How Do You Score Intent Signals From G2 and Website Visits?

How do you score intent signals from G2 and website visits?

Give each signal points based on how close it sits to a buying decision, let the points fade over a few weeks, and combine the total with a separate fit score. Pricing and comparison activity should count most. Blog reads should count least. Then test the score against deals that actually closed before sales relies on it.

Intent data promises to tell you which accounts are shopping right now. In practice, it is easy to buy a feed, pipe it into the CRM, and watch reps ignore it within a month. The data was not the problem. The scoring was. Every signal counted the same, nothing expired, and fit was ignored.

This framework combines two common sources: third-party intent from a review site like G2, and first-party intent from visits to your own website. The goal is one score that tells a rep which accounts deserve attention this week, and why.

What is the difference between third-party and first-party intent?

First-party intent is activity on your own properties, such as visits to your website, opened emails, and product sign-ups. Third-party intent is activity elsewhere, such as research on review sites or publisher networks. First-party signals are usually more specific to you. Third-party signals can show interest earlier, before an account ever visits your site.

Both have blind spots. First-party data misses buyers who research everywhere except your site. Third-party data often works at the company level, so you know an account is active but not always who inside it. Used together, each covers some of the other's gaps.

The mistake is treating them as equal. A visit to your pricing page is direct evidence of interest in you. Research on a category page elsewhere might be interest in a competitor. Weight them accordingly.

What does G2 intent data show?

G2 describes its Buyer Intent data as showing companies researching your software and your category over the past 30 days. It lists signal sources such as category research, competitor comparisons, pricing page visits, and direct engagement with your G2 profile. G2 also names Salesforce and HubSpot among the tools it connects to.

Those signal types are not equal. An account viewing your profile and comparing you with a competitor is much closer to a decision than one browsing a category page. Score them separately rather than as one "G2 activity" point.

Check G2's own documentation for what your plan includes and how the data reaches your CRM. The details of access and integration are worth confirming before you design scoring around them.

Which website visits signal real buying intent?

Visits to pricing, comparison, integration, security, and demo pages signal the strongest intent, because people read those pages when they are evaluating a purchase. Case studies and product pages show moderate intent. Blog posts and general content show interest in a topic, which is useful context but weak evidence of buying.

Repeat visits matter more than single ones. One pricing page view could be curiosity. Several visits from the same account in a week, especially from different people, look like an evaluation. Count distinct visitors and repeat visits, not just page views.

If you want sales to react quickly to the strongest signal, set up an alert when a target account hits your pricing page. I covered that setup in how to alert sales when a key account visits pricing.

How do you weight the signals in one score?

Assign points by stage. For example, give pricing and comparison activity the most points, product and integration pages a middle value, and content reads the least. Add points for repeat activity and multiple people from one account. Then subtract points as signals age, so activity from last week outweighs activity from last quarter.

A simple starting scale might give a pricing page visit ten points, a G2 comparison involving you eight points, a case study read five points, and a blog read one point. These numbers are a starting point, not a rule. Adjust them once you see which signals appear before real deals.

Decay is essential. Without it, accounts collect points forever, and old research looks like current interest. One simple approach is to halve a signal's value after a couple of weeks and drop it after a month or two. Pick a window that matches your typical sales cycle.

How do you combine intent with fit?

Keep intent and fit as separate scores and read them together. Fit tells you whether an account matches your ideal customer profile. Intent tells you whether it is active now. High fit with high intent gets immediate attention. High intent with poor fit gets a light touch. High fit with no intent goes to nurture or outbound.

Merging them into one number hides the reason behind the score. A rep seeing "75" does not know whether that is a perfect-fit company with no activity or a poor-fit company researching heavily. Two numbers, or a simple grid, make the next action obvious.

If your fit scoring is not set up yet, start there. I explained how to build one reps will actually use in how to build a lead score sales will trust.

How should sales act on the score?

Give each combination of fit and intent a clear play. High fit and high intent gets personal outreach within a day, referencing the topic they researched without being creepy. Medium combinations get a sequence. Low fit gets nothing extra. Write the plays down so every rep responds the same way.

Tell reps what the signal was, not just the score. "Three people from this account viewed pricing and compared you with a competitor this week" is actionable. "Intent score 82" is not.

Be careful how you use the information in outreach. Saying "I saw you were on our pricing page" can feel invasive. Use the signal to pick the right message, such as a note about pricing questions or a comparison guide, rather than to announce that you were watching.

How do you know the score is working?

Check whether high-scoring accounts become opportunities more often than low-scoring ones. Look back at recent closed deals and see whether the score rose before the deal started. If it did not, adjust weights. Review this monthly at first, then quarterly once the score proves itself.

Track how often reps act on high scores, too. If reps ignore the score, find out why. Usually it is because the score flags accounts that are obviously wrong, which points back to fit criteria or noisy signals.

Drop signals that never show up before real deals. Every signal you track has a cost, in data fees, in setup, or in rep attention. Keep the ones that predict pipeline and cut the rest.

What should you do next?

List the intent signals you already have, from your website analytics and any G2 or other third-party feed. Assign starting points by how close each sits to a decision, add decay, and keep fit as a separate score. Then compare the scores of your last ten closed deals to see whether the model would have spotted them early.

Start with first-party signals if you have no budget for third-party data. Your own pricing and comparison pages are often the strongest intent you can get.

I build websites and the automations behind them for lead generation, and intent scoring is where website data finally reaches the sales team. If you want help building a score your reps will act on, reach out and let's chat.