In this post:
A fit score rates how closely an account matches the companies you sell to, and an intent score rates how likely that account is to buy soon. Fit is about who they are. Intent is about what they're doing this month.
You need both, and you need them as two separate numbers. When a CRM adds them into one lead score, an ideal customer who hasn't thought about you all year and a poor-fit company that's suddenly reading your pricing page can end up with exactly the same value. The rep looking at that number has no way to tell which of the two is in front of them.
Below, the two scores sit side by side, with what goes into each and how to read them together so they decide who gets a call this week.
TL;DR
Fit runs on slow-moving facts (industry, size, region, tech stack, role) measured against your ICP. Intent runs on fast-moving behavior (pricing visits, replies, product sign-ups, research spikes on third-party sites). Score them separately.
One blended number hides the reason behind it. An account at fit 92 and intent 24 averages to the same 58 as an account at fit 31 and intent 85, and those two need opposite handling.
Put fit on one axis and intent on the other. High on both means a rep acts now, high fit with low intent goes to nurture, high intent with low fit gets a careful qualifying step, and low on both gets skipped.
Weight first-party behavior above bought intent data, and read trigger events as timing. Bombora, for example, only counts a company as spiking on a topic at a score of 60 or more against its own baseline.
Fit holds for quarters while intent fades within weeks, so let intent decay and leave fit alone. HubSpot's score decay can shrink old events over 1, 3, 6 or 12 months.
The main tools already keep the pair apart. HubSpot builds fit, engagement and combined scores, Salesforce Account Engagement pairs a score with a grade, and GitLab only treats a 6sense account as sales-ready when fit and buying stage both pass.
The usual mistakes are the blend, chasing intent without fit, pushing hard on fit without intent, and scoring against an ICP nobody has tested. On outbound, fit builds the list and intent orders it. On inbound, both land at the form fill and decide who reaches a rep within minutes.
Fit score vs intent score at a glance
The quickest way to see the difference is to put the two next to each other.
Fit score | Intent score | |
|---|---|---|
Question it answers | Should we sell to this account at all? | Is this account likely to buy soon? |
Typical inputs | Industry, employee count, revenue, region, tools in use, job title and seniority | Pricing page visits, demo requests, email replies, product sign-ups, topic research spikes, trigger events |
Where the data comes from | Enrichment providers and your CRM | Website and product analytics, your sequencer, intent data providers |
How fast it changes | Over quarters, when the company grows, moves or changes its stack | Within days or weeks, and it should fade on its own |
What it decides | Which accounts belong on your list | Which accounts on that list get attention now |
How it usually breaks | An ICP built on opinion instead of won deals | Noise from students, competitors and job seekers who click a lot |
The two are independent. A company can sit at the top of one scale and the bottom of the other, and both readings can be right at the same time. That independence is the whole argument for tracking them apart.
Two accounts, one blended score of 58
Here is the situation that makes the case. Account A is a 400-person software company in your best industry, with exactly the kind of sales leader who buys from you. Nobody there has visited your site in months. Fit 92, intent 24.
Account B is a 15-person consultancy outside your market. Someone there has been on your pricing page three times this week and downloaded two guides. Fit 31, intent 85.
Average the two the way many lead scoring setups effectively do, and both accounts land on 58. They'd sit side by side in a rep's queue with nothing on the record to tell them apart.
The right moves are close to opposite. Account A belongs in a patient nurture track or a well-timed outbound touch, and the day its intent climbs it should jump the queue. Account B needs a quick qualifying question, or a route to a self-serve plan, before anyone books an hour-long demo with it.
Split scores keep that difference visible. A rep who sees "fit 92, intent 24" knows what to do without opening the record, and so does any workflow you build on top.
The four quadrants, and what to do in each
Plot fit on one axis and intent on the other and every account falls into one of four boxes. The box matters more than the exact number, because each box calls for a different kind of effort.
High fit, high intent: act now. These are the right companies showing they're ready. They go straight to a rep, with a response time written into your routing rules.
High fit, low intent: nurture and watch. Usually the biggest box and the one holding most of next year's revenue. Keep these accounts in targeted outbound and ads, and set an alert so a rise in intent moves them up. Acting on that rise is what signal-based selling is built around.
Low fit, high intent: go carefully. This box fools teams more than any other, because activity feels like opportunity. Some of these are students, competitors, agencies researching on behalf of someone else, or companies too small to afford you. A few are a sign that your ICP is too narrow. If won deals keep coming out of this box, change the ICP instead of explaining the pattern away.
Low fit, low intent: leave them. No outreach and no nurture spend. Look again only when their enrichment data changes, for example after a funding round or a big hiring push.
Where you draw the lines is a judgment call. We tend to start with the median of each score as the split, then move the lines after a quarter of checking which box the won deals actually came from.
What actually goes into a fit score and an intent score?
A fit score is built from facts that rarely change. On the company side that's firmographic data such as industry, headcount, revenue and location, plus technographic data about the tools they already run, all weighed against your ideal customer profile.
On the person side it adds role and seniority. The same company can be a great fit through its Head of RevOps and irrelevant through an intern in finance, so a contact-level fit score and an account-level one are both useful.
The weights should come from your own history. An ICP scoring model turns the traits your best customers share into points, and the cleanest way to find those traits is to build your ICP from CRM data rather than from a workshop whiteboard.
An intent score is built from evidence that changes week to week. It usually draws on three kinds of input, and they don't deserve equal weight.
Your own engagement data. Pricing page visits, demo requests, replies to sales emails, free trial sign-ups. This is behavior on your product and your site, so it carries the most weight.
Bought intent data. Research activity on publisher networks and review sites, reported at account level. Bombora's guidance is that businesses with scores of 60 or more on a topic are considered spiking, measured against that company's own baseline. It shows up earlier than your own data but is noisier, which is why we use it as a tiebreaker rather than a trigger.
Trigger events. A funding round, a new sales leader or a hiring push. These buying signals say a fitting account is more likely to move, and a trigger event often explains why intent started rising in the first place.
Put together, the intent score is your read on buying intent turned into something a workflow can act on. If you want the longer version of where this data comes from and how much to trust it, our piece on what intent data is goes through the sources one by one.
Fit moves by the quarter, intent by the week
A company's industry and size barely change from one quarter to the next. An account that fit your ICP in January almost certainly still fits in April, so you can compute fit when enrichment refreshes and trust it in between.
Intent is the opposite. A pricing visit from last Tuesday says a lot. The same visit from last spring says very little, and a research spike can cool off before your next campaign goes out.
That's why intent needs decay and fit doesn't. HubSpot's lead scoring tool has a decay setting that automatically reduces an event's points based on how long ago it happened, with intervals of 1, 3, 6 or 12 months. For most B2B intent signals we pick the shortest one. A company doesn't become a worse fit because it went quiet, so the fit score shouldn't lose points for silence.
The different speeds also suggest a simple way to run the pair. Build your target list on fit and keep it stable. Then let intent, recalculated daily, pick which accounts on that list get worked this week. Mixing them forces you to keep recomputing a number that's half permanent and half perishable, and it ends up useless for both jobs.
How HubSpot, Salesforce and 6sense keep them apart
None of the big platforms asks you to throw the two dimensions away, which says something about how often a single number lets teams down.
HubSpot lets you build fit, engagement and combined scores. Fit qualifies records on their property values, engagement qualifies them on their actions, and turning on a combined score still creates separate fit and engagement properties next to the total. Engagement in HubSpot's sense is the first-party slice of intent. We cover the full setup in our guide to HubSpot lead scoring.
Salesforce Account Engagement (formerly Pardot) uses a score for activity and a letter grade for fit. Its Trailhead module on lead qualification models describes joining the two by combining an increasing score, an estimate of interest, with an increasing grade, an estimate of potential fit.
6sense reports a profile fit rating and a predicted buying stage for each account. GitLab's public marketing handbook shows how one company uses them: its 6sense qualified account rule only switches on when profile fit is Moderate or Strong and the in-market stage is Decision or Purchase. That's fit and intent used as a gate with two locks, not added into a total.
Where fit and intent scoring usually breaks
The blend. One number with no way to see which dimension drove it. Every routing rule and report built on it inherits the confusion, and reps end up second-guessing the score.
Intent without fit. Jumping on every account that moves. It feels productive because these accounts reply, but a lot of the calendar fills up with companies that were never going to sign.
Fit without patience. Hammering high-fit accounts that show no intent at all with sequence after sequence. They weren't ready, and now they've learned to ignore your domain. These accounts needed lighter, slower nurture until something changed.
Stale fit. A fit score built on an ICP nobody has checked against real results. It ranks accounts with great confidence against the wrong picture of a good customer. Pull the last twelve months of won and lost deals, look at what they scored when they first came in, and adjust the weights where the two disagree.
All four come back to the same discipline. Keep the scores separate, check them against deals you actually closed, and let each one do the single job it's good at.
Using both scores on outbound and inbound
On outbound, fit does its work before the first email goes out. It filters your total market down to the companies worth contacting, and intent then decides the order. In the setups we build, Clay usually pulls the firmographic, technographic and signal data, calculates both scores and writes them back to the CRM, so reps see them on the account record and not in a spreadsheet.
On inbound, both scores land the moment someone fills in a form. Fit comes from enrichment on the company, and intent comes from what the person just did plus anything else the account has been doing. A high-fit, high-intent lead is the one a rep should reach within minutes, which is where speed to lead and clear lead routing turn two numbers into a phone call. A high-intent, low-fit lead goes through lead qualification before anyone offers a demo.
None of this replaces lead scoring. Fit and intent are what a good lead scoring model is made of, with fit deciding who belongs in the pipeline and intent ranking them by readiness. Once you have a year or two of outcomes, predictive lead scoring can learn which combinations convert for you. Even without any machine learning, splitting one score into two is a large step up.
In short
The fit score and the intent score answer questions that only look similar. One asks whether an account is worth selling to. The other asks whether this is the moment. A single number can hold one answer or the other, but not both in a form anyone can act on.
Keep them as two fields, let intent fade while fit stays put, and read them together as a grid instead of a total. The quadrants then do most of the deciding: who gets a call today, who stays warm, who gets a careful question first and who gets left alone.
The quiet benefit is trust. When reps can see why an account ranked where it did, they use the ranking. When they can't, they go back to their own hunches.
This is the kind of change we make inside a client's own HubSpot or Salesforce: two score fields, the routing rules that read them, and a one-page note on every weight. After that the scores belong to your team, which is where they should be, since you'll notice your buyers changing before anyone outside does.
Frequently asked questions
Should you build the fit score or the intent score first?
Build fit first. Knowing which accounts are worth pursuing at all is the more basic question, and a simple fit filter already stops most wasted outreach before you buy any intent data. Add intent once you have signals worth feeding it, such as pricing page visits, replies or a third-party feed. Keep the two as separate fields from day one so you never have to untangle a blended score later.
Is an intent score the same as an engagement score?
Not quite. An engagement score usually counts what a person does with your own assets, such as page visits, form fills and email replies. An intent score is broader and can also include research activity on third-party sites and trigger events like funding or new hires. In tools like HubSpot, the engagement score is effectively the first-party part of intent, and many teams use the two terms loosely.
Is a fit score the same as an ICP score?
They overlap heavily. An ICP score rates a company against your ideal customer profile, which is the account-level core of a fit score. A full fit score often adds a person-level layer on top, checking whether the contact has the right role and seniority. So a strong ICP score with the wrong contact can still give you a weak overall fit for that specific lead.
Can you combine fit and intent into one score at all?
You can, as long as the parts stay visible. HubSpot's combined score, for example, keeps separate fit and engagement properties next to the total, and some teams use labels like A1 or C3 so the letter shows fit and the digit shows intent. The problem is a single total with nothing behind it, because then nobody can tell why an account ranked where it did.
How reliable is third-party intent data?
It is useful but noisy. Third-party intent data usually works at the account level, so it tells you a company is researching a topic, not which person is doing it or why. Research spikes can come from students, consultants or a team evaluating something unrelated. Treat it as an early hint that raises an account's priority, and let your own first-party signals decide when a rep actually reaches out.
What should sales do with a high-intent lead that is a poor fit?
Slow down before booking a full demo. Ask one or two qualifying questions, by email or a short call, to check budget, size and use case. If the answers rule them out, point them to a self-serve plan, a partner or useful content. If several of these leads turn into good customers, that is a sign your fit criteria are too narrow and need updating.
Do fit and intent scores apply to people or to companies?
Both, and B2B teams usually need both levels. Fit at the company level asks whether the account matches your ICP, while fit at the person level checks role and seniority. Intent works the same way: one person's pricing visit is a lead-level signal, while several people from one company researching your category is an account-level signal. Rolling person-level intent up to the account gives a truer picture.
Which tools calculate fit and intent scores?
Most CRMs and marketing platforms can. HubSpot builds fit, engagement and combined scores, and Salesforce Account Engagement pairs an activity score with a fit grade. Account platforms like 6sense report profile fit alongside a predicted buying stage, while providers like Bombora supply topic-level intent. Clay is often used to pull these inputs together, calculate custom scores and write them back to the CRM.
How do fit and intent scores relate to MQLs and MQAs?
They are usually the inputs that decide those stages. A marketing qualified lead or marketing qualified account is typically defined by crossing both a fit bar and an intent bar, not one combined threshold. Requiring both stops busy but poor-fit contacts from becoming MQLs, and keeps strong-fit accounts that show no activity from being handed to sales before they are ready to talk.
Share this post
Intent
Data
Prospecting
Glossary
Related glossary items
Understanding the core GTM concepts
Ready to build your pipeline
