, record evidence for each one, score them, and find the right person at the ones that pass. Two made-up businesses carry the examples throughout, so you can see the same steps applied twice: a commercial landscaping contractor and an IT managed-service provider. Every company, number and result in those examples is hypothetical and labelled as such. Neither is a case study.
What a prospect list is for
A prospect list answers one question: who gets a message this month, and why them? It is a working document for a specific stretch of selling, and it is different from a database of every business you could ever sell to.
That distinction shapes how you build it. If the purpose is "everyone who could buy," the list grows without limit and nobody scores it, because scoring thousands of rows is unpleasant. If the purpose is "the next thirty companies I will contact," you can afford to look closely at each one. Closer looking is what turns a name into a reason.
Two terms appear throughout. A prospect is a business you have reason to think might buy and have not yet spoken to. A trigger is something observable that makes a purchase more likely right now: a new building, a growing headcount, a notice asking for quotes, a change of ownership, a contract that is about to renew. Triggers are a large part of what separates a warm approach from a cold one. A trigger you can point to is worth more than matching on size and industry alone.
Step 1: Describe the buyer in a few sentences
Before searching anything, write down who buys from you. Four parts are enough:
- What you sell, in the buyer's terms. Not your service list, but the problem that gets solved.
- Who has that problem. The kind of organization, the size range, and where they are.
- What usually triggers a purchase. The events that make someone start looking.
- Who to leave out. Customers you cannot serve well, or would not want.
The U.S. Small Business Administration frames market research as a way to find customers, and its guidance asks the questions a buyer description has to answer: how large the demand is, where customers are located, what they currently pay for alternatives, and how crowded the market already is. It also separates data that already exists (good for general questions such as industry trends) from research you gather yourself by talking to people (better for specific questions about your own offer). A buyer description starts from the first kind and gets sharpened by the second. Your existing customers are the best source for the second kind, and the most useful thing to ask them is how they decided.
Here is how the two hypothetical businesses might write theirs.
Hypothetical example: the buyer, in four lines. Neither business below is real.
A commercial landscaping contractor, a small company serving one metro area:
- Sells: grounds that look cared for year-round, without the property manager having to coordinate crews.
- Buys: property managers and owners of multi-tenant office parks and retail centers, and homeowner associations for common areas, within about forty minutes' drive of the yard.
- Triggers: a new or newly acquired property, a complaint about the grounds, a change of management company, a contract coming up for renewal.
- Leaves out: single-family residential lawns, and any property under a multi-year contract with a national provider we cannot match on price.
An IT managed-service provider, a small team that runs other companies' computers, security and help desk:
- Sells: someone competent looking after the office's technology, so the owner stops being the unofficial IT department.
- Buys: professional offices of roughly ten to fifty people, such as accounting, legal and medical practices, within the state.
- Triggers: a staff member who handles IT as a side task leaving, a new office or an expansion, a security scare, a software end-of-life notice.
- Leaves out: companies with an in-house IT team, and anything that needs on-site staff outside our service area.
Commercial cleaning companies write the same four lines too: buyers such as medical offices, schools and property managers, with triggers like a new tenant or a failed inspection. The point of showing two other industries is that the shape does not change. Only the contents do.
Step 2: Turn the description into filters
The buyer description is prose. A search needs filters: an industry, a place and a size. Two public tools help you set them, and both are free.
An industry code. The North American Industry Classification System, NAICS, is, according to the Census Bureau, the standard that federal statistical agencies use to classify business establishments. It runs from two digits (a broad sector) to six (a specific national industry), and the Census Bureau notes that County Business Patterns and the Statistics of U.S. Businesses both use it. Finding your buyers' codes gives you the vocabulary that many directories, licensing lists and government datasets share. In the 2022 NAICS manual, for example, 561730 is Landscaping Services, 531312 is Nonresidential Property Managers, 541512 is Computer Systems Design Services, 541513 is Computer Facilities Management Services and 541211 is Offices of Certified Public Accountants. Your buyers' codes are the ones to look up for your own list, and your own industry's code is useful too, because it tells you which competitors the data counts.
A count of how many there are. The Census Bureau's County Business Patterns program publishes, for establishments with paid employees, the number of establishments, employment during the week of March 12, and first-quarter and annual payroll. The data is organized by industry and by the employment size of the establishment, and it is available for the nation, states, counties, metro areas and ZIP codes. Two cautions from the program's own description matter here. It leaves out some industries, including crop and animal production, rail transportation, the postal service and public administration, and most establishments reporting government employees. And it protects individual employers from being identified by using noise infusion, which alters reported values slightly. The Census page also notes that it is evaluating how to comply with a Commerce Department order that prohibits the use of noise infusion, so check the current methodology page before relying on exact figures. A count of establishments is a count of businesses that exist, not a list of prospects: it says nothing about who wants to buy. For list building you need an order of magnitude, not an exact count: is the pool in your area forty companies, four hundred or four thousand?
That number decides what kind of list you are building. Here is arithmetic for the hypothetical landscaping contractor.
The same step tells you when your idea of a buyer is too narrow. If the pool you can see is very small, widen the filters before you invest in researching it.
Step 3: Find candidates
With a description and filters, you are looking for specific named businesses. There is no single source. A short list of the usual places, and what each is good and bad for:
- Trade-association and chamber-of-commerce member directories. They list active businesses and often a contact. Coverage is partial, because only members appear.
- Licensing and registration lists. Many fields require a license, such as medical practices, contractors and some professional services, and the licensing body often publishes who holds one. The data is authoritative about existence and usually thin on contacts. Terms of use vary by place, and some lists restrict commercial reuse, so read the terms of each one before copying it.
- Public notices and bid postings. Local government and some private owners post requests for quotes. These are the clearest triggers you will find, with a date attached. They are limited to buyers who use formal procurement.
- Local business listings and maps. Good for finding what is physically nearby and for seeing the business's own description. Listings go stale, so confirm each business is still operating.
- The businesses' own websites. A company's own site tells you what it does, how large it looks, whether it has several locations and who is named on the team page.
- Your own customers and your network. Ask who else in their building, their association or their supply chain faces the same problem. A name introduced by someone who knows the buyer is warmer than anything a directory yields.
- News and announcements. A new location, a funding round, an acquisition or a hiring wave are triggers. They also arrive unevenly, so this source supplements the others.
These are common places to look, drawn from how prospecting is usually done, and you should expect to use three or four of them for any one list. Each overlaps the others, which is useful: a business that shows up in a directory and in a recent notice is better evidence than one that shows up in a directory alone.
Resist collecting names beyond what you can research. A long column of unresearched names feels like progress and delivers none. It is better to stop at forty names you looked at than to reach four hundred you did not.
Step 4: Record the evidence along with the names
This is the step most lists skip. For every candidate, write down not only who they are but what you saw and where you saw it. Three fields do most of the work: the reason to contact (a trigger, or the match to your buyer description), the source (a link or a named document), and the date you saw it.
The date matters because evidence expires. A trigger from last week is a reason to write today. A trigger from last year is background. Months later, when you pick the list up again, the dates are what tell you which rows to trust.
Keeping evidence also protects you from your own optimism. Writing "looks like they might need this" in a notes column is easy and proves nothing. Writing "posted a request for landscaping quotes on the town's procurement page on [date]" proves something, and gives the first message a specific, true opening sentence.
Here is a column layout you can copy into a spreadsheet. Keep it small: a list with twenty columns does not get kept up to date.
Company Website City / area Industry (your buyer's category or NAICS code) Size estimate (employees or sites) and how you estimated it Reason to contact (a trigger, or "fits the buyer description") Evidence source (link or document name) Date evidence seen Fit score (0 to 2) Trigger score (0 to 2) Access score (0 to 2) Total score Contact name and title (leave blank if unknown) How you know the contact (staff page, directory, phone call, guess) Status (not started, researching, first message sent, replied, not now, no) Next step and date Do not contact (yes or no, and why)
The last column matters more than it looks. A prospect who asks not to be contacted, or whom you promised to leave alone, has to stay visible on the list so that nobody contacts them again by accident. The "how you know the contact" column does a similar job: it stops a guess from hardening into a fact as the list gets passed around.
Step 5: Score the candidates
Scoring turns forty researched rows into an order of work. Keep it light. A rubric with four or five criteria, each scored 0, 1 or 2, is enough, and the criteria should come straight from your buyer description.
- Fit. 0: outside your buyer description. 1: close, with one doubt. 2: clearly the kind of buyer you described.
- Trigger. 0: no trigger found. 1: an old or indirect trigger. 2: a recent, dated trigger you can cite.
- Access. 0: no way to reach a decision-maker found. 1: a general inbox or switchboard only. 2: a named person in a relevant role.
- Capacity. 0: you could not serve them well. 1: you could serve them with effort. 2: a comfortable fit for your team and area.
Add the scores. Decide your cutoff in advance, before looking at any row. For example: contact everything at 6 or above first, research again anything at 4 or 5, and set aside the rest. Deciding the cutoff in advance prevents you from rounding up the companies you happen to like.
Two cautions. First, a score is a ranking aid. It is not a prediction, and a 7 is not a 70 percent chance of anything. Second, if everything scores high, the rubric is too generous, and if nothing does, your filters are probably off. Either result is information.
Hypothetical example: three rows each, scored. The names are invented and do not refer to real companies.
Landscaping contractor:
- Ridgeline Commons (a hypothetical office park): Fit 2 (multi-tenant, common grounds), Trigger 2 (a notice on the owner's page asks for grounds-maintenance quotes, seen this week), Access 1 (a property manager is named on the site, without a direct email), Capacity 2. Total 7: contact first.
- Lakeside Villas HOA: Fit 1, Trigger 0, Access 1, Capacity 2. Total 4: research again before writing.
- Harbor Row Plaza: Fit 2, Trigger 0, Access 0, Capacity 0 (it sits outside the service area). Total 2: set aside.
IT managed-service provider:
- Alder & Finch Accountants (hypothetical, about twenty staff): Fit 2, Trigger 2 (the firm is advertising for an "IT coordinator", which suggests IT is currently a side task), Access 2 (a named office manager on the team page), Capacity 2. Total 8: contact first.
- Brightwater Dental Group: Fit 2, Trigger 0, Access 1, Capacity 1. Total 4.
- Kestrel Legal: Fit 1, Trigger 0, Access 0, Capacity 1. Total 2.
The first row's top score depends on a source you could cite. If you cannot write down where the trigger came from, it does not count as a 2.
Step 6: Find the right person
A prospect list has a company on every row and, for the rows that pass, a person. The right person depends on how the buyer decides, and the title that sounds right is often the wrong one. We wrote a longer treatment of this for one industry in How to Find the Person Who Hires Cleaning Companies, and the logic carries over: start from who pays for the thing you sell, not from a title.
Two habits are worth building into the list itself.
Record how you know. A name from a company's own staff page, a name from a directory and a name you inferred from an email pattern deserve different levels of trust. The "how you know the contact" column from step 4 is where that goes. Before you send to an inferred contact, check it some other way, for example by calling the main line and asking who handles the area.
Allow for "no name found." Some businesses will not publish who decides. A row with a company and no person is still useful: a polite call to the front desk, or a message to a general address that asks who looks after the area, is a legitimate first step. A guessed address on a named person is not.
Keep the list small and keep it alive
Build in batches. A first batch of twenty-five to fifty researched rows is enough to start sending, and you will learn more from the replies than from another hour of research. After the first round, look at what the replies say about your filters. If the best responses all came from one size band or one trigger, adjust the buyer description and build the next batch accordingly.
Then maintain it. Every row needs a status and a next step with a date, and every row's evidence date should be visible. A list that gets looked at weekly stays accurate. One that gets looked at quarterly is mostly out of date by the time anyone uses it. Records of who asked not to be contacted deserve particular care, because they are the part you cannot afford to lose.
Contacting the people on your list: two sets of rules
This article is about building the list, but a list exists to be used, and two rules shape how. This is general information, not legal advice. Rules differ by place and by what you sell, and a lawyer who knows your situation can tell you how they apply.
Email. The Federal Trade Commission's guide to the CAN-SPAM Act states that "The law makes no exception for business-to-business email." Commercial messages must carry accurate header information and a subject line that matches the content, identify themselves as advertisements, include a valid physical postal address, and give a clear way to opt out. The FTC says opt-out requests must be honored within 10 business days, and that the opt-out mechanism must keep working for at least 30 days after the message is sent. It also states that each separate email in violation can carry penalties of up to $53,088. That is a ceiling in FTC enforcement, not a typical outcome. In a notice published in September 2026 the FTC said it would not adjust its civil penalty amounts during 2026 and would keep applying its 2025 levels, so the figure should stand through the year, but check the guide again before you rely on it. Opt-outs belong on your list as "do not contact" rows.
Phone. The FTC's guidance on the Telemarketing Sales Rule says most calls between a telemarketer and a business are exempt from the rule, with exceptions, including business-to-business calls to induce the retail sale of nondurable office or cleaning supplies. It also says that calls to a business line that solicit individual employees for personal purchases are not business-to-business calls. We did not research state telemarketing laws, which can differ, so check yours before a calling campaign.
Both pages are short and worth reading in full before you send anything. Our companion article on commercial cleaning prospecting templates covers first messages, follow-up timing and a pre-send checklist, and while its examples are from cleaning, the compliance and follow-up sections apply to any B2B sender.
A week to a first working list
If you would like a concrete schedule, here is one that fits around other work.
- Day 1: Write the four-part buyer description. Ask two customers how they decided to buy from you, and add what you learn.
- Day 2: Find your buyers' industry codes, set your area and size filters, and estimate how many plausible companies there are.
- Days 3 and 4: Collect candidates from three or four sources. Stop at forty to sixty names.
- Day 5: Fill in the reason, source and date for each. Cross off any row where you cannot write a true reason.
- Day 6: Score the rows with the rubric and decide the cutoff before you look at the totals.
- Day 7: Find a contact for the top fifteen to twenty-five, record how you know each, and write the first messages. Send them the following week, in small batches.
At the end you will have a list that is short, every row of which has a reason and a source. A quick copy of a directory rarely produces that, and it is what makes the first message easy to write.
Where software fits
Good Loux publishes this Journal and sells software for this kind of work, so weigh what follows accordingly. The method above works with a spreadsheet and a few hours. Software helps with the slow parts, and we can describe only what ours does today.
Whatever tool you use, the discipline is the same: a reason, a source and a date on every row.
Sources
We read these directly on November 10, 2026, and checked the CAN-SPAM, Telemarketing Sales Rule and County Business Patterns statements quoted or paraphrased above against the pages again on November 10. Pages change, so check the originals.
- U.S. Census Bureau: County Business Patterns, About the Data for what the program reports (establishments, employment, payroll, by industry and size), its geographies, its exclusions and its disclosure notes
- U.S. Census Bureau: Understanding NAICS for what NAICS is and how its two- to six-digit levels work
- U.S. Census Bureau: 2022 NAICS Manual for the industry titles quoted in step 2. We extracted the text of the PDF and checked the titles; we did not rely on its definitions beyond those titles.
- U.S. Small Business Administration: Market research and competitive analysis for the market questions, the distinction between existing data and direct research, and its list of free federal data sources
- Federal Trade Commission: CAN-SPAM Act: A Compliance Guide for Business for the email requirements, the business-to-business statement, the 10 business day opt-out period and the penalty figure
- Federal Trade Commission: Civil Penalty Inflation Adjustments, notice of September 15, 2026 for the statement that the FTC's 2026 civil penalty amounts are unchanged from its 2025 levels
- Federal Trade Commission: Complying with the Telemarketing Sales Rule for the business-to-business exemption and its exceptions. The page is long; we read its first 100,000 characters.
Everything about how Good Loux works comes from the product itself, not from the sources above. The buyer descriptions, the scoring rubric, the made-up companies and the arithmetic are our own illustrations, not research findings. The list of places to find candidates reflects common practice rather than a study, and nothing here is a claim about any real company.


