Lead Scoring for a Small Business in Seven Moves

Lead scoring for a small business comes down to four steps. Write one sentence saying what the score predicts, sort the behaviours you already record into tiers by commitment, give each tier several times the one below, and keep the model small enough to explain in a minute.

Lead scoring for a small business, in seven moves

Forget the forty-signal model. Nobody running a business alone needs it.

What you need is a score you can explain to yourself in a minute. A score you don't trust is a column you never sort by. So the job isn't clever. It's small, written down, and checkable.

Here is the order I'd do it in. One afternoon for the decisions, a little longer for the build.

Move one: say what the score predicts

Before a single point, write one sentence. This person is worth a sales conversation with me now, and a meaningful share of people like them go on to become real opportunities.

Not "engagement." Engagement is what you get when you score without a target. It rewards your most enthusiastic readers, who are frequently students, competitors, job seekers and people who like your writing.

Then check the sentence against the people who actually bought from you. Does it describe them? If not, rewrite it. Every later decision gets tested against that line.

Also know roughly how long your sales cycle is, from first contact to close. You will need it for decay.

Free guide · 21 pagesRation the one thing you cannot scale.The Lead Scoring Playbook: deciding which leads deserve your time.Get the free guide →

Move two: put fit at the door

Fit is who they are. Make it a gate: fits, doesn't fit, unknown. Don't add it to the points.

If your fit data is mostly empty, and for a newsletter list it usually is, use the three states and move on. The longer argument is in fit vs intent in lead scoring.

Move three: rank the signals before you number them

Lead scoring for a small business starts here: list every behaviour you can observe. Then sort them by how much commitment each costs the person. Don't number yet. Rank first.

Now the numbers almost choose themselves. In lead scoring for a small business, each tier is worth several times the one below. The ratios matter far more than the values. If a pile of cheap behaviour can add up to the expensive one, you've built a traffic counter.

Hard offers are different. When someone asks to talk to you, don't check whether their score is high enough. Talk to them. Keep these signals out of the arithmetic entirely, so you're never tempted to set a threshold a direct request could miss.

Move four: cap everything that repeats

This is the technical rule almost everyone forgets. Page views, opens, visits: anything that can repeat needs a ceiling on what it can contribute in total. And the whole score needs a ceiling too.

Without caps, someone who reads your posts daily for a year outranks someone who asked for a call yesterday. That is the opposite of what you wanted. The mechanics are in how to cap lead scoring signals.

Move five: keep only signals you really capture

For most small businesses, a dozen signals is generous.

Go through your tiered list and keep only the behaviours you are genuinely recording today, reliably, for most contacts. A signal you intend to capture next quarter doesn't belong yet. It scores nothing for anyone, and it makes the model look more complete than it is.

Add it when the data starts flowing. Not when you plan it.

Move six: use a short ladder, and stop the score early

You need few stages: subscriber, lead, marketing-qualified, sales-qualified, then opportunity and customer.

The design decision is where the score stops. It can move people through the early stages, because those are judgements about readiness. It must never declare someone an opportunity or a customer. Those are facts about deals. A score that can mark someone a customer will eventually do it wrongly.

And add one safeguard. Nobody becomes marketing-qualified on passive signals alone. They must also have done at least one deliberate thing. Otherwise your threshold, however sensible it looked, quietly promotes every regular reader of your newsletter. The full treatment is in how to define a marketing qualified lead.

Move seven: make every score explain itself

For any contact you should be able to see the contributions that add up to their number, largest first. This form. This many visits. This download, capped here, decayed there.

A number with no reasons gives you every excuse to ignore it. "Applied for the programme, read the pricing page twice this week" has the opening line of your call written for you.

Reasons also find bugs. A strange score with visible reasons is usually obviously wrong.

Two small details. Show a contact the process has never seen as unscored, not as zero. Zero says cold. Unscored says nobody has looked. And recompute on a schedule, a nightly run by your tools, not by you remembering.

Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →

Check the threshold before you work from it

Once the first scores exist, look at how they spread. Marketing-qualified should be a minority, the people worth your time.

If a large share of your contacts sit above the line, the threshold is too low and the label distinguishes nobody. Raise it before you start calling.

Launch strict and relax. Never launch loose and retract. A loose start costs you a month of evenings on calls that should never have been booked.

Then read real examples above and below the line. Do you agree with where they fall? If you don't, your rule is wrong, not your instinct.

Lead scoring for a small business: what to leave out

• Signals you don't record. They look thorough and score nothing.
• Big negative numbers. Use explicit exclusions with reasons.
• Weights tuned on a handful of deals. You'd be fitting to noise.
• A formula you can't say out loud. If it needs exceptions, it needs cutting.

The inputs also drift. A renamed tag or relaunched form can silently zero a valuable signal, with no error anywhere. Keep a running list of inputs in use that map to no signal, and check it whenever you launch something.

A walk-through with an archetype

Say you sell a productised audit to other one-person businesses. Not a real client, just a shape to test the moves against.

The target sentence: this person runs a small business, is looking at their own numbers, and could plausibly pay for an audit now. Fit gate: they run a business of their own, so they fit. Someone with a job, researching for a friend, does not. Someone whose role you can't tell is unknown.

Signals, by tier. Passive: opens an email, reads a post. Engaged: comes back several times in a fortnight. Deliberate: completes your assessment. Hard offer: books a call. Everything repeatable gets a cap. Marketing-qualified needs the assessment or something like it, whatever the pile of reads.

Now run two people through it. A loyal reader opens every email and reads every post. The caps stop the pile growing, and without a deliberate action they never reach marketing-qualified. They stay a reader. Good, that is what they are. A second person reads two pages, visits pricing twice and completes the assessment. They cross, with reasons attached, and you open the call knowing why.

That is the model working. It didn't need forty signals. It needed one honest sentence and a few rules you could say out loud.

Where small businesses get stuck

Three places, in order of how often I'd expect them.

First, the temptation to add more before the first version has run. Run it. Look at the spread. Read the examples. Then add.

Second, treating the first threshold as final. It's a starting guess. Strict is safer, because relaxing a rule feels fine and retracting one feels like failure.

Third, skipping the check against what closed. If you do only one thing after building, plan that. The audit is what turns a guess into a model, and how to audit a lead scoring model shows the shape of it.

Free guide · 21 pagesRation the one thing you cannot scale.The Lead Scoring Playbook: deciding which leads deserve your time.Get the free guide →

Then what

A score only produces business at two points: when it hands someone to you and you act, and when its record is compared with what closed. Building is the easy half. The rest is in how to hand off a hot lead to yourself.

That is lead scoring for a small business done right: small, written down, explainable. A score you will actually sort by.

Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →

Frequently asked questions

How many signals does lead scoring for a small business need?

A dozen is generous. Keep only behaviours you genuinely record today, reliably, for most contacts. A signal you plan to capture next quarter scores nothing and makes the model look more complete than it is.

What should the score predict?

Something specific: this person is worth a sales conversation with you now, and a meaningful share of people like them become real opportunities. Without that sentence, you end up scoring engagement.

Should the score decide who is a customer?

No. It can move people through the early stages, which are judgements about readiness. Opportunity and customer are facts about deals, and a score that sets them will eventually be wrong.

How do I know the threshold is right?

Look at the distribution. Marketing-qualified should be a minority of contacts. If most sit above the line, raise it before you work from the queue.

The full method, with the checklists, is in The Lead Scoring Playbook, a free 21-page playbook.