Lead generation for a business with no sales team
With no sales team, lead generation isn't a funnel. It's forty named businesses on a list, looked at one by one, emailed slowly, and worked through again next month. This page has where leads actually come from for a small local business, how to pull real ones out of Google Maps in twenty minutes, how to throw most of them away before you write, and the send rates that keep you out of the spam folder.
Almost everything written about lead generation is written for a company with a sales team — a top of funnel, an SDR to work it, a CRM with stages, a quota. If that's not you, most of the advice inverts. You don't need more leads; you need fewer, better ones, because the constraint isn't the list, it's the three hours a week you have to spend on it and the fact that you're the person who also has to do the work when it lands.
So the useful question isn't "how do I get more leads". It's: how many real conversations do I need this quarter, and what's the smallest number of businesses I have to look at to get them? For most one-person B2B businesses the answer is startlingly small. A hundred properly researched companies, emailed over a month with a short follow-up sequence, is enough to produce two or three real conversations — and two or three conversations is a good quarter for a business that can only take on so much work anyway.
The running example on these pages is Northline Bike Repair: one founder, a van, Denver, mobile bike repair. Its lead generation problem is specific and it's the shape most small B2B has. Northline wants bike shops to hand it overflow — the tune-ups their workshop can't fit in during spring. There are roughly sixty bike shops in the Denver metro. That's the whole market. You cannot run a funnel over sixty companies; you can know all sixty by name.
Where leads actually come from
Eight channels are realistically open to a small local business. They are not equal, and the two at the top of this table produce most of the work for most small businesses in the country — which is inconvenient, because they're the two nobody sells you a tool for. The numbers below are what a one-person business should expect, not what a channel is capable of at scale.
| Channel | What it costs | First lead | What to expect | Suits |
|---|---|---|---|---|
| Referrals and past customers | Free · an hour | Days | The highest close rate of anything here. Ten specific asks — a name, not 'anyone you know' — typically produce one or two jobs. | Everyone, first |
| Partnerships and overflow | Free · your time | 2–6 weeks | Few leads, each worth a lot and recurring. One good partner can be a quarter of a small business's work. | Trades, services, B2B |
| Outbound from Google Maps | Free · 3 hrs a week | 1–2 weeks | 3–4% reply rate across all cold email; 10–20% on a small, genuinely researched list. Roughly 2–3 conversations per 100 well-chosen companies. | Local B2B |
| Google Business Profile and local search | Free · a few hours | 2–8 weeks | A slow, compounding trickle of people already looking for you. Reviews are the lever, not keywords. | Anyone with an address or a service area |
| Paid ads (Meta, Google) | $300–1,000 / month | Days | Fastest to switch on, and the only channel where the cost is fixed and the return isn't. Needs a page that converts and a way to attribute the sale. | B2C, and B2B with a clear offer |
| Paid lead marketplaces (Thumbtack, Angi, Bark) | $15–80 a lead | Same day | Instant volume, low intent, and you're quoted against four others on price. Useful to fill a gap; corrosive as a strategy. | Home services, in a slow month |
| LinkedIn and industry communities | Free · 2 hrs a week | 1–3 months | Slow to start, then leads arrive pre-warmed. Posting beats connecting; answering questions beats posting. | Consultants, agencies, B2B services |
| In person — trade counters, markets, meetups | Free · half a day | Weeks | Low volume, very high conversion, and it makes the cold email that follows warm. Underrated by everyone who reads about lead generation online. | Local anything |
Reply-rate figures are from 2026 analyses of cold-email sequences at scale (Saleshandy, 53m messages; Instantly, full-year 2025) — the 3–4% is the all-campaigns average, and the 10–20% is what small, hand-researched lists produce. Marketplace lead prices are typical US list prices and vary by trade and city. Everything else is the range small businesses report, not a measured benchmark.
The honest reading of that table: do the top two first and do them properly, because they cost nothing and convert best, and only then add the third. Outbound is on this page because it's the one that scales with effort rather than luck, and because it's the one you can start on a Tuesday with no audience, no budget and no reputation. But a founder who cold-emails sixty strangers before asking ten past customers for a name has done the hard version of an easy job.
Pulling a real list out of Google Maps
Google Maps is the best free B2B database for local businesses, and it's better than a bought list in the way that matters: everything on it is a business that currently exists, at an address, with a phone that rings. It has an owner who updated it. That's already more than most $200-a-month lead databases can promise.
Write the search the way a customer would
Not "B2B partners" — the thing they call themselves, plus where. "Bike shops in Denver". "Physiotherapy clinics near Lakewood". "Commercial landscapers, Aurora". Run it three or four times with the neighbourhood changed, because Maps ranks by proximity and one search only shows you one part of the city.
Zoom the map out to your actual service area and search again from the new centre. A van that covers thirty miles should be running the search from three or four centres, not one.
Take five fields, not fifteen
Name, website, phone, review count and rating. That's enough to qualify with and everything else is a distraction. Review count in particular is the most useful free proxy for size that exists for a local business — a shop with 40 reviews and a shop with 900 are not the same company, whatever their websites say.
Skip anything with no website. Not because a business without one is bad, but because you can't research it in two minutes, which means you can't write to it in a way that earns a reply.
Cut the list in half before you look at anything
Chains and franchises where the decision isn't made locally. Businesses that are visibly a competitor rather than a partner. Anything permanently closed, which Maps still lists. Anything more than your realistic travel time away. For Northline, sixty Denver bike shops becomes about thirty-five once the two chains and the outlying towns come off.
Find a name before you find an address
An email to
info@is a message to a shared inbox nobody owns. Look for a named person on the About or Team page, in the Google reviews the owner replies to (owners sign their replies), or on the shop's Instagram. If a name exists, the guess pattern is usually visible from any address on the site — first@, firstlast@, f.last@.If there's genuinely no name, the generic address is still worth one email — but write it to a person: "whoever looks after the workshop schedule".
Stop at forty
Forty is about what one person can research properly in a week and email carefully over the following fortnight. A list of four hundred is not four times as good; it's a list you will not read, which means generic emails, which means the reply rate drops by a factor of two or three and your domain takes the complaints. The list is not the asset. What you know about the list is the asset.
| Business | Reviews | Website | Keep or cut, and why |
|---|---|---|---|
| Front Range Wheelworks | 412 · 4.7 | Yes | Keep. Big enough to have a queue in spring, one location, owner replies to reviews by name. |
| Baseline Bike Co | 88 · 4.9 | Yes | Keep. Small, service-heavy, visibly proud of turnaround time — overflow is their exact problem in April. |
| Two Rivers Cyclery | 37 · 4.4 | One page, no email | Keep, but last. Too small to have overflow, and there's no way in without walking through the door. |
| Cap Hill Bikes | 203 · 4.6 | Yes | Keep. Careers page has a mechanic vacancy open — they're short-staffed, which is the whole pitch. |
| National chain, Colfax branch | 1,140 · 4.1 | Corporate | Cut. The service contract is signed in another state. |
| Summit Mobile Bike Repair | 64 · 5.0 | Yes | Cut. Same business as Northline. This is a competitor, not a partner. |
| Golden Ridge Cycles | 156 · 4.8 | Yes | Cut for now. Forty minutes each way — the overflow job stops paying at that distance. |
That's the whole method, and the point of showing it as a table is that the fourth column is the work. Anyone can produce the first three columns in ten minutes. The fourth column is why Northline's email to Cap Hill Bikes will open with the mechanic vacancy and get read, and why the email to the chain never gets written.
Qualify before you write, not after
The standard sales qualification frameworks — BANT and its descendants — ask about budget, authority, need and timeline. Every one of those is a question you ask a person on a call. You don't have a call. You have a website, a Google listing, an Instagram and two minutes, and you have to decide from those whether this company is worth a carefully written email or a place on a list you look at again in six months.
Three questions do almost all the work at this stage, and they're deliberately answerable from the outside:
- Do they visibly have the problem?
- Not 'could they benefit'. Is there something on the site, the listing or the reviews that shows the problem exists today — a hiring post, a two-week wait time, an apology in a review reply, a service they've stopped offering.
- Can they pay for it?
- Compare their prices to yours. A shop charging $90 for a tune-up can pay you $45 for one. A one-person outfit with a Squarespace page and no prices probably can't pay anyone yet, however much they'd like to.
- Is there a path in?
- A named person with a findable address, a shared customer, a mutual acquaintance, or something you've genuinely used or been to. A generic contact form is not a path in; it's a lottery ticket.
Score every business on the list against those three before writing anything, sort by the score, and write to the top ten properly. The rest get a shorter email or nothing at all. The full method, including where scoring goes wrong and a scorer you can run on a company right now, is on the lead qualification page.
Then send slowly, on purpose
This is where most small-business outbound dies, and it dies for a technical reason that has nothing to do with the writing. Gmail and Outlook decide whether your mail reaches an inbox largely on your domain's reputation, and a domain you registered last month has none. Send forty emails on day one from a brand-new domain and a meaningful share of them land in spam — you won't know which, because nothing tells you, and you'll conclude the message was wrong when the message was never read.
The numbers that keep you out of trouble are small and boring:
- A domain needs an age
- Let a newly registered domain sit for about a week before it sends anything, then warm it with genuine, low-volume, replied-to conversation for two to three weeks. Three to four weeks from registration to real outreach is the normal timeline.
- 20–30 a day per mailbox
- That's the practical cold-outreach ceiling on a healthy mailbox, against a Google Workspace hard limit of 2,000 external recipients a day. The published limit is not the useful limit — the filters engage long before it.
- Bounces under 2%
- A bounce rate above a few percent is the clearest signal to a mailbox provider that you didn't check your list. Verify addresses before the first send, not after the first bounce.
- Complaints under 0.10%
- Google's guidance is to stay below 0.10% and never reach 0.30%. At 30 emails a day that is roughly one complaint a month before you have a problem — the margin for annoying people is thin.
At thirty a day, five days a week, a hundred-company list with a three-email sequence takes about three weeks to work through. That is the correct speed. It also happens to be the speed at which you can actually read the replies, which is the real constraint — a founder who sends 300 a day and can't answer them has built a machine for annoying people. The DNS records, the warm-up schedule and a planner that gives you a week-by-week send cap are on the cold email page.
The weekly rhythm that survives a busy week
Lead generation fails in small businesses for the same reason bookkeeping does: it's never today's emergency. The fix is to make it small enough that a bad week doesn't kill it. Three hours, split across the week, in fixed slots:
| When | How long | What you do |
|---|---|---|
| Monday morning | 40 min | Pull fifteen new businesses from Maps, cut the obvious no's, find a name for each. Nothing written yet. |
| Tuesday morning | 50 min | Two minutes on each website. Score them. Write the top five emails properly — the other ten get the short version or wait. |
| Tuesday–Thursday | 10 min a day | Send at the cap, in a window, spread out. Read replies the same day; anyone who replies comes out of the sequence immediately. |
| Thursday afternoon | 30 min | Follow-ups on anything sent nine days ago that went quiet. One paragraph, new information, not 'just bumping this'. |
| Friday, ten minutes | 10 min | Look at three numbers only: replies, bounces, meetings booked. If bounces are over 2%, stop and fix the list before sending anything else. |
Run that for a quarter and the arithmetic is roughly: sixty new companies a month, a 4-touch sequence, two or three real conversations a month, one of which becomes work. That's an unglamorous number and it is what actually happens. The businesses that do much better than it are not sending more — they're the ones whose first email mentioned something true and specific about the company, which is a research result, not a copywriting one.
One more thing that matters more than any of this: follow up. Across 53 million cold emails analysed in the first half of 2026, 44% of all positive replies came from follow-ups rather than the first message, and the first follow-up alone accounted for a quarter of them. A founder who sends one email and stops is throwing away about half of what the list was worth.
Common questions
How many leads does a small business actually need?
Far fewer than the advice implies. Work backwards: if you can take on four new clients this quarter and you close one conversation in three, you need twelve conversations. At two or three conversations per hundred well-researched companies, that's four to six hundred companies over a quarter — or, more realistically for a one-person business, a hundred a month done properly. A list of ten thousand is a sign that nobody read any of them.
Is scraping Google Maps legal?
Looking up businesses on Maps and writing down their public details is ordinary research and businesses publish that information so people contact them. Automated bulk scraping is a different matter and breaches Google's terms of service. The safe version is to use the Places API or search manually — and either way, what you do with the addresses afterwards is governed by CAN-SPAM in the US and PECR and GDPR in the UK and EU, not by how you found them. This isn't legal advice.
How long before outbound produces anything?
Two to six weeks to the first real reply if the domain is already warm; four to eight if you're starting from a domain you just registered, because three to four of those weeks are warm-up. The first month usually produces replies and no work. The second month produces work from the first month's follow-ups. Judging outbound after three weeks is judging it before it has run.
Should I use LinkedIn instead of email?
Use both, on different people. LinkedIn works for consultants, agencies and anyone selling to someone with a job title; connection requests with a note get read more often than cold email does. It's poor for local trades, where the owner of a bike shop or a plumbing firm isn't on LinkedIn at all and their email is on their website. Automated LinkedIn tools risk your account, which is a worse thing to lose than a domain.
What reply rate should I expect?
Across all cold email, 3–4% is the honest average — 2026 analyses of tens of millions of messages put it at 3.4–3.7%. Small, hand-researched lists do much better: campaigns under 200 prospects run roughly double the reply rate of campaigns over 500, and 10–20% is achievable when every email mentions something specific about the company. A 1% reply rate means the list was wrong, not the copy.
Do I need a CRM?
Not at forty leads. A spreadsheet with the company, the person, what you noticed about them, the date of each touch and the outcome is enough, and it's better than a CRM you don't fill in. The moment you need one is when you can't remember whether you emailed someone — usually somewhere past a hundred and fifty active names, or when a second person gets involved.
Cold calling or cold email?
For local B2B, a call to a shop at 10am on a Tuesday often beats an email, because the owner is standing next to the phone and you get an answer in ninety seconds instead of nine days. The catch is that it doesn't scale past what your nerve allows, and it produces no record. The combination that works is email first, then a call to the ones who opened and didn't reply — you're no longer a stranger by then.
What do I do with the ones who say no?
Keep them. A no is usually a no-right-now — the shop isn't busy in November and is drowning in April. Note the reason, put a date on it, and come back once, six months later, with a single line referencing the last conversation. That re-contact list quietly becomes the best-converting list you have after a year, and it costs nothing to maintain.
The list, pulled and scored before you open the laptop.
Velofound finds real local businesses on Google Maps, looks each one up and scores it hot, warm or cold, drafts the outreach and sends it paced and capped. Anyone who replies drops out of the sequence. Nothing sends without you. Free to start.
Start free →