At its simplest, there are two types of lead scoring: demographic and behavioral.
Demographic scoring focuses on the attributes of the individual lead. What role do they have? Where are they located? How big a team do they work with? And, beyond that, what type of company are they from? What's their size, industry, and location?
Behavioral scoring, also called activity scoring, looks at how that person interacts with your brand. Do they go to your website, download content, attend your events, and read or click into your emails?
Demographic scoring determines if you're interested in the lead (because it matches your ICP), and behavioral scoring determines if the lead is interested in your company and products.
Second, scoring is primarily used to rank leads so that marketing and sales know who to follow up with and in which priority. If your company receives so many leads that you need to sift the likely from the unlikely, scoring will help you draw that line and know who to talk to first. On the other hand, if you need them to know a certain amount about your offerings before you have a conversation, scoring can help you examine when each individual has matured enough in their knowledge of your business to warrant a conversation.
Despite how central scoring is to RevOps, most teams aren't confident they're doing it well. The 2025 Openprise State of RevOps Data Quality report found that only 35% of revenue operations leaders have complete confidence in their ability to effectively score leads. The most common culprit isn't a bad scoring model — it's bad data feeding into it. A demographic score is only as accurate as the underlying firmographic and contact data it's built on, and a behavioral score is only as meaningful as the breadth of signals being captured.
The problem with traditional numeric lead scoring
Before we get into how to do lead scoring well, let's talk about a common failure mode: the numeric scoring model that marketing builds, sales ignores, and nobody trusts.
A number-based scoring system — where each activity or attribute earns a certain number of points — sounds logical. But in practice, it creates a few persistent problems.
First, it breeds a lack of transparency. How did this lead get to 75 points? What does 75 mean compared to 67? Has the boundary changed because of a different set of actions by leads? Or have the business's objectives or targets changed? This type of scoring needs constant updates and vigilance or it falls into disuse. Either way, it breeds a lack of trust that ultimately ends in marketing working on scoring, sales disregarding it, and both teams being unhappy.
Second, the timeframe of the activity that earned the score is often unclear. Does attending a webinar within the last 10 days have the same value as attending a webinar months ago? Most traditional scoring systems treat them identically, even though one is a much stronger buying signal than the other.
A better approach: grades, not scores
At Openprise, we've flipped that on its head. We actually don't throttle leads at all, are fully transparent between marketing and sales, and let sales determine how deep into the prospect pool they want to go.
Instead of using numbers, we grade all the leads. A, B, C, or D. This is based on the demographic information and their suitability as a prospect for us. If there's also recent or surge engagement, they get a plus (A+, B+, etc.). Then we decay activity over time, so that four weeks after an activity, that activity is no longer part of the scoring formula. If we haven't been able to follow up on that engagement within a month of the activity, we determine that the lead wasn't really that engaged after all and they don't get their +. That's it. Sales can focus only on their A leads or decide they have the bandwidth to reach into the Bs.
This grading approach only works, though, when the underlying data is accurate. Consider a real-world example: Greg Collins is a Demand Gen Lead based in San Francisco. His company's headquarters is in Tokyo, and the parent company has thousands of employees. If that raw data flows into a scoring model without being cleaned, Greg gets 0 points — wrong location, company too large. He never surfaces to sales.
After his record is standardized, normalized, segmented, and enriched — his actual role, location, and business unit properly resolved — Greg scores 40 demographic points. Nothing about Greg changed. The data about Greg changed. With accurate data, Greg is a prime candidate for outreach. Without it, he's invisible.
This is why the sequence matters: data quality and enrichment have to come before scoring, not after. A well-designed scoring model built on dirty data will produce confidently wrong results. For more on how data quality problems show up directly in scoring, see our post on how scoring issues are a data quality problem.
But wait, there's more: account scoring
That's not the only place to think about scoring. While leads are important, because sales happen between people, account scoring is also important. Only companies with products that only require a quick "buy now" have a single person as the buying team. The purchase of most B2B products requires a team to consider the purchase or, at the very least, going through procurement. For this reason, it's important to assess (and categorize or calculate a score for) not just the individuals, but the entire account.
Openprise also grades accounts the way we grade companies: A, B, C, D, based on how they fit our ideal customer profile (ICP). If they're a fit and they have activity across the organization, they also can get a +. So an A+ lead at an A+ company gets an immediate follow-up and an A+ lead at a B+ company less so. Even an A+ lead at a B company is of less interest because while it represents one engaged person at a company that might be a prospect, the lower account grade lowers the overall score.
Once again, the most important things to focus on are the prospect company's:
- Suitability for your products (grade)
- Overall activity level (+) (because you'll need to convince more than one person at the account to purchase your product)
Going beyond website and email: intent, technographic, and product signals
We've talked about lead scoring using a company's ICP and individual activity, but what about company activity?
Activity scoring can include things like direct activities (like website engagement, form fills, or event attendance), but also you can consider more anonymized information, like surge scoring, which signals intent by measuring the intensity of a company's web activity on particular topics. Companies like Bombora and G2 deliver this type of research that's associated with a company, not an individual. Incorporating intent scoring into your account scoring as part of your overall lead scoring model helps give you a more holistic score. You can even go a step further and consider product usage for trial or freemium products as a key indicator as well!m capabilities
Scoring just website and email activity gives you an incomplete picture. The most accurate models incorporate signals from multiple sources, including:
- Intent signals: Third-party data from providers like Bombora and G2 that shows a company's research activity on relevant topics across the web — even before they've visited your site. A company consuming content about data quality, lead routing, or RevOps automation is showing buying intent that pure behavioral scoring will miss entirely.
- Technographic data: Whether a prospect's current tech stack is compatible with — or already adjacent to — your solution is a strong ICP signal. Openprise's Data Marketplace makes this data available for enrichment directly within your scoring model.
- Product and trial usage: For companies with a freemium or trial product, usage events are often the strongest leading indicator of conversion. How many features has the user activated? Have they invited teammates? Have they hit the product's key "aha moment"? These signals frequently outperform traditional marketing engagement data.
- Momentum scoring: Rather than looking at a score at a single point in time, momentum scoring tracks the rate of change — how fast is a lead's score rising? A prospect who has gone from no activity to three high-intent actions in a week is a different conversation than one who has been at 45 points for three months.
Most marketing automation platforms only let you run one scoring model at a time, which forces teams to choose between competing hypotheses about what good looks like. Openprise lets you run and A/B test multiple models simultaneously — against real sales data — so you can validate which signals actually predict conversion for your specific business, rather than assuming your first model is right.
As you can see, if you've got the data there's no shortage of indicators you can use to build your scoring models.
Interested in hearing more about how we do lead and account grading at Openprise? Download our white paper, The comprehensive survival guide on lead scoring.
.jpg)
