How Do You Design a Fit Score in HubSpot Step by Step?

Direct answer: A fit score in HubSpot is built by first defining your ideal customer profile with sales, then creating a dedicated Fit score object in HubSpot's Lead Scoring tool using property based rules, assigning weighted positive and negative points to firmographic attributes, setting thresholds and testing against real records, and activating the score in workflows combined with an engagement score.

Why does fit scoring in HubSpot require a new approach now?

HubSpot retired the old blended score property throughout 2025, replacing it with a Lead Scoring tool that separates fit (who a lead is) from engagement (what a lead does).

For years, startups relied on a single HubSpot Score property that mixed identity with behavior. That model is now obsolete. According to iodigital, HubSpot began retiring this model in 2025: you could not create new score properties starting May 1, 2025, could not edit existing ones starting July 1, 2025, and existing score properties stopped updating starting August 31, 2025. Legacy scores not used in an active asset are also being periodically deleted from Q4 2025 onward.

The replacement, HubSpot's Lead Scoring tool, deliberately splits two distinct questions. Fit measures who they are (company size, industry, job title, revenue), while engagement measures what they do. As one analysis puts it, mixing these dimensions makes it impossible to explain why a lead qualified. For a founder running a lean GTM stack, this split forces you to codify your ICP once, cleanly, instead of guessing which combination of clicks and titles equals a good customer.

How do you define ICP criteria before opening HubSpot?

Get sales and marketing aligned on explicit, verifiable firmographic signals before touching the tool, since this strategy work determines whether the fit score succeeds.

This is not a HubSpot task, it is a strategy task, and skipping it is the single biggest reason fit scores fail. Guidance from Weidert is explicit that scoring criteria should come from numerical scores assigned to attributes chosen collaboratively by marketing and sales.

Build a short list of firmographic signals:

• Company size
• Industry vertical
• Job title or seniority
• Annual revenue
• Geography
• Tech stack

Separately, list disqualifying traits such as competitors, students or job seekers, unsupported geographies, or company sizes below your minimum viable deal size. As one migration guide notes, this migration forces the ICP definition work you've deferred, referenced in analysis from Simple Machines Marketing.

How do you create the fit score object and choose its scope in HubSpot?

Navigate to Marketing > Lead Scoring, choose whether you're scoring contacts, companies, or deals, pick fit as the score type, and set a score ceiling like 0 to 100.

Fit scoring lives inside the dedicated Lead Scoring tool, not the old Properties panel. In your HubSpot account, navigate to Marketing > Lead Scoring, then create or edit a score.

HubSpot asks who and how: who means contacts, companies, or deals; how means engagement, fit, or combine. For a startup, contact-level fit scoring is usually the right starting point since it captures the individual buyer's role, while company-level fit scoring matters more when selling to multi-stakeholder accounts.

Subscription gating matters here, per HubSpot's knowledge base:

• Contacts on Marketing Hub only can get engagement, fit, or combined scores
• Companies on Marketing Hub or Sales Hub can get the same three types
• Deals on Sales Hub only can get combined engagement and fit scores

Fit scoring specifically requires Marketing Hub or Sales Hub Professional or Enterprise. Also set your score ceiling: HubSpot's Score limit dropdown offers ranges from -100 to 100 up to -10,000 to 10,000. Most lean teams should keep it simple with a 0 to 100 scale so thresholds stay intuitive for reps unfamiliar with the tool.

How do you build property groups with positive and negative rules?

Fit scores are built exclusively from property based rules on contact, company, or associated records, with positive points for ICP traits and negative points for disqualifiers.

Unlike engagement scores, which are built from events, fit scores use only property groups. Per HubSpot's documentation, for fit scores you'll only have the option to add property groups.

HubSpot also supports pulling data from associated records, which matters for B2B startups selling to teams: click Add property criteria, then select Contact or Company, or Associated object for a related record's property.

Build both sides of the ledger:

• Positive points: target industry, right-sized company, senior title.
• Negative points: competitors and job seekers identified by specific email domains, per an example cited by Hubjoy.

Weight criteria based on actual win-rate correlation rather than spreading points evenly. If gated content downloaders convert at a notably higher rate, apply heavier weight to that signal since it closely correlates to revenue outcomes.

Importantly, fit scores should not decay over time the way engagement scores do, since fit describes identity, not activity: apply decay to engagement only, as fit scores are static.

How do you set thresholds and test the fit score before turning it on?

Translate the raw score into High, Medium, and Low labels, then validate the logic against real closed-won and closed-lost records before activating it live.

A raw number means little to a rep triaging pipeline. HubSpot's threshold feature categorizes records by label, for example setting 0-49 as Low, 50-74 as Medium, and 75-100 as High, per HubSpot's release notes.

Before flipping the switch, validate against real records. HubSpot lets you click "Test contact" to evaluate your score property. One implementation guide reinforces the same discipline: run your score logic against a few examples to confirm top prospects aren't underweighted and low-quality leads aren't scoring too high, per Man Digital.

For startups with thin data, this manual sanity check often catches errors, like a title filter that accidentally excludes your best customer segment, before it reaches a rep's queue.

How do you activate the fit score and combine it with engagement?

Connect the fit score to a workflow that triggers alerts or tasks at your threshold, then pair it with an engagement score to build a routing matrix that prevents mismatched leads from reaching AEs.

A fit score only creates value once it drives action. Connect it to workflows so records crossing your threshold trigger something concrete, such as choosing a start trigger based on lead score above 50, then adding actions like sending an internal notification or creating a task for sales.

For most startups, the highest leverage move is pairing fit with an engagement score into a Combined score. One framework maps this as a 3x3 grid: fit A/B/C times engagement 1/2/3 equals 9 routing categories, described by Thalox.

This grid resolves the classic ambiguous case: medium fit plus high engagement creates ambiguity, since the prospect shows interest but doesn't match ICP, so route to SDR for qualification rather than direct to AEs, because high activity doesn't compensate for wrong fit.

Finally, treat the score as a living system. Revisit it as your ICP sharpens, because lead scoring tools don't solve scoring problems, process design does, and the best model fails if sales doesn't trust it and marketing doesn't maintain it.

Frequently asked questions