Lead scoring for small sales teams

Large companies score leads because they have thousands. Small teams need it for the opposite reason: they have very little time. If two people are doing all the selling, every hour spent on a poor-fit lead is an hour not spent on a good one.

Lead scoring is simply a written, shared answer to 'who do we call first?'. Without it, the order is set by whoever shouted loudest or arrived most recently. Speed matters too: research published in Harvard Business Review found that most companies respond to online leads far too slowly, and a clear score is what lets you respond fast to the right ones.

Fit is how closely a company matches your ideal customer profile: industry, size, location, business model, technology. Fit changes slowly. Intent is evidence that they are likely to buy soon: they replied, asked about pricing, are hiring for a relevant role, or just opened a new location. Intent changes quickly.

Score them separately. A high-fit, low-intent company belongs in a nurture list. A low-fit, high-intent company deserves a polite, quick qualification call — and often a 'we're not the right fit'. High fit and high intent goes to the top.

Copy this and adjust the criteria to your ideal customer profile. Keep it to eight to twelve criteria; more makes it harder to maintain, not more accurate.

1. 1Score last quarter's leads. Apply the template to 30–50 leads you already know the outcome of.
2. 2Compare scores with outcomes. Did the deals you won score higher than the ones that went nowhere? If not, a criterion is wrong or weighted badly.
3. 3Set three bands. For example: 70+ gets contacted this week with researched outreach; 40–69 goes into a lighter sequence; under 40 is parked.
4. 4Make the top band small. If more than about a fifth of leads are 'hot', the threshold is too low to be useful.
5. 5Review monthly. Look at which criteria actually predicted replies and meetings, and move points accordingly.

An IT support firm uses the template. Lead A is a 25-person accounting practice in their city (15 + 10 + 10), whose careers page lists an IT administrator vacancy (problem evidence 15, trigger 10), with the managing partner's email published on the site (10). Score: 70. Lead B is a 400-person firm in another state that downloaded a guide (10 for inbound) but is outside the size range and region, with no named contact. Score: 25. Without scoring, B might get the call first because it came in through the website. With scoring, A does, and B gets a polite automated reply.

The hardest criterion to score by hand is 'has the problem you solve', because it needs someone to read the company's website, reviews and job posts. That is exactly the reading AI is good at. A useful setup has AI propose a fit score with evidence for each point, and a person accepts or overrides it.

Be wary of black-box scores. If a tool gives a lead 87 and cannot tell you why, you cannot check it, and you cannot fix it when it is wrong. Prefer scores that show the reasons.

• Let AI do: reading websites and job posts, matching against your brief, summarising evidence, flagging disqualifiers.
• Keep for people: final judgment on borderline leads, intent signals from conversations, changes to the model.

A scoring model only helps if it changes what people do each morning. The simplest working routine for a team of one to five sellers looks like this:

1. 1Score at the moment a lead is added. Fit criteria should be filled in before a lead enters the pipeline, not weeks later.
2. 2Sort the day by score. Start each day with the highest-scoring leads that have not been contacted, then the highest-scoring replies.
3. 3Update intent immediately. A reply, a question about price or a referral changes the score on the spot.
4. 4Record why leads were rejected. A short reason ('too small', 'already has a provider') is the raw material for better disqualifiers.
5. 5Close the loop on outcomes. Mark every lead as won, lost or no response so the monthly review has real data.

Frameworks such as BANT (budget, authority, need, timing) are useful for the qualification conversation, after someone has replied. They are less useful for scoring cold leads, because you rarely know a stranger's budget or timing from the outside. A practical split is to use the points model above to decide who to contact, and a short framework like BANT during the first call to decide whether to keep going. Record the answers, because they become your best intent signals for similar companies later.

• Scoring activity, not buying signals. Opening a message is not intent; asking about price is.
• Never revisiting the model. A model that is never checked against won and lost deals drifts into fiction.
• Too many criteria. Twelve well-chosen criteria beat forty.
• No negative scores. Disqualifiers should push a lead down hard.
• Scores nobody sees. The score has to live where the team works — the inbox or pipeline — or it will be ignored.

Prime Leads Finder's Lead Finder gives each company a 0–100 fit score against your product brief and shows the reasons, which you review before import. Conversations and stages then live in the inbox and pipeline, and Sales Insights helps you see which segments reply and convert so you can adjust your weights. Leads can sync to Atrix CRM through CRM Sync. If you want to try it, request early access.