You can't read real purchase intent from pageviews and time on site alone. Those numbers feel easy, but they often send sales and marketing after the wrong people while real buyers slip through. To treat your website like a real growth channel, you need to look past click paths and focus on who's visiting and what they're actually signaling.
Page depth and time on site alone often create a big chunk of false interest, especially during research spikes and planning seasons. When you mix identity, context, and behavior, you can rank visitors by purchase likelihood instead of guessing from session stats.
Why Pageview-Only Intent Models Keep Letting You Down
Traditional models score a visit by pageviews per session, time on site, bounce rate, and visits to a few "high intent" URLs like pricing or demo. They act like every browser tab is equal, ignoring who the visitor is, what account they work for, and what that account is doing in other channels.
That creates common problems:
- Research traffic and students that look like buyers
- Job seekers reading your company and product pages
- Bots or internal visits inflating "engagement"
- Seasonal spikes that make average numbers look better than they are
A rule like "3+ pricing page views = high intent" often just means one team keeps coming back to the same page or interns are researching. Sales ends up chasing the same account over and over, while fresh high-fit visitors get no special treatment.
Reading Website Visitor Intent with Identity First
To fix this, start with identity, not pageviews. Identity resolution ties devices, cookies, emails, and offline records back to real people and accounts. Instead of "New Visitor 123," you know it's a director on a buying team or a consumer profile with a clear history.
Identity changes how you read intent:
- A generic 3-page visit tells you almost nothing
- A 3-page visit from a director at a target account who also opened your email and downloaded a guide says a lot
Teams that adopt identity-first models often see identified traffic rise from 10–20% of visits to 35–50% within a quarter, depending on their audience and consent rates. That jump in known visitors is what makes better scoring and routing possible.
Expanding Intent Signals Beyond Clicks and Sessions
Once identity is in place, widen the set of signals. Strong intent models pull from three areas:
- On-site behavior: pages viewed and scroll depth, repeat visits and recency, content type (solution, educational, support)
- Off-site intent: third-party research activity, search and review site patterns, ad engagement and view-through
- Identity-linked context: job changes or seniority, firmographic shifts like hiring surges or funding, channel habits
A simple blended approach: start with a base score when someone hits two or more key pages, add points if they're at a target account or in the right role, add more if you see recent off-site research or ad engagement.
Over a full quarter, teams that test blended signals against behavior-only rules often see 15–30% lifts in opportunity creation or demo conversion, even with the same traffic volume.
Turning Intent Signals Into Actionable Playbooks
Scores do nothing on their own. Every high-intent model should answer four questions for each visitor: route to sales, nurture and watch, suppress from heavy remarketing, or enrich and log for later?
Straightforward rules:
- High identity + high intent: route to sales within minutes
- Medium score: send to product-led nurture or light outbound
- Low score: cap ad frequency and let them self-educate
Then line this up by channel: email triggers for people who reached pricing but didn't finish, ad bid adjustments up for rising accounts and down for anonymous bounce traffic, and stronger on-site "Talk to Sales" prompts for known high-intent return visitors.
Seasonal timing matters. During midyear budgeting, if intent rises across a cluster of target accounts, send more direct outreach, build small custom landing pages, and give sales talking points based on what visitors read. Teams that act on these windows often see 10–20% better demo-to-opportunity rates.
Measuring Whether Your Intent Detection Is Working
- Baseline metrics: match rate, qualified-visit rate, high-intent visitor volume over time
- Outcome metrics: demo-to-opportunity conversion, opportunity-to-win rate, average deal cycle for high-intent vs. low-intent segments
- Testing metrics: lift between old and new scoring models, SDR meetings set per contacted visitor
A simple 60–90 day pilot: benchmark your funnel with pageview-only scoring, turn on blended identity-first scoring for half of your traffic, keep the rest on the old model, and compare closed-won revenue per 1,000 visitors and outreach efficiency.
Building a Next-Generation Intent Strategy in 90 Days
- Days 1–30: audit current tracking and routing, choose an identity and intent partner, set target match rates and audience segments
- Days 31–60: implement identity resolution, configure blended scoring, launch at least two activation playbooks
- Days 61–90: run side-by-side tests with your old model, tune thresholds, set a quarterly review of match and conversion rates
When you evaluate identity and intent platforms, focus on broad identity coverage across consumer and business profiles, consistent match performance, and the ability to plug intent data into the tools your teams already use.
See our approach to website visitor intent detection, or book a demo.