Skip to content
If you treat those two visitors the same way, whether that means ignoring both or chasing both with the same urgency, you’ll waste that invaluable data.
But if you can spot the signs of buying intent – and then identify those companies that look like they’re in buying mode – you can fuel your B2B sales and marketing teams with warm leads that convert much better than cold ones.
Here’s how to read the signals properly, and how to avoid the mistakes that lead teams to chase the wrong accounts.
Why buying intent signals matter more in B2B
In B2B, purchases don’t happen on impulse and rarely happen with a single decision-maker. A buying decision usually involves a group of stakeholders, spread across departments, who research independently before a conversation ever happens with sales. By the time someone fills out a contact form, a meaningful part of the evaluation is often already done.
That’s exactly why intent signals from anonymous, pre-form-fill behavior matter so much. If you’re only paying attention to inbound form submissions, you’re only seeing the tail end of a much longer research process, and you’re seeing it later than your competitors might be.
Buying signal: repeat visits from the same company
A single visit from a company could mean almost anything. It could be a curious researcher, a competitor doing due diligence, an employee who clicked a link by accident, or a job candidate checking out your careers page.
But multiple visits from the same organization within a short window is a much stronger signal. It suggests more than one person, or one very engaged person, is actively investigating what you offer.
As a rough guide, two or more visits within a seven to 10 day window is generally worth flagging. Beyond that window, the visits may reflect ongoing casual awareness rather than active evaluation, so recency matters as much as frequency.
Buying signal: viewing specific sales or marketing pages
Someone who views a blog has a very different intent from someone who’s viewed your pricing page, product comparison content, case studies, or a demo request page. The latter usually signals that someone is further along in evaluating a purchase.
To really tap into the buying signals of page views, it helps to actually map your site’s pages to funnel stages before you start scoring anything. You can start by looking at top, middle, and bottom of funnel to differentiate them.
A useful tip is watching for page sequences, not just individual page visits. For example, a visitor who goes from a blog post to a comparison page to your pricing page in a single session is behaving very differently from someone who lands on pricing directly from a paid ad and bounces.
Buying signal: time on page and depth of engagement
You need to look for the visit that spends several minutes scrolling through detailed content, because that’s a sign they’re really interested. Metrics like time on page and scroll depth, where your analytics setup can capture them, help separate genuine engagement from an accidental click, a bounced visit, or a bot.
Be careful not to over-index on this signal in isolation. A long time on page could reflect genuine interest, or it could reflect someone who left a tab open in the background. It’s most useful as a supporting signal alongside page selection and visit frequency, not as a standalone indicator.
Buying signal: visit frequency increasing over a short window
Intent tends to build before it converts, and it rarely does so instantly. A company that visited once a month and is now visiting several times a week is showing accelerating interest. This trend (the rate of change rather than the absolute number of visits) is often a better indicator of near-term buying activity than any single visit, however deep it was.
This is also a useful way to catch accounts that might otherwise slip past a simple threshold-based system. For example, a company with a sudden spike in activity, even if it hasn’t crossed your scoring threshold yet, is often worth a manual look.
False signals worth watching for
Not every strong-looking pattern is what it appears to be, and treating every signal as equally trustworthy is how teams end up chasing dead ends. A few situations worth building in checks for:
- Vendors and partners researching you, not buying from you. Recruiters, competitors, and existing suppliers can generate visit patterns that look like buying intent but aren’t.
- A single engaged employee outside the buying process. Someone in a research or content role might read everything on your site out of genuine interest without ever being part of a purchase decision.
- Traffic spikes tied to a specific campaign or press mention. A sudden jump in visits from a company right after a press release or LinkedIn post may reflect curiosity about the news, not active evaluation of a purchase.
None of these mean the signal is worthless, but they’re a good reason to combine visitor data with firmographic fit (does this company match your ideal customer profile at all) before treating any single account as sales-ready.
Turn signals into a lead scoring model
These insights are only powerful if you can use them at scale, which is where a lead scoring model can help.
A simple point-based model works well as a first version, and you can refine it once you have enough closed-deal data to see which signals actually correlate with revenue.
All you have to do is agree how many points to assign specific behaviors. For example:
- +1 for a repeat visit within 7 to 10 days
- +2 for a visit to a high-intent page (pricing, demo, case studies)
- +2 for a second contact from the same company engaging
- +1 for increasing visit frequency week over week
- -1 for visit patterns that match known non-buyer profiles (competitors, recruiters, existing vendors)
Set a threshold and agree anything above that gets flagged for immediate outreach.
You should review the model quarterly because as you close more deals, you’ll often find that certain signals (like the second-contact indicator) turn out to be far more predictive than others, and the weighting should shift to reflect that.
