Account scoring gets confused with lead scoring, and that confusion costs more time than anything else during implementation. Lead scoring rates a person. Account scoring rates a company, in two independent dimensions that have to be calculated separately.
Fit score tells you whether a company matches your ideal customer profile. Heat score tells you whether that company is moving right now. These are two different numbers, and multiplying them together is a mistake, because you lose the information about which one is low.
A company with high fit and low heat is an account to warm up over a quarter. A company with low fit and high heat is a distraction that will eat a rep’s time. A single number cannot tell those cases apart.
Fit is built from traits that do not change week to week: industry, size, revenue, technology stack, sales model, team structure, the markets they operate in.
You do not invent the points at a whiteboard. You pull the customers who stayed longest and pay most out of the CRM and look for what they have in common. Then you do the same for the ones who churned in the first year. The traits that separate those two groups are your fit criteria. Everything else is a hunch.
The most common mistake is scoring on industry, because industry is convenient to filter by. Industry rarely separates good customers from bad ones. What usually does is the situation the company was in at the moment of purchase: a fresh funding round, a new person in a role, a system being replaced.
Recalculate fit once a quarter, no more often. Companies do not change size every week.
Heat is built from behaviour and events that carry a date. Website visits, email opens, LinkedIn engagement, downloads, demos watched, and on the company side: hiring, stack changes, leadership moves, funding.
Two rules, without which heat does not work. First, it is counted at company level, not person level: three people from one company on your pricing page is one strong signal, not three weak ones. Second, points must decay. A visit six months ago says nothing about today, and without decay the list turns into a ranking of every company that ever clicked anything.
Recalculate heat weekly. It is what sets the order of work for the coming days.
| Fit score | Heat score | |
|---|---|---|
| What it measures | Whether the company matches your ICP | Whether the company is moving now |
| Built from | Company traits | Dated behaviour and events |
| Recalculated | Quarterly | Weekly |
| Do points decay | No | Yes, this is required |
| Used for | Who gets on the list | What to work on this week |
Tiers are not a label in the CRM, they are a statement about how much work an account gets. If Tier 1 and Tier 3 are handled the same way, the split means nothing.
Tier 1 is the highest-fit accounts, handled individually: their own account plan, named people in the buying committee, content built for that company’s situation. Realistically that is a dozen to a few dozen companies, because that is what a team can carry.
Tier 2 is good-fit accounts handled in groups of a few to a dozen companies with the same problem. Content is shared across the segment, personalised at the level of industry and situation.
Tier 3 is the rest of the list, handled by campaigns and advertising against the account list. Heat does most of the work here: an account moves up only once it starts moving.
The boundaries between tiers are set by team capacity, not by a point threshold. Start from how many accounts one rep can genuinely run individually, then pick the threshold that lets that many through.
A spreadsheet. With a list in the hundreds, two point columns and a weekly review beat a platform, because everyone can read it and nobody waits for access.
Moving into the CRM or a dedicated tool makes sense once the weekly review stops fitting into a week, which happens somewhere above a thousand accounts and three reps.
What not to do at the start: do not build a model with fifteen variables. Five fit criteria and five heat sources are enough to make the list useful, and a model nobody understands stops being used in week three.
Should I multiply fit by heat? No. The product hides which number is low, and that is the only thing that tells you what to do with the account. Keep two columns.
How many fit criteria? Five to start. A fifteen-variable model looks serious and goes stale within a month.
Does scoring replace talking to sales? No. The Tier 1 list goes past sales before any work starts. The accounts reps strike off are the most valuable information you have about what the model gets wrong.
How do I know scoring is not working? Reps work their own way and ignore the list. The problem is in the criteria, not in team discipline.
Where heat comes from is covered separately in B2B buying signals and B2B intent data. The full model is in the guide to Account-Based Marketing for B2B SaaS, and the terms are in the ABM glossary.
Daniel Tupczyński runs MOAT Marketing, an Account-Based Marketing partner for B2B SaaS companies. MOAT builds and runs ABM systems alongside the client’s own sales and marketing team, from market research and prospecting into Tier 1 accounts through to insights for product and strategy, so the know-how stays in the company. He also runs marketing at a B2B SaaS company.