Lead scoring: ranking enquiries without inventing a formula

Lead scoring exists to answer one question: of the enquiries sitting in front of you, which gets attention first. It is a prioritisation device rather than a measurement, and that distinction matters because a score that is interesting but does not change the order of your day is costing you more than it returns. Keeping a record of who enquired is a different job, and most teams need that one solved before scoring is worth attempting.

What a lead score should be built from

• Fit: whether they resemble the customers you already serve well — size, sector, location, the thing they do.
• Intent: what they actually did. Requesting a quote outranks downloading a guide by a wide margin.
• Timing: whether they stated a date, a budget cycle or a deadline, which is the most underweighted signal of the three.
• Reachability: a direct contact beats a generic address, because a lead you cannot speak to scores zero in practice.
• Not seniority alone — the person enquiring is frequently researching on behalf of somebody who decides.

Why most lead scoring models are too clever

Because they are built on points that nobody can justify. Twelve behaviours each worth between two and fifteen points, summed into a number out of a hundred, produces a ranking that feels objective and is mostly arbitrary — and nobody can explain why a lead scored sixty-three. A model built from three or four factors, with the weights written down in a sentence each, ranks nearly as well and can actually be corrected when it is wrong. The test of a scoring system is whether somebody disagreeing with a score can point at which factor is wrong.

Score against outcomes you can see, which means waiting until you have enough closed enquiries to compare. Before that point you are guessing at which signals matter, and a guess formalised into a formula is harder to abandon than a guess left as an opinion.

Building one that earns its place

1. Take the last fifty enquiries and sort them by what happened. That list is your training data and it is already in your records.
2. Find the two or three attributes the good ones shared, and check them against the bad ones before believing it.
3. Write the score as a small number of bands — high, medium, low — rather than a figure out of a hundred.
4. Attach an action to each band, because a score with no consequence is a decoration.
5. Review it quarterly against what actually closed, and remove any factor that did not predict anything.
6. Keep it simple enough that a new person can apply it by hand on their first day.

When not to bother

If you receive ten enquiries a week, call all of them. Scoring is a response to having more enquiries than capacity, and applying it below that threshold adds ceremony while losing the one advantage a small company has, which is answering everybody quickly. The honest sequence is: capture every enquiry reliably, respond to all of them, and introduce scoring only at the point where you demonstrably cannot.

Ettex CRM holds the fit and intent attributes as fields rather than as prose, so a band can be filtered rather than calculated, and the enquiry sits beside the record of what happened to it. That record is what makes the next version of the score better than the current one, which is the only way scoring improves.