Analytics · Fracmo Blog

Building a Lead Scoring Model With a Spreadsheet

Published September 11, 2026 · 5 min read

Cover art: a rising series of bars
Illustration: NetWebMedia

Sales complains that half the leads marketing sends over are not worth the time, but there is no system for telling good leads from bad ones.

Why this keeps coming up

Every business runs into this eventually, and most figure it out the expensive way — through a mistake, not a plan. Sales complains that half the leads marketing sends over are not worth the time, but there is no system for telling good leads from bad ones. What follows below is how to approach it deliberately instead of reactively.

The problem rarely shows up as a single dramatic failure. It shows up as a slow accumulation of small, avoidable costs — an hour lost here, a confused customer there, a decision made twice because nobody wrote down the first one. None of those costs are individually large enough to trigger a fix, which is exactly why they tend to persist for months or years without anyone stepping back to address the root cause.

The stakes are rarely dramatic in any single instance, which is exactly why this tends to get deprioritized in favor of whatever feels more urgent that particular week. Over a year, though, the cumulative effect of handling this well versus handling it reactively is usually larger than it looks from inside any one month.

The concrete approach

List the handful of traits that best-fit customers actually share

List the handful of traits that best-fit customers actually share — company size, specific need mentioned, source channel — based on real closed-won data. The value here comes almost entirely from consistent follow-through rather than from a cleverer version of the idea. Skipping this step, or doing a half version of it, tends to look a lot like building an overly complex scoring algorithm that nobody on the team can explain — a shortcut that feels harmless in the moment and shows up as a real cost later.

Step 2

Assign simple point values to each trait rather than building a complex weighted algorithm nobody can explain or maintain. This is straightforward to describe and easy to skip under deadline pressure, which is exactly why most teams never actually get around to it. Skipping this step, or doing a half version of it, tends to look a lot like basing point values on assumption instead of real closed-won and closed-lost data — a shortcut that feels harmless in the moment and shows up as a real cost later.

Step 3

Set a threshold score that defines "sales-ready" and route leads below it to a nurture sequence instead of directly to sales. None of this is complicated in theory — the difficulty is almost always in actually doing it consistently rather than understanding what to do. Skipping this step, or doing a half version of it, tends to look a lot like never reviewing scored lead outcomes against actual close rates — a shortcut that feels harmless in the moment and shows up as a real cost later.

Step 4

Review scored leads against actual close rates quarterly, and adjust point values based on what the data shows, not intuition alone. It costs very little to implement, which is precisely why it is worth prioritizing over a more expensive fix aimed at the same underlying problem. Skipping this step, or doing a half version of it, tends to look a lot like keeping the scoring logic hidden from sales, who need to trust it to use it — a shortcut that feels harmless in the moment and shows up as a real cost later.

Step 5

Keep the model in a shared spreadsheet or simple CRM field that both sales and marketing can see and understand. The value here comes almost entirely from consistent follow-through rather than from a cleverer version of the idea. Skipping this step, or doing a half version of it, tends to look a lot like building an overly complex scoring algorithm that nobody on the team can explain — a shortcut that feels harmless in the moment and shows up as a real cost later.

Where this goes wrong

  • Building an overly complex scoring algorithm that nobody on the team can explain.
  • Basing point values on assumption instead of real closed-won and closed-lost data.
  • Never reviewing scored lead outcomes against actual close rates.
  • Keeping the scoring logic hidden from sales, who need to trust it to use it.

How to tell it is actually working

The team stops needing to re-litigate the same decision every time it comes up, because the answer is already written down somewhere everyone can find it. That shift is worth watching for deliberately, since it is easy to miss in the day-to-day and easy to credit to something else entirely.

Bottom line

None of this requires a large team or a big budget — it requires a deliberate process instead of reacting in the moment. Small businesses that get this right treat it as a repeatable habit, not a one-time fire drill.

Start with the smallest version of this that can be done this week, not the fully polished version that keeps getting pushed to next quarter. A rough version in place today beats a perfect version that never ships, and most of what is described above can be revised and improved once it exists — it is much harder to improve something that was never started.

See exactly what each Fracmo plan ships. Every deliverable and every price is public — $249, $999 and $2,490 a month, month-to-month, no discovery call to see a number.

See Fracmo pricing →

Keep reading

FAQ

Questions people actually ask

Does lead scoring require dedicated software?
No — a simple spreadsheet-based model using a handful of traits from real closed-won data can work well at small scale.
How often should a lead scoring model be reviewed?
Quarterly, checking scored leads against actual close outcomes, and adjusting point values when the data shows a mismatch.

Put a fractional CMO to work this week.

Start self-serve — your CRM is live the moment you sign up, and every plan is month-to-month with pricing published in the open.

Start with Fracmo →

Not ready? Run the free AI-visibility audit or compare plans →