This applies to B2B teams running visitor identification with a defined ICP and a CRM that records account ownership. The tiers and windows are practitioner rules, not measured benchmarks, so set the window to your own sales cycle and review false positives against pipeline.
What Revenue Teams Consider An Actionable Visitor Signal
Revenue teams treat a website visit as actionable only when an ICP-fit account or person views a buying page such as pricing or a competitor comparison, comes back within days, and a named owner is set to respond, because 94% of buying groups rank preferred vendors before first contact. Short of that, website visitor identification output is raw traffic.
Each test removes a different kind of noise:
- Fit: The company, and ideally the person, matches the ideal customer profile. A perfect page path from a company you cannot sell to is still noise.
- Page context: The visit lands on a page that shows evaluation, such as pricing, a competitor comparison or a demo page. Pocus's playbook treats careers-page and login visits as low intent.
- Repetition: The behaviour comes back inside a set window. One visit says little, and a return within days says the account is still looking.
- Ownership: A named person is set to respond, with a response time. Without an owner, a strong signal sits in a dashboard.
Raw traffic passes none of these tests by default. A visit count, a list of identified companies or a spike in sessions tells a rep that something happened, not what to do next.
The bar matters because the shortlist forms before sellers are involved. In 6sense's survey of more than 4,000 buyers, the vendor preferred before buyers engaged sellers went on to win 80% of deals. The same survey found the split between independent research and seller engagement moved from 70/30 to 60/40, so buyers still spend most of the journey on their own.
A visitor signal is one of the few views a seller gets into that independent stretch. It only helps if it reaches a rep while the shortlist is still forming. Common Room's own playbook calls visits from high-fit accounts on high-intent pages an excellent buying signal, and Heinz Marketing's checklist ties every trigger to a documented response. The four tests above are where those views meet.
How This Guide Separates Expert Opinion From Observed Facts
This guide labels every point as either observed data, such as Forrester's survey of nearly 18,000 business buyers, or a named practitioner's published rule, such as the visit tiers Knock2 and Pocus set out, so readers can see which thresholds are evidence and which are opinion.
The evidence falls into three groups:
- Observed data: Buyer surveys from 6sense and Forrester, and Workato's response-time test, in which researchers submitted demo requests to 114 B2B companies.
- Practitioner opinion: Published playbooks from Pocus (Alexa Grabell, co-founder and CEO), Heinz Marketing (Sarah Threet), Knock2 and Vector's VP of Marketing, Jess Cook.
- Tool documentation: The scoring and routing documentation for Warmly, Vector, Common Room, Salesforce and Gong.
Which Website Visitor Signals Show Buying Intent
Pocus's playbook rates visits from an ICP to the pricing page or a competitor comparison page as high intent and careers-page or login visits as low intent, which gives revenue teams a page-level filter before any score is applied. The table sets out Pocus's full list, which is practitioner opinion rather than measured data.
High- and low-intent website signals in Pocus's playbook (practitioner opinion)
| Signal | Why it counts or not | Typical response |
|---|---|---|
| ICP visit to pricing page | High intent, buying context | Sales if decision-maker |
| ICP visit to competitor comparison | High intent, weighing options | Sales if decision-maker |
| Target account surge in a short window | High intent, several people active | Account owner review |
| Rising ICP engagement with webinars, white papers and blog posts | High intent, earlier stage | Route by visitor role |
| ICP visit to careers page | Low intent, likely job seeker | Leave out of alerts |
| Visit from a team the product does not serve | Low intent, no fit | Leave out of alerts |
| Homepage visit to log in | Low intent, existing user | Leave out of alerts |
Page context comes before volume. Ten visits to blog posts and a careers page say less than one return to pricing, and Knock2's framework treats any return visit to pricing or demo as its top tier on its own.
The page filter also decides whose data you trust. Vector's VP of Marketing, Jess Cook, argues that third-party intent only shows that someone at a company looked at a topic, and that it "doesn't tell you a real buyer is in-market." Her point about shared data is sharper. If every seller who buys the same feed sees the same topic surge, the surge is not an edge.
That view comes from a vendor that sells first-party, contact-level data, so read it as Vector's stated position. The logic still holds for any team. Behaviour on your own pricing, comparison and demo pages is seen by you alone, while a topic surge is seen by every competitor paying for it.
A practical page filter has three lists:
- Buying pages: Pricing, competitor comparison, demo and integration pages that show evaluation.
- Context pages: Webinars, white papers and blog posts, which count only when engagement rises at an ICP account.
- Excluded pages: Careers, login and support pages, plus any visit from a team the product does not serve.
Person-Level Or Account-Level Signal, Which Should Sales Act On
Forrester's survey found that the average buying decision involves 22 people, so a named visitor is a starting point for the account, not proof that the buyer has raised a hand. Forrester counts 13 internal stakeholders and nine external participants in that group, with larger groups for expensive or complex purchases.
The case for the person is practical. Common Room's playbook puts it bluntly, saying "reps can't outbound a logo. They need contact-level visibility." Jess Cook gives a concrete example. A director at a target account who visited pricing, clicked an ad and came back three days later is worth more to her than an account-wide surge. The difference between company-level versus person-level identification is the difference between knowing who to call and knowing only which company to research.
The counterweight comes from the same sources. Cook writes that the buyer "isn't always whoever happened to click the most." With 22 people in an average decision, the most active visitor may be an analyst gathering options for someone else.
Coverage figures need the same care. Vector says roughly a third of US B2B site traffic resolves to a named contact, but that is Vector's own claim, not an independent measurement. Before tiers depend on person-level matches, check match accuracy on known accounts against companies you already know.
A workable decision rule:
- Act on the person when the role is a decision-maker and the visit lands on a buying page.
- Treat it as an account signal when several people from one account visit inside the window, and send it to the account owner.
- Hold individual-contributor visits for marketing nurture until the account shows more activity.
What Threshold Should Trigger Sales Outreach
Knock2's framework treats a third engaged session inside a 14-day window, or any return visit to pricing or demo as the point to hand an account to a named SDR or AE the same day, with one visit logged and held. Knock2 defines an engaged session as at least 10 seconds or two or more pageviews.
The table puts Knock2's tiers, which are practitioner opinion, next to the thresholds three of the covered tools encode.
Escalation tiers and tool thresholds
| Rule set | Threshold | What happens |
|---|---|---|
| Knock2 tier 1 | First engaged session, ICP account | Log and hold |
| Knock2 tier 2 | Second engaged session in window | Slack alert, light touch |
| Knock2 tier 3 | Third session or pricing or demo return | Same-day named SDR or AE |
| Warmly | Score 90 to 100 | Engage immediately |
| Warmly | Score 70 to 89 | Prioritise outreach |
| Salesforce Account Engagement | Team-set threshold score | Auto-assign to a rep |
| Common Room | Team-set score ranges | No default cut-off |
Knock2's default window is 14 days, shortened to 7 for short sales cycles and stretched to 30 for long ones. Quiet accounts reset to zero once the window closes, so an old visit never props up a new alert.
Tier 1 is where most noise sits. Holding it out of rep channels is the first step in filtering low-intent visitor alerts, and it keeps Slack for the tiers that need a response.
The tools encode thresholds in different ways. Warmly publishes fixed score bands, with 90 to 100 meaning engage immediately, 70 to 89 prioritise, 50 to 69 nurture and below 50 not in market. Salesforce Account Engagement lets a team set a threshold score that auto-assigns a rep. Common Room lets teams edit the score ranges and publishes no default.
Vendor bands carry a catch. Warmly's model sets the weights behind its score, so a team cannot change why an account crossed 90. Those bands are a vendor's view of intent, not the team's own definition.
A decision rule for thresholds:
- Set the window to the sales cycle, using 7, 14 or 30 days as Knock2 suggests.
- Count engaged sessions, not pageviews, so a bounce does not move an account up a tier.
- Let a return to pricing or demo jump straight to the top tier.
- Write the vendor band you act on into your own rules, rather than taking the default as the definition.
Who Should Own A Visitor Signal, Sales Or Marketing
Sales should own the higher tiers through the account owner in the CRM, and marketing should own the first tier and individual-contributor visits. Pocus splits ownership by role this way, and Heinz Marketing's checklist tells teams to document signal triggers and the GTM responses tied to each, so ownership is settled before the first alert fires.
In Heinz Marketing's model, marketing runs threshold-triggered nurtures, sales follows up with plays tied to behaviour, and CSMs get alerts from product usage. The CSM role there rests on product data, not on website visits.
Pocus draws the line by role. Decision-maker visits to high-intent pages go to personal sales outreach, and individual-contributor visits go to marketing for lead warming. That rule keeps sales time for visits that can move a deal and still uses the rest.
The tools assume the CRM holds the answer. Common Room keeps team segments updated from each rep's book of business in the CRM, so a rep's alerts cover the accounts that rep owns. Salesforce Account Engagement assigns a prospect to a sales rep automatically once the prospect reaches the thresholds the team sets.
Both depend on clean ownership fields. Before routing goes live, connecting visitor identification to your CRM decides whether a match lands on the right account and the right owner.
A decision rule for ownership:
- Account owner in the CRM gets the tier 2 Slack alert and light touch, and owns tier 3 on owned accounts.
- SDR queue takes tiers 2 and 3 on unowned ICP accounts.
- Marketing holds tier 1 and nurtures individual-contributor visits.
- CSMs take product-usage alerts, as Heinz Marketing's model sets out.
How Fast Should Sales Follow Up On A Visitor Signal
In Workato's test of 114 B2B companies, more than 99% did not respond to a demo request within five minutes, and teams using routing tools still averaged over three hours, so a visitor signal with no owner and no routing rule usually goes cold.
99%
More than 99% of the B2B companies in Workato's test missed a five-minute response to a demo request.
2026 · 114 B2B companies · Source
The rest of the test is harder reading. Nearly one in five companies never replied by email. Companies with lead routing tools averaged 3 hours 32 minutes, against nearly 13 hours for companies without them.
Those were demo requests, the clearest hand-raise a buyer can make. Everyone says to act fast on intent, yet observed data shows most teams miss even hand-raisers by hours. A visitor signal is weaker and quieter than a form fill, so it needs a routing rule even more.
The practitioner rule is same-day outreach for the top tier. Knock2 escalates a tier 3 account to a named SDR or AE the same day. Warmly updates first-party website signals in real time and third-party signals within hours, so the delay usually sits with the team, not the data.
Timing matters because the window is short in buyer terms. 6sense found buying cycles averaged 10 months in 2025, down from about 11 months in 2024, and most buying groups had a preferred vendor before first contact. A same-day reply reaches the account while that preference is still being formed.
How Warmly Scores Visitor Intent
Warmly scores accounts on a 0 to 100 scale that mixes website visits, pricing-page views and Bombora research data, and its own bands say a score of 90 or more means engage immediately while 50 to 69 means nurture. Warmly is a visitor identification platform with AI sales agents.
Warmly
B2B go-to-market platform with AI agents that identify website visitors and engage them through chat, email and retargeting, for sales and marketing teams.
Warmly identifies website visitors and engages them through AI agents, chat, email and retargeting. It serves sales and marketing teams that want person-level website intent turned into conversations.
- Platforms
- API
- Compliance
- GDPR, SOC 2
- Ratings
- G2 4.5/5 (250 reviews)
Signal: First-party inputs are website visits with frequency and recency, pages viewed with pricing treated as high intent, email engagement and chat. Third-party inputs are Bombora topic surges, job postings, technographics, hiring and news. Research intent comes from a data co-op of over 5,000 B2B websites, with Bombora and Demandbase as data partners.
Scoring: Scores of 90 to 100 are Urgent, 70 to 89 High, 50 to 69 Medium and below 50 Low. First-party website signals update in real time and third-party signals within hours.
Owner path: Warmly's Inbound Agent turns intent signals into conversations and meetings.
Limit: The model sets the weights, so a team sees the band but not the reasoning behind it. The third-party share of the score is also data other sellers can buy.
HubSpot announced in June 2026 that it will acquire Warmly. Contracts, pricing and integrations stay unchanged for now, but the deal affects any roadmap or integration bet built on Warmly.
How Vector Reads Contact-Level Intent
Vector ties intent to a named contact instead of an account, and its VP of Marketing argues that first-party behaviour, such as a director returning to pricing within days, outweighs a third-party topic surge. Vector is an account-based marketing platform that turns those signals into contact-level ad audiences.
Vector
Account-based marketing platform that turns buyer signals and account scoring into contact-level ad audiences for B2B marketing teams.
Vector turns buyer signals and account scoring into contact-level ad audiences. It serves B2B marketing teams running account-based programmes.
- Public API
- Yes
- Compliance
- GDPR
Signal: Vector de-anonymises on-site and off-site activity at contact level, not only company level. It partners with a network of 400,000+ publishers, says it processes 10 billion+ signals daily, and uses AI to separate casual mentions from buying intent. Custom intent models track brand, competitor and category research.
Limit: First-party data only covers people who reach you, as Cook notes, and the share of traffic Vector resolves to a named contact is Vector's own claim. Teams set their own outreach rules on top of Vector's contact data, so the threshold work from the earlier section still applies.
How Common Room Scores Leads And Accounts
Common Room lets teams weight fit and behaviour themselves and set the score ranges that count as good or excellent, then sends real-time Slack alerts when a high-fit contact visits a high-intent page. Zoom completed its acquisition of Common Room in July 2026, and Common Room's CEO said the team will continue to serve customers.
Common Room
Common Room is a go-to-market platform that builds buyer intelligence from CRM, product and engagement data and gives sales, RevOps and marketing teams AI agents and outreach tools to act on it.
Common Room builds buyer intelligence from CRM, product and engagement data. It gives sales, RevOps and marketing teams AI agents and outreach tools to act on it.
- Platforms
- Chrome extension
- Public API
- Yes
- Compliance
- GDPR, SOC 2
Signal: Website visits from high-fit accounts on high-intent pages are what Common Room's playbook calls an excellent buying signal, identified at contact level where possible.
Scoring: Scores combine fit, such as industry, annual revenue and economic buyer, with behaviour, such as LinkedIn engagement and product usage. Teams edit the weights and the ranges for not a fit, fair, good and excellent. Scores refresh daily from those rules.
Owner path: Real-time Slack alerts reach reps when a high-fit contact or account visits. A segment can pull high-scoring leads owned by a rep in Salesforce, and a workflow can add them to an email sequence.
Limit: Common Room publishes no default weights or cut-offs, so the threshold work from the earlier section is on the team.
How Salesforce Turns Scores Into Assignments
Salesforce Account Engagement keeps an activity score apart from a fit grade and assigns a prospect to a sales rep automatically once both reach the thresholds a team sets, which makes the CRM the place where a visitor signal gets an owner. Einstein Behavior Scoring adds a 0 to 100 activity score.
Salesforce
Cloud CRM platform for sales, service and marketing teams, with built-in AI agents.
Salesforce is a cloud CRM platform for sales, service and marketing teams, with built-in AI agents. Account Engagement is its marketing automation layer for scoring and grading prospects.
- Free trial
- 30 days, no card needed
Signal: The score tracks activity and the grade tracks fit. A page visit can carry its own points, and a Trailhead example scores "View Pricing Page" at +10 as an illustration, not a default.
Scoring: Einstein Behavior Scoring rates prospects 0 to 100 from activities such as link clicks and form submissions, weighing recency. Scores typically update every 4 hours.
Owner path: Once prospects reach the thresholds, they are assigned to a sales rep automatically. Assignment can also run through page actions, completion actions, automation rules or by hand.
Limit: Einstein works best with a year of engagement data and at least 20 prospects linked to opportunities. Duplicate prospects' scores are not combined, and person accounts are not supported.
Teams comparing visitor identification tools for Salesforce should check how each one writes visits into these score and assignment rules.
How Gong Measures Account Activity
Gong rates account activity from inbound and two-way engagement only, and website visits reach it through integrations such as Leadfeeder by Dealfront that sync visit events to the activity timeline once a day. Gong is a revenue AI platform that analyses customer interactions for sales, RevOps and customer success teams.
Gong
Revenue AI platform that analyzes customer interactions to help sales, RevOps and customer success teams forecast, coach and act on deals.
Gong analyses customer interactions to help sales, RevOps and customer success teams forecast, coach and act on deals.
- Compliance
- GDPR, ISO 27001, SOC 2
Signal: Only inbound and two-way customer engagement counts. Outbound emails and texts do not, and recent activity weighs more.
Scoring: Accounts are rated High, Medium, Low or Inactive across timeframes of three weeks, three months, six months and one year. Very high volume does not automatically give the highest level.
Owner path: Leadfeeder by Dealfront pushes website-visit events to Gong activity timelines, syncing once a day. It needs a CRM connected to both Gong and Leadfeeder, plus a paid Dealfront subscription.
Limit: The numeric score is internal and not displayed, and users cannot change the calculation. A once-a-day sync is too slow for a same-day tier.
Gong acquired RightBound, announced in December 2025, to add third-party account and contact signals such as financial data, news, buying group information and job history. Treat those signals as announced, not as part of activity level today.
What Do Experts Say About Visitor Identification
Practitioners agree that page context and repeat visits matter more than raw traffic, and Heinz Marketing's Sarah Threet names the common failure plainly, writing that most teams treat all engagement as good engagement. They split on two questions, whether third-party surges belong in a score and whether sales should act on a person or the account.
Most teams treat all engagement as good engagement.
Where practitioners agree (opinion):
- Page context beats volume. Pocus rates pricing and comparison visits as high intent, and Knock2 lets a pricing or demo return jump to its top tier.
- Repetition inside a window matters. Knock2 builds its tiers on engaged sessions in a 14-day window, and Jess Cook's director example turns on a return three days later.
- Triggers need documented responses. Heinz Marketing tells teams to write down each trigger and the response tied to it.
Where practitioners disagree (opinion):
- Third-party surges. Warmly includes Bombora data in its score. Cook argues that a surge every seller can see is no edge.
- Person or account. Common Room says reps need contact-level visibility. Cook warns the buyer is not always the person who clicks most.
The observed data settles neither argument, but it frames both. Forrester's count of 22 people in an average buying decision supports watching the account, not only the busiest visitor. Workato's test shows that response speed, not signal choice, is where most teams lose hand-raisers.
How To Define An Actionable Visitor Signal
A working signal definition fits on one page and names the ICP filter, the buying pages, the visit window, the tier that earns outreach and the owner for each tier, as Knock2's tiered escalation and Heinz Marketing's checklist both set out. Write it before alerts go live, then work through the checklist below.
- Name the ICP filter by industry, company size and the roles that buy
- List the high-intent pages and the excluded pages
- Define an engaged session, such as 10 seconds or two pageviews
- Set the window to the sales cycle, using 7, 14 or 30 days
- Set the tiers and the response for each one
- Name the owner per tier, from account owner to SDR queue to marketing
- Set a response time and a routing rule for the top tier
- Check scoring data needs, such as Einstein's year of engagement data
- Review false positives against pipeline every month
Each item maps to the evidence above. The session and window rules come from Knock2. Documented responses per trigger come from Heinz Marketing. The routing rule answers Workato's finding that companies with routing tools averaged 3 hours 32 minutes against nearly 13 hours without them.
Buyer behaviour supports weighting late-stage pages. Forrester found 94% of business buyers use AI during their buying process, and 6sense found 58% of buyers engaged sellers earlier to clarify details AI tools left out. 6sense also found 94% of buying groups ranked preferred vendors before first contact. Because most buyers reach sellers with a ranked shortlist, the definition should give pricing, comparison and demo pages the heaviest weight.
FAQ
Is a single pricing page visit an actionable signal?
Usually not on its own. A single pricing-page visit is usually logged and watched rather than sent to sales, unless it comes from a decision-maker at an ICP account. Knock2's framework holds a first engaged session at tier 1, while Pocus routes decision-maker visits on pricing pages to personal sales outreach.
Do vendor intent scores replace lead scoring in the CRM?
No. Vendor scores such as Warmly's 0 to 100 bands use weights the vendor's model sets, while Salesforce Account Engagement assigns a rep only when a prospect reaches thresholds the team sets. Treat the vendor score as one input to CRM scoring, where ownership and assignment rules already live.
How many signals should a revenue team track at first?
Start with the few signals practitioners agree on, which are ICP fit, visits to pricing, comparison and demo pages, and repeat engaged sessions inside one window. Pocus's playbook and Knock2's tiers both build on them. Add third-party surges or content engagement later, once false positives are reviewed against pipeline.
Is a high Gong activity level a buying signal?
Not by itself. Gong's activity level counts inbound and two-way engagement, with recent activity weighted more, so a High level shows an engaged account rather than a buying decision. Gong also says very high volume does not automatically earn the top level, and the numeric score behind it is hidden.
Sources
Sources
- The Timeline for Influencing B2B Buyers Is Shrinking: Insights From 6sense's 2025 Buyer Experience Report · businesswire.com
- Website Visitors Signal Playbook - Pocus Playbook Library · pocus.com
- The State Of Business Buying: Risk-Averse Buyers Demand Proof, Not Promises · forrester.com
- Repeat Website Visitors: The B2B Sales Trigger You Miss · knock2.ai
- B2B Lead Response Times: What We Learned from 114 Companies · workato.com
- Read the Room: Operationalizing CLG Signals That Actually Matter · heinzmarketing.com
- trailhead.salesforce.com
- help.gong.io
- collective.gong.io



