Qualify a lead by actually looking it up
Qualifying a lead before the first email isn't a discovery call — it's two minutes on their website. BANT asks four questions you can't answer yet; three others you can answer from the outside, and they predict a reply better. Here's what to look at, a scorer that turns what you found into hot, warm or cold with the reasoning, and what each verdict should actually change about the email you write.
A lead arrives as a name and a URL. Nothing else about it is known. In the next two minutes you have to decide one of three things: write to this company today by hand, put it in a sequence and spend the time elsewhere, or don't write to it at all. That decision is lead qualification for a small business, and it is entirely different from what the phrase means inside a company with a sales team — where a lead has already raised a hand and qualification means deciding how fast to chase it.
The two minutes are not wasted time before the real work. They are the work. A first email that names something true and specific about the company reads as a person; one that could have been sent to four hundred companies reads as what it is. The measured gap is large: campaigns aimed at fewer than 200 prospects run roughly twice the reply rate of campaigns over 500, and the mechanism isn't list size — it's that the person with the short list read the websites.
Northline Bike Repair, the one-founder mobile bike repair in Denver used as the example across these pages, has thirty-five shops on its list of possible overflow partners. Two minutes each is seventy minutes, once, and it decides which five get a hand-written email and which thirty get three sentences or nothing. That is the entire economics of it.
Why BANT is the wrong tool here
BANT — Budget, Authority, Need, Timeline — was written at IBM for enterprise reps qualifying inbound enquiries on a call. Each letter is a question you ask a person who has agreed to talk to you. Before a first cold email you have none of those conditions, and each letter fails in its own specific way.
- Budget
- Nobody has a budget line for a supplier they've never heard of. A small business doesn't allocate; it decides, out of this month's money, when something hurts. Asking about budget qualifies out every good lead you have.
- Authority
- In a nine-person company everyone is the decision-maker and there's no org chart to check. Worse, chasing 'the decision-maker' makes you skip the person who actually cares — the workshop manager drowning in bookings, who will take your email to the owner for you.
- Need
- The only letter that survives, and it's the one that gets watered down. 'Could they benefit' is not need. Need is something visible today: a job advert, a two-week wait, an apology in a review reply.
- Timeline
- Unknowable from the outside, and mostly irrelevant. A small business's timeline is 'when it becomes the worst thing on the list'. Your job is to be memorable when that happens, which is a follow-up problem, not a qualification one.
The successor frameworks — MEDDIC, CHAMP, ANUM — are better sales tools and worse pre-email tools, because they ask for more information from a conversation you haven't had. They belong after the first reply, when qualification stops being a research exercise and becomes an actual conversation. Before that, you need questions that a website can answer.
The three questions that survive
These three are answerable from the outside, and between them they carry most of what predicts a reply. Notice that none of them asks how much you would like this company to be a customer.
1. Do they visibly have the problem — today?
Not "would they benefit". Everyone would benefit. Is there something public that shows the problem exists right now? An open vacancy for the role that does the thing you do is the strongest single public signal a small company emits — it's a company saying, in writing, that it is short of capacity. After that: a stated lead time, a service quietly dropped from the site, a review reply apologising for a wait, a job page that's been open for two months.
2. Can they pay for it?
Compare their prices to yours, not their size to your hopes. If a shop charges $90 for a tune-up, paying Northline $45 to do one is inside its normal spending and is a decision, not a purchase. If a business has no prices anywhere and one page on Squarespace, the honest read is that it can't pay anyone yet — and a keen lead who can't pay costs you more time than a cold one, because keenness is convincing.
3. Is there a path in?
A shared customer, a mutual acquaintance, a named person with a findable address, or something you have genuinely used or been to. A contact form is not a path in. This is the one dimension you can improve without waiting: an hour spent finding the workshop manager's name is worth more than an hour spent rewriting the email.
Score a lead you're looking at
Open a company's website in another tab, spend the two minutes, and answer these five. The weights are the ones described above — the visible problem and what's changing carry most of the score, size and budget act as gates rather than engines, and a company that has already solved the problem is a no whatever the total says. It loads with a real case from Northline's list: a mid-sized shop with a mechanic vacancy open. Nothing you type leaves your browser.
- Verdict
- Hot · 72/100 A visible problem, something changing, and a way in. This is worth an hour of your Tuesday.
- Raises it · Sizesmall enough to decide in a day, big enough to have the problem
- Raises it · What's changingan open vacancy is the single most reliable public signal that a company is short of capacity
- Neutral · The problema reasonable guess, but the first email has to test it rather than assert it
- Write it todayWrite it today, by hand. Open with the role they're hiring for — one specific true sentence about them — then one question and nothing else. Under 80 words, no attachment, no calendar link in the first email.
These are the same weights Velofound's Lead qualifier agent applies — the visible problem and what's changing carry most of the score; size and budget are gates rather than engines; an incumbent who has already solved it is a no whatever the total says. The difference is that the product does the looking-up itself, from the company's website and listing, before the score exists. Nothing typed here leaves your browser, and a score is a starting point, not a verdict on a business you know better than a rule does.
This is the scoring, with the looking-up left to you. In Velofound the same weighting runs inside a Lead qualifier agent that does the research itself — it reads the company's website and its Google listing, works out size, what's changed and whether the problem is visible, scores it hot, warm or cold, and hands the hot ones to Sales follow-up to draft an email that mentions what it found. You approve or bin the draft; nothing sends on its own.
What the verdict actually changes
A score that doesn't change your behaviour is decoration. Three verdicts, three genuinely different next actions — and the difference is mostly how much of your time each one gets.
| Verdict | Your time | What happens | Share of a list |
|---|---|---|---|
| Hot | 20 min each | Written by hand, today, opening with the specific thing you found. One question, under 80 words, no attachment and no calendar link in the first email. If they don't reply, you call them. | 5–10% |
| Warm | 3 min each | Into a three-email sequence over about eighteen days. The first email still mentions something specific — a template with a real first line beats a personal essay you never send. | 40–50% |
| Cold | 0 min | Not emailed. Onto a watch list with a date six months out and one line saying what would have to change. Re-scored then, in ten seconds. | 40–55% |
| Shop | What two minutes found | Score | What Northline does |
|---|---|---|---|
| Cap Hill Bikes | 8 staff. Careers page has a bike mechanic vacancy, posted seven weeks ago. Tune-up listed at $95. Owner replies to reviews and signs them 'Dana'. | 76 · Hot | Hand-written email to Dana today, opening with the vacancy. 'You've been looking for a mechanic since July — I'm a mobile one you don't have to hire.' |
| Two Rivers Cyclery | Two people. One-page site, last updated 2023, no prices, no names, no careers page. 37 Google reviews, most recent four months ago. | 31 · Cold | Not emailed. Watch list, March. What would change it: a second mechanic, or a review complaining about turnaround. |
The uncomfortable part of that table is the last column of the first row: forty to fifty-five per cent of a list you spent a morning building doesn't get emailed at all. That is the correct outcome and it's the hardest part of the discipline to keep. Every email sent to a cold lead costs you a small piece of your domain's reputation and buys you a bounce — and it displaces the twenty minutes that should have gone to a hot one.
Four ways scoring goes wrong
Scoring how much you want them
The biggest account on the list gets marked hot because it would change your year. That's a wish, not a signal. The tell is that you can't point at anything on their site that produced the score. Force yourself to write the sentence you'd open the email with; if you can't write one that's true and specific, the lead isn't hot, it's wanted.
Scoring size instead of situation
Bigger companies have more money and more inertia. A fifteen-person firm that just opened a second location is a far better lead than a two-hundred-person firm where nothing has changed since 2019, because change is when a company is willing to try a new supplier. Size is a gate — can they pay, at all — not the engine of the score.
Scoring once and never again
A score is a photograph of a company on one Tuesday. Six months later the quiet shop is drowning, the drowning shop has hired two people, and the one that ignored you has a new owner. Re-scoring a cold list costs ten seconds per company because you already know where to look, and re-contacts convert better than first contacts — you're not a stranger, you're the person who wrote in March.
Treating the score as the decision
Rules are consistent, which is exactly why they miss things. You know that Baseline's owner is retiring, that the shop on Colfax fell out with its last supplier, that this trade goes quiet in November. None of that is on a website. A score is a way of spending your attention in the right order; the judgement stays yours, and the same is true when the scoring is done by software.
Common questions
What is lead qualification?
Deciding, before you spend time on a lead, whether it's worth the time — and how much. For a business with a sales team it means ranking inbound enquiries. For a small business doing outbound it means the opposite: looking at a company you found, from the outside, and deciding whether to write to it by hand, put it in a sequence, or leave it alone. The output is a verdict that changes what you do next, not a number in a CRM field.
Is BANT still useful?
After the first reply, yes — budget, authority, need and timeline are reasonable things to establish in a conversation. Before the first email, no: all four require someone to answer you. Qualifying a cold lead has to run on what a company publishes about itself, which is why the three questions on this page are about visible problems, prices and paths in rather than internal facts.
How long should qualifying one lead take?
Two minutes, and it should be capped at two minutes. Five tabs: careers, about, services and prices, recent Google reviews, last social post. If you can't form a view in two minutes the answer is usually warm, and warm means a sequence rather than more research. The failure mode isn't spending too little time, it's spending twenty minutes on a company you then email anyway.
What's the difference between a lead score and a lead grade?
In big sales tools, a grade describes fit — how much this company looks like your best customers — and a score describes behaviour, like opening emails or visiting the pricing page. Before a first email there is no behaviour to score, so what this page calls a score is really a grade: fit plus visible urgency. Behaviour scoring becomes useful later, once they've clicked something.
Can AI qualify leads accurately?
It can do the looking-up reliably and consistently, which is most of the work and the part humans skip when they're busy. It reads a website, a Google listing and public pages, extracts size, recency and visible problems, and applies the same weights every time — no fatigue on lead thirty-eight. What it can't know is what you know: who's retiring, who fell out with whom, which trade goes quiet in November. That's why a score should arrive as a recommendation with its reasoning attached, and why nothing should send on the strength of one.
Should I ever email a cold-scored lead?
Rarely, and never in bulk. The exception is a genuinely cheap one-liner sent once — no sequence, no follow-up — where the cost of being wrong is a single ignored email. What you shouldn't do is put cold leads into a three-touch sequence to make the numbers look busier: that's three times the complaint risk on your domain for the least likely reply on the list.
How many leads should be hot?
Between five and ten per cent of a properly built list, which is one reason the list should be forty and not four hundred. If a third of your list is scoring hot, the criteria have drifted — usually 'visibly has the problem' has quietly become 'could benefit'. If none of it is, the list is wrong: you're searching for the wrong kind of company, or in too small an area.
Does a good score mean they'll buy?
No. It means the email is worth writing. A hot score predicts a reply, not a sale, and the honest conversion from a first cold email to work done is small at every stage: roughly two or three conversations per hundred well-researched companies, and a fraction of those become jobs. Qualification improves the odds and shortens the list; it doesn't turn outbound into a sure thing.
Or have it do the looking up.
Velofound pulls real local businesses from Google Maps, researches each one, and scores it hot, warm or cold before anything is written — then drafts the email around what it found and leaves the send to you. Free to start.
Start free →