Fit and intent scoring in 2026: a practical dual-score model (with examples for PLG, mid-market, and enterprise)

Your buyers do the research first. Your scoring model decides whether you show up at the right moment or show up late with "just bumping this." Gartner found that 61% of B2B buyers prefer a rep-free buying experience (gartner.com), and a follow-up survey put that figure at 67% in 2026 (gartner.com). That means your first touch has to be timed and earned, not sprayed.

The short version

• Run a dual score: fit (0-100) plus intent (0-100).
• Fit answers should we sell to them? Intent answers should we sell to them now?
• Add a next-action matrix so the rep or the agent stops guessing.
• Use the copy-paste templates below to ship a working fit and intent scoring model today. No data science, no "revops project."

Why single-score lead scoring breaks in 2026

A single score mixes two different truths:

• Fit: firmographics, stack, exclusions. "Can this account realistically buy?"
• Intent: behavior and change. "Are they in a buying window?"

Blend them and you get the classic failure:

• A perfect-fit account with zero intent becomes a fake "hot lead."
• A high-intent account outside your ICP hijacks the rep's week.

Email engagement also got less reliable as a primary signal. Bulk-sender rules pushed teams toward cleaner infrastructure and less tracking-heavy outreach. You still track engagement, you just treat it as a supporting signal, not gospel. Mailgun's breakdown of the Google and Yahoo bulk-sender requirements (mailgun.com) is a solid operational summary if your team missed the last two years.

The 2026 dual-score standard: fit (0-100) plus intent (0-100)

Here is the model:

• Fit score (0-100) is mostly static. Update it weekly or monthly.
• Intent score (0-100) is volatile. Update it in near real time and decay it fast.
• Priority is not "fit + intent." Priority is the action you take based on both.

This is the system Chronic runs end to end. Chronic is an autonomous revenue operator: it scores fit and intent, prioritizes the queue, writes and sends the outreach from warmed mailboxes, handles replies, and books the meeting, surfacing approvals only for the decisions that matter. The product version of this scoring engine is here: Chronic AI lead scoring.

Fit score template (0-100): copy, paste, ship

Fit scoring should feel strict. It should say "no" a lot.

Fit score structure

• Firmographics (0-45)
• Technographics (0-25)
• ICP exclusions and penalties (0 to -40)
• Role and buying-center match (0-10)
• Data confidence (0-10)

Cap the total at 0-100.

0-45: firmographics scoring (example weights)

Use what matters to your ICP. Here is a default B2B SaaS layout.

Company size (0-15)

• 1-10 employees: 2
• 11-50: 6
• 51-200: 12
• 201-1000: 15
• 1001-5000: 10
• 5000+: 6 (unless enterprise is your motion)

Revenue band (0-10)

• Unknown: 3
• <$5M: 2
• $5M-$25M: 6
• $25M-$100M: 10
• $100M+: 8

Industry match (0-10)

• Core ICP industries: 10
• Adjacent: 6
• Everything else: 2

Geo and compliance (0-10)

• Sells in your supported countries: 6
• Has required compliance profile (SOC 2, HIPAA, FINRA, etc.): +4 if relevant

0-25: technographics scoring

Technographics are fit accelerants, not personality tests.

BuiltWith tracks more than 100,000 web technologies across hundreds of millions of sites (blog.builtwith.com), which is why many teams use it as a starting point for technographic filters.

CRM present (0-10)

• Salesforce present: 10
• HubSpot present: 8
• Pipedrive/Zoho/Close/Attio: 6
• Unknown: 3
• "No CRM" signals: 0

Sales engagement / outbound stack present (0-5)

• Outreach/Salesloft/Instantly/Reply.io: 5
• None: 1

Data enrichment / intent tools present (0-5)

• ZoomInfo/Clearbit alternatives/Lusha: 3-5
• None: 1

Your "wedge" tech (0-5) Examples:

• You sell to companies running Webflow + Segment: +5
• You sell to teams on Shopify Plus: +5
• You sell to companies on Snowflake: +5

0 to -40: ICP exclusions and penalties

This is where scoring becomes useful.

Hard disqualifiers (set fit to 0, or apply -40)

• Competitor, partner conflict, reseller only
• Students, consultants, agencies (if not your ICP)
• Countries you cannot sell into
• Industries you will not support

Soft penalties (-5 to -20)

• Known churn-risk patterns for your business (example: "crypto exchange," "adult," "lead-gen spam shop")
• Unrealistic size band (too small or too large)
• No budget owner function exists (example: an IT tool sold to a 12-person company)

0-10: role and buying-center match

Score this at lead level even if fit is account-level.

• Target buying roles (VP Sales, RevOps, SDR leader): +10
• Adjacent roles: +5
• Everyone else: +1

If you need tighter definitions, build your ICP first, then score against it: Chronic ICP builder.

0-10: data confidence multiplier

Bad data creates fake precision.

• Verified domain + enriched company + verified contact: 10
• Missing one of those: 7
• Missing two or more: 4
• Sketchy free-email domain: 1

Fit score output fields (copy-paste)

Store these fields so scoring is not a black box:

• Fit score (0-100)
• Fit tier: A (80-100), B (60-79), C (40-59), D (<40)
• Top 3 fit reasons (text)
• Top 2 exclusions/risks (text)
• Data confidence (0-10)

If you want this to update itself instead of rotting in a spreadsheet, that is the whole point of letting an operator own enrichment. The deeper version of this argument is here: self-updating account data in 2026.

Intent score template (0-100): copy, paste, ship

Intent is a race. You either show up in the window or you do not.

Intent score structure

• Digital behavior (0-35)
• Company change signals (0-35)
• Tech install changes (0-15)
• Outbound engagement (0-15)

Then apply:

• Decay (time-based)
• Spam-safety rules (quality gating)

0-35: digital behavior intent (your properties, your site)

Pricing page and high-commercial pages (0-15)

• Pricing page view: +8
• Visited pricing twice in 7 days: +12 total
• Visited integration docs, security page, ROI page: +5 each (cap 15)

Product usage for PLG (0-12)

• Activated core event (your "aha" moment): +8
• Invited teammate: +4
• Hit a usage limit: +6 Cap at 12.

Comparison intent (0-8)

• Viewed a /vs/competitor page: +8
• Viewed two competitor comparisons in 14 days: still cap 8, do not double count

0-35: company change signals intent

This is where "timing" actually lives.

Job changes (0-15)

• Known champion changes jobs into an ICP account: +15. UserGems has published analysis showing that opportunities involving a previous champion close at materially higher rates (usergems.com). Treat champion moves as a top-tier signal.
• New VP Sales / RevOps hire (net-new leader): +10
• SDR Manager hire: +6

Hiring signals (0-10)

• Hiring for roles that imply your category problem (example: "Sales Operations Analyst," "Outbound SDR," "RevOps Systems"): +6
• Hiring spike in the relevant department (3+ roles in 30 days): +10

Funding, expansion, procurement motion (0-10)

• Funding round announced: +10
• New region expansion press release: +6
• "RFP" / procurement job post: +8

0-15: tech install change intent

These are "budget moved" signals.

• Added Salesforce or HubSpot recently: +8
• Added a sales engagement tool: +6
• Installed your competitor: +10 (they are buying, you are just late)
• Removed your competitor: +12 (switch window)

0-15: outbound engagement intent (email, replies, meetings)

Email engagement is noisy. Replies are not.

Replies (0-15)

• Positive reply (asks a question, asks for pricing, asks for a call): +15
• Neutral reply ("not now," "circle back Q3"): +8
• Referral reply ("talk to Jane"): +12
• Negative reply ("stop emailing me"): -25 and suppress

Clicks (0-3)

• Clicked a link: +3. Keep this low; click tracking gets messy fast.

Opens (0-1)

• Open: +1. Opens are a rounding error in 2026.

Reply classification rules (copy-paste)

This is the simplest usable set:

• Positive: asks for a demo, asks for details, shares pain, loops in a buyer
• Neutral: not now, wrong person but gives guidance, timing note
• Negative: remove, unsubscribe, hostile, threat
• Auto: out of office, bounce, automated firewall notice

Intent decay (non-negotiable)

Intent expires, fast. Apply this decay to the whole intent score:

• 0-3 days since last signal: 1.0x
• 4-7 days: 0.8x
• 8-14 days: 0.6x
• 15-30 days: 0.4x
• 31-60 days: 0.2x
• 60+ days: 0.1x

This one change stops your queue from turning into a graveyard.

The next-action matrix (this is where teams win)

Scoring without actions is decoration. Here is a practical matrix.

Define tiers

Fit tiers

• A: 80-100
• B: 60-79
• C: 40-59
• D: <40

Intent tiers

• Hot: 70-100
• Warm: 40-69
• Cold: 0-39

Next-action matrix (agent or rep)

Fit \ Intent	Hot (70-100)	Warm (40-69)	Cold (0-39)
A Fit	Book a meeting in 24h. Multi-channel. Personalized opener from signals.	7-touch sequence over 14 days. Add one strong proof point.	Light nurture. Add to signal watch. No heavy outbound.
B Fit	Book a meeting in 48h. Email + LinkedIn.	Standard outbound sequence over 21 days.	Nurture only. Re-score quarterly.
C Fit	Qualify fast. 1-2 touches max. If no traction, drop.	Do not sequence. Wait for an intent spike.	Ignore.
D Fit	Suppress. Do not waste cycles.	Suppress.	Suppress.

"Book a meeting" playbook (copy-paste)

When fit is A and intent is hot:

1. Touch 1 (same day): short email, 80 words max, one signal, one question
2. Touch 2 (next day): follow up with a concrete resource (security, ROI, integration)
3. Touch 3 (day 3): call if you have a direct line, voicemail optional
4. Touch 4 (day 5): "close the loop" message with two time slots

Chronic runs this prioritization and sequencing on its own, all the way to the booked meeting, and asks you to approve before it does anything that touches your reputation. See: Chronic sales pipeline.

A working fit-and-intent spreadsheet template

Create a sheet with these columns.

Account columns (fit)

• Company
• Domain
• Employees
• Revenue
• Industry
• Geo
• CRM detected
• Key tech detected
• Exclusion flags
• Fit score (0-100)
• Fit tier
• Fit reasons (text)

Lead columns (intent)

• Contact
• Title
• Email
• Last signal date
• Signal types (multi-select)
• Intent score raw
• Intent decay multiplier
• Intent score final
• Intent tier
• Reply classification (if any)
• Next action (auto-generated)

If you would rather not maintain the enrichment by hand, that is the kind of work the operator does for you: lead enrichment.

Worked example 1: PLG inbound assist (high intent, mixed fit)

Scenario

• Product: PLG SaaS
• Motion: self-serve signups, sales assists on expansion
• Account: 120 employees, US, SaaS
• Contact: Senior Manager, Ops
• Signals in the last 48 hours:
  • Hit a usage limit
  • Invited 2 teammates
  • Viewed pricing and the security page

Fit score calculation

• Company size (51-200): 12
• Revenue unknown: 3
• Industry (SaaS core): 10
• Geo supported: 6
• Compliance unknown: 0

Firmographics subtotal: 31/45

Technographics:

• HubSpot present: 8
• Sales engagement unknown: 1
• Data tools unknown: 1
• Wedge tech match: 2

Technographics subtotal: 12/25

Role match:

• Ops manager (adjacent): 5/10

Data confidence:

• Verified domain and contact: 7/10

Exclusions: none

Fit score = 31 + 12 + 5 + 7 = 55 (C/B edge)

Intent score calculation

Digital behavior:

• Pricing page: +8
• Security page: +5
• Hit a usage limit: +6
• Invited teammates: +4

Subtotal: 23/35

Company change: none. Tech change: none. Outbound engagement: none.

Raw intent: 23. Decay multiplier (0-3 days): 1.0x. Intent score final = 23 (cold)

That feels wrong, and it is. Here is the PLG nuance: product usage should weigh heavier than generic web pages. So adjust the PLG weights:

• Product usage bucket becomes 0-35
• Website bucket becomes 0-15

Re-score quickly:

• Usage limit +6, invited teammates +4, activated core event (assume yes) +8, repeated active days +6, added a second workspace +6 (example). Now you are above 30.
• Pricing +6, security +4.

Raw intent: 40. Intent = 40 (warm)

Next action

Fit C/B edge + intent warm:

• Action: inbound-assist qualification. Send a human, in-product-style email. One question, then route.
• If they confirm a budget owner exists, upgrade fit, then sequence.

Chronic can run this on its own: score the product events, score fit, then prioritize. The scoring engine lives here: AI lead scoring.

Worked example 2: mid-market outbound (high fit, timing spike)

Scenario

• ICP: 200-1000 employee B2B SaaS
• Account: 650 employees, Series C, hiring 6 SDRs
• Stack: Salesforce detected, Outreach detected
• Signals:
  • New VP Sales started this week
  • Hiring an SDR team
  • Visited your "Salesforce integration" page from a branded search

Fit score

Firmographics:

• Size (201-1000): 15
• Revenue (assume $25M-$100M): 10
• Industry core: 10
• Geo supported: 6
• Compliance: 2

Subtotal: 43/45

Technographics:

• Salesforce: 10
• Outreach: 5
• Data tools: 3
• Wedge tech: 3

Subtotal: 21/25

Role match (you target VP Sales and RevOps):

• Contact is RevOps Director: 10/10

Data confidence:

• Verified: 10/10

Exclusions: none

Fit = 43 + 21 + 10 + 10 = 84 (A)

Intent score

Digital behavior:

• Integration page: +5
• Branded search arrival: +6

Subtotal: 11/35

Company change:

• New VP Sales: +10
• Hiring SDR surge: +10

Subtotal: 20/35

Tech change: none (already on Salesforce and Outreach). Outbound engagement: none yet.

Raw intent: 31. Decay: 1.0x. Intent = 31 (cold/warm edge)

Now add the reality: multiple signals inside 7 days compound. Create a "signal stack bonus":

• 3+ different signal categories in 7 days: +15
• 2 categories: +8

This account has company change (exec hire), hiring, and digital behavior. That is 3 categories.

Intent = 31 + 15 = 46 (warm)

Next action

Fit A + intent warm:

• Action: 14-day outbound sequence.
• Personalize the opener to the VP Sales start and the SDR hiring.
• Skip "congrats on the new role." Everyone does that. Say what the role forces: pipeline math, tooling, process.

An operator can write this faster and to your voice. Chronic's writer is here: AI email writer.

Do not blow deliverability by blasting five domains with tracked links. If you need the 2026 infrastructure baseline, this is the SOP-style breakdown: cold email infrastructure in 2026.

Worked example 3: enterprise ABM-lite (high fit, low intent, do not spam)

Scenario

• ICP: enterprise, regulated
• Account: 12,000 employees, banking
• Stack: Salesforce
• Signals:
  • 1 anonymous visit to pricing
  • 1 G2 review read (if you can track it)
  • No champion, no hiring spike, no tech change

Fit score

Firmographics:

• Size 5000+: 6 (unless you weight enterprise higher, and you should)
• Revenue $100M+: 8
• Industry core: 10
• Geo supported: 6
• Compliance match: 4

Subtotal: 34/45

Technographics:

• Salesforce: 10
• Other: 6

Subtotal: 16/25

Role match:

• Unknown contact yet: 1/10

Data confidence:

• Medium: 7/10

Exclusions: none

Fit = 34 + 16 + 1 + 7 = 58 (C)

This is a common enterprise trap. Great logo, weak targeting. If enterprise is your ICP, change the firmographic weights so 5000+ employees scores higher. You do not "fix" this with intent.

Intent score

• Pricing visit: +8
• Everything else: 0

Raw: 8. Decay: 1.0x. Intent: 8 (cold)

Next action

Fit C + intent cold:

• Action: ABM-lite watchlist. No sequence, no spray. Build the account, map the buying group.
• Trigger outreach only if:
  • an exec is hired in your function
  • a champion changes jobs into the account
  • a competitor install is detected
  • there are multiple visits to integration/security pages in 7 days

This is where an autonomous operator earns its keep. It watches the account quietly and fires when the window opens, without adding another dashboard for you to babysit: AI lead scoring and sales pipeline.

Implementation: ship this in 48 hours

Day 1: define ICP and exclusions (one page)

• 3 must-have firmographics
• 3 must-have technographics
• 5 exclusions
• 5 "yellow flags"

Then codify it in a builder: Chronic ICP builder.

Day 2: instrument intent signals

Minimum viable intent:

• Pricing, security, integration pages
• Activation and limit-hit product events (PLG)
• Job changes (champions and execs)
• Hiring spikes (role-specific)
• Tech installs and removals

Day 3: wire next action into the queue

If the output is not a single field like NEXT_ACTION = SEQUENCE_14D, it will not run. If your reps still triage by hand, you built a reporting system, not a pipeline system.

For a clean prioritization concept, this post lays out the "queue" idea: the modern SDR queue: fit + intent + timing.

Common mistakes that wreck dual scoring

Mistake 1: treating "fit" as "has money"

Money does not mean they buy your thing. Fit means the problem exists and procurement is plausible.

Mistake 2: treating opens as intent

Opens are a liar. Replies are the truth. Site paths are the hint.

Mistake 3: no decay

If intent does not decay, your queue becomes a museum.

Mistake 4: no suppression rules

Negative replies must suppress instantly. No "maybe we try again next month."

Mistake 5: scoring without enrichment

If half your accounts have missing employee counts and unknown stacks, your model becomes vibes. Fix the inputs: lead enrichment.

FAQ

What is a fit and intent scoring model?

A fit and intent scoring model runs two separate 0-100 scores. Fit measures whether the account matches your ICP. Intent measures whether the account is in a buying window right now. The model outputs a next action so reps or an agent execute without manual triage.

What is a good threshold for "hot" intent in 2026?

Start with 70+ as "hot" if you use multiple signal categories plus decay. If you rely mostly on email engagement, raise the threshold, because that data is noisy. The real rule is speed: act within days, not weeks.

How do we handle PLG differently from outbound?

PLG intent should weight product events more than pageviews. "Invited teammate," "hit usage limit," and "reached activation milestone" should dominate the intent score. Pageviews are supporting evidence.

Should we buy third-party intent data?

Only if you already do the basics well. First-party intent (site and product) plus public change signals (job changes, hiring, tech changes) usually gets you most of the value. Third-party intent often adds noise unless you have strong filtering and fast activation.

How often should we refresh fit and intent scores?

• Fit: weekly or monthly, plus updates whenever enrichment changes.
• Intent: near real time, with a time-decay multiplier applied daily.

How does Chronic apply dual scoring in real workflows?

Chronic is an autonomous revenue operator. It scores fit and intent, prioritizes the queue, writes personalized outreach, sends it from warmed mailboxes, and pushes sequences forward to the booked meeting, asking you to approve the decisions that matter. Start with AI lead scoring, then connect lead enrichment.

Put it into production this week

• Copy the fit template. Add your exclusions. Score 200 accounts.
• Copy the intent template. Add decay. Score daily.
• Implement the next-action matrix as a single field that drives sequencing.
• Stop debating lead quality in Slack. The model decides, the operator executes.

If you want the whole loop run for you, from scoring to the booked meeting, that is the job Chronic is built to do: you set the goal and the guardrails, and it handles discovery, outreach, replies, and scheduling, surfacing approvals only when a human should weigh in.