Dual scoring that actually books meetings: fit + intent, with a stop-sending rule

Most teams treat scoring like a New Year's resolution. They talk about it. They build a spreadsheet. Nothing changes, and pipeline stays flat.

Dual scoring fixes that, but only if you ship the part everyone avoids: rules that change actions. Who gets emailed. When. With what message. And when to stop sending.

This guide shows how to build fit and intent scoring, then wire it into outbound with a hard stop-sending rule. At the end, it shows how Chronic, an autonomous revenue operator, runs the whole motion for you instead of leaving you to assemble it by hand.

1) Fit and intent in one sentence

Fit and intent scoring is a two-part model that ranks accounts by:

1. ICP fit: how likely an account is to succeed with your product.
2. Buying intent: how likely that account is to buy right now.

Most teams mash these into one "lead score" and then wonder why the top "A+ leads" never reply. Fit and intent are different forces. Treat them separately so your outbound can do different things:

• High fit + low intent: light touch, value-first, longer window.
• Low fit + high intent: tempting, but usually a time sink.
• High fit + high intent: all gas. Best message, fast follow-up, meeting priority.

2) Build your fit model (ICP) before you touch intent

Intent without fit is how you end up chasing companies that read one article once.

Pick 6 to 10 fit inputs that actually matter

Use inputs you can explain to a rep in 15 seconds. If you need a data scientist, you already lost.

Firmographics

• Industry
• Employee range
• Revenue range (optional, often messy)
• Geography (if you have constraints)

Technographics

• Tools you integrate with
• Tools you replace
• Stack signals that they run modern GTM, or the opposite

Structural constraints

• Security or compliance requirements you cannot meet
• Segment exclusions (education, government, and so on)

Fit-adjacent triggers

• Recently funded
• New exec hire (sales, marketing, revops)
• Hiring velocity in relevant teams

If you want an ICP that does not rot in a doc, start here: Chronic ICP Builder.

Set your hard-no filters first

Gating comes before scoring. Examples of a hard no:

• Under 10 employees (no budget, no process)
• A regulated segment you cannot serve
• Competitor tech that blocks your product entirely

Hard no means score = 0, never enroll.

Assign fit points, and keep it boring

Start with a 0 to 100 fit score. You can get fancy later, but later never comes, so ship boring now.

• Industry match: 0 to 25
• Company size band: 0 to 25
• Tech stack match: 0 to 25
• Trigger strength: 0 to 15
• Geo and constraints match: 0 to 10

Your weights should reflect your real close rates. If you do not know them yet, set weights on operator judgment and plan to recalibrate monthly.

Fit scoring collapses without clean inputs. Chronic Lead Enrichment keeps them fresh.

3) Build your intent model (timing) from signals you can actually capture

Intent is not "they exist." Intent is behavior, change, or motion. It is also noisy: one signal lies, multiple signals converge. Gartner calls third-party intent fundamental for improving outbound effectiveness, but stresses combining multiple signals to find real buying activity (gartner.com).

Category A, website behavior (first-party)

• Pricing page view
• Integration or docs view
• Competitor comparison page view
• Multiple visits within 7 days
• Time-on-page thresholds (use carefully)

Category B, org changes (trigger intent)

• New VP of Sales, RevOps, or Marketing Ops
• Promotions (manager to director, director to VP)
• Hiring spree (SDR, AE, or RevOps roles)

Category C, tech installs (technographic intent)

• A new tool installed that correlates with your use case
• Tool churn (they removed a competitor)
• A new data provider, sales engagement platform, or CRM migration

Category D, paid and content engagement

• Retargeting ad clicks from target accounts
• Webinar signup from ICP accounts
• High-intent content ("migration," "security," "pricing," "RFP template")

Category E, email engagement (treat as weak intent)

• Reply beats click beats open, in that order
• Multiple clicks on specific assets
• Forwarding signals (rare, but gold)
• Negative signals: spam complaints, repeated no-engagement

4) The dual scoring table you can copy today

A simple model that works without a PhD.

Fit score (0 to 100)

Fit input	Criteria	Points
Industry	Core ICP industry	+20
Industry	Adjacent industry	+10
Company size	In sweet spot (e.g. 50 to 500)	+20
Company size	Outside but plausible	+10
Stack match	Uses key integration tool	+15
Stack match	Uses competitor	+10
Constraint	Disqualifying segment	0 (hard no)
Trigger	Funding in last 12 months	+10
Trigger	New RevOps or Sales leader	+15
Geography	In supported regions	+5

Intent score (0 to 100)

Intent input	Criteria	Points
Website	Pricing page visit (last 14 days)	+25
Website	Integration or docs visit	+15
Website	3+ visits in 7 days	+20
Org change	New VP+ in a function you sell to (last 60 days)	+20
Tech	Competitor installed or removed	+15
Ads/content	Webinar or demo request	+40
Email	Reply (any)	+50
Email	Click on a high-intent link	+10
Negative	"Not a fit" reply	-100 (stop)
Negative	Unsubscribe or spam complaint	-100 (stop)

Two notes that decide whether this works:

• Intent decays. Website intent from 90 days ago is a ghost. Put time windows on points.
• Replies override scoring. A human response beats your model every time.

5) Turn dual scoring into actions (the part that books meetings)

Scoring that does not change routing is analytics cosplay. You need a ruleset for sequence enrollment, message type, channel choice, follow-up priority, and when to stop sending.

Create four segments using fit + intent thresholds

Keep the thresholds simple. A good starting point is high fit = 70+ and high intent = 60+. That gives you four boxes:

1. High fit + high intent (HFHI)
2. High fit + low intent (HFLI)
3. Low fit + high intent (LFHI)
4. Low fit + low intent (LFLI)

Enrollment rules, simple and ruthless

HFHI: enroll immediately

• Start the sequence within minutes of the signal where you can
• Use your strongest, most specific (trigger-based) message
• Add a call task if your motion uses the phone

HFLI: enroll in a slower, value-first sequence

• Longer spacing, lower personalization cost
• Educational angle
• Goal is to create engagement signals, not to book a meeting tomorrow

LFHI: conditional enroll

• Enroll only if the intent signal is strong and recent (a demo request, or a pricing view plus a docs view)
• Otherwise route to marketing nurture or ignore
• This box is where teams waste hours chasing mirages

LFLI: never enroll

• Keep it on a watchlist, re-score weekly
• Do not burn your domain on deadwood

If you want this to run inside one system instead of a spreadsheet, Chronic runs dual scoring and prioritization here: Chronic AI Lead Scoring.

6) Message rules: what you say changes with fit and intent

HFHI: mirror the signal, ask for the meeting

You earned the right to be direct.

1. Call out the signal, without being creepy
2. Tie it to a likely priority
3. Make a tight meeting ask with a clear outcome

Example:

• "Saw your team is hiring SDRs and ramping outbound."
• "That usually means lists, deliverability, and routing get messy fast."
• "Worth 12 minutes to map your scoring and stop-sending rules so reps only touch accounts that can buy this month?"

HFLI: teach, then trigger

No fake urgency. That is how you earn spam clicks.

1. Call out fit (industry plus role)
2. Offer a useful pattern
3. Ask a low-friction question

Example:

• "Most Series B SaaS teams hit the same wall: volume dies, targeting gets sloppy."
• "We run fit and intent scoring with a stop-sending rule so outbound does not torch the domain."
• "Who owns scoring and routing today, RevOps or SDR leadership?"

For personalization patterns that do not read like a generic compliment, use this: Personalization that wins in 2026.

7) The stop-sending rule (non-negotiable)

Teams love to "always follow up." Mailbox providers punish them for it. Google's bulk sender guidance is explicit: keep spam complaint rates below 0.3% in Postmaster Tools, and offer one-click unsubscribe for bulk mail (support.google.com). That makes a stop-sending rule survival, not a nicety.

Your rule needs two layers: contact-level stops that respect the person, and domain-level stops that protect deliverability.

Contact-level stop rules

Stop sending immediately when any of these happen:

• Unsubscribe or spam complaint
• "Not interested," "remove me," or any explicit ask to stop
• A hard bounce
• "We already use X" and you are not competing for displacement
• "Wrong person" with no referral and no reliable way to reroute

Also stop after a no-engagement cap. For example: 6 touches across 21 days with zero clicks and zero replies. Move that contact to a 60 to 90 day recycle, then re-enter only if new intent appears.

Domain-level stop rules

These pause campaigns when deliverability risk spikes:

• Complaint rate approaching provider thresholds (watch Gmail Postmaster if you have volume)
• Sudden bounce spikes (a list-quality problem)
• Open rates collapsing across mailboxes (a reputation hit or filtering change)

If you do not have a weekly deliverability routine, fix that before you crank volume: Cold email deliverability ops in 2026. The plain truth is that the stop-sending rule is how you keep sending next month.

8) Reply handling and meeting-booking priority

Scoring decides who enters. Replies decide everything after.

Reply classification

Every reply becomes one of these:

1. Positive (meeting interest)
2. Objection (timing, budget, authority)
3. Not now (future)
4. Not a fit (stop)
5. Wrong person (reroute)
6. Out of office (pause then retry)

Priority order

When you run volume, speed is a weapon. Handle replies in this order:

1. Positive reply from HFHI
2. Positive reply from HFLI
3. Objection from HFHI
4. Positive reply from LFHI
5. Everything else

If your team answers "maybe later" faster than "yes," you deserve the quarter you get.

Scheduling

• Offer two specific times, then a calendar link.
• Confirm the agenda in one line.
• Log the activity and tag the original fit and intent scores, so you can learn from what booked.

9) Make it run without hero reps

Set a scoring cadence

• Recompute intent daily if you have live signals.
• Recompute fit weekly or monthly; fit shifts slower.
• Review thresholds weekly for the first 4 weeks, then monthly.

Track the metrics that matter

Forget "leads scored." Track:

• Meetings booked per 100 enrolled, by segment
• Reply rate and positive reply rate, by segment
• Time to first touch after an intent spike (HFHI should be minutes, not days)
• Unsubscribe and complaint rates, by segment and by sequence

Relevance is the whole game. Forrester reports that 86% of B2B purchases stall and 81% of buyers are dissatisfied with the provider they choose (forrester.com). Buyers are overwhelmed and picky, so targeting beats volume.

Recalibrate weights using booked meetings, not vibes

Every 30 days:

• Pull booked meetings by segment.
• Find which fit inputs correlate with bookings.
• Find which intent inputs correlate with replies and bookings.
• Cut points for vanity intent (random blog visits, opens).
• Add points for commitment intent (pricing, docs, demo, stakeholder activity).

10) Where Chronic fits

You can build all of this with a stack of tools and a RevOps specialist to keep it running. Or you can hand the goal to an autonomous revenue operator and let it run the motion for you.

Chronic does the work end to end:

• Defines your ICP and routing with ICP Builder
• Keeps account data clean with Lead Enrichment
• Scores accounts on fit and intent with AI Lead Scoring
• Writes the sequences with AI Email Writer
• Sends from managed, warmed mailboxes, handles replies, and books the meeting inside your Sales Pipeline

The stop-sending rule is built into how it operates, so it protects your domain by default and surfaces only the decisions that need you.

Other tools cover pieces of this:

• Apollo runs data and engagement, but you still stitch the logic and governance (Chronic vs Apollo).
• HubSpot runs CRM and sequences, but dual scoring logic tends to turn into custom-property soup (Chronic vs HubSpot).
• Salesforce runs anything if you pay enough and hire admins to keep it tuned (Chronic vs Salesforce).

Chronic runs the whole motion until the meeting is booked.

FAQ

What is the difference between fit scoring and intent scoring?

Fit scoring measures whether an account matches your ICP and can succeed with what you sell. Intent scoring measures whether that account shows signs of active buying motion right now. Fit stays stable. Intent spikes and decays.

What is a good starting threshold for high fit and high intent?

Start with high fit = 70+ and high intent = 60+ on a 0 to 100 scale. Recalibrate after 30 days using booked meetings per segment, and keep thresholds stable long enough to learn from them.

Which intent signals matter most for outbound?

The ones closest to a purchase: pricing page views, integration and docs views, demo requests, and org changes in buying roles. Email opens are weak. Replies are the strongest signal of all.

How do I add a stop-sending rule without killing pipeline?

You stop sending to people who do not engage and people who opt out. That improves deliverability, which raises inbox placement, which means more replies from the right accounts. You replace "more follow-ups" with better targeting and timing.

How often should we update the scoring model?

Recompute intent daily if signals change quickly, and fit weekly or monthly. Recalibrate the weights monthly for the first quarter, then quarterly once the model is stable.

What is the biggest mistake teams make with fit and intent scoring?

They score leads but do not change actions. No routing, no sequence differences, no priority rules, no stop-sending rule. They end up with a pretty dashboard and the same outbound that was not working last month.

Build it this week, then enforce it

1. Ship a basic fit model (0 to 100) with 6 to 10 inputs.
2. Ship a basic intent model (0 to 100) with time windows.
3. Create the four-box routing (HFHI, HFLI, LFHI, LFLI).
4. Enforce enrollment rules and reply priority.
5. Enforce the stop-sending rule like your domain depends on it, because it does.
6. Recalibrate monthly using booked meetings, not feelings.