B2B lead scoring: fit first, timing second
Most writing about lead scoring gives points for opened e-mails and visits to the pricing page. That works for leads who already came to you. In outbound prospecting there is no behaviour to count, because the company does not know you yet. Two questions remain: does it fit your offer, and does it have a reason to buy now. Here is how we combine them into one number, with the weights and the rules.
Published 8 Oct 2026 · 5 min read
Paste your websiteTwo questions that are easy to confuse
Fit says whether a company could be your customer at all. Industry, size, region, business model, the problem you solve. This answer changes slowly. A company that fitted last year probably fits today.
Timing says whether the company has a reason to buy right now. A new site, a won tender, an EU grant, a board change, hiring for a role your offer supports. This answer changes fast and loses value within months.
The most common mistake is ranking by timing. A list of companies with the freshest signals looks attractive, but without fit it is a list of companies that are buying something, just not from you. That is why fit weighs more than timing in our formula.
The formula: five components and their weights
Each component is a number from 0 to 100. The final score is their weighted average, also out of 100.
| Component and weight | What it measures |
|---|---|
| Fit, 40% | Industry, size, geography and whether the company plausibly has the problem you solve. A region outside the voivodeships you sell to lowers fit. |
| Intent, 30% | Strength and recency of signals relevant to your offer. No signals caps it at 25. An undated or weak signal caps it at 55. |
| Company quality, 15% | Whether the company can be verified: a description, a tax number, a registry record, headcount. A company struck off the register loses points. |
| Freshness, 10% | Age of the newest signal. Full value up to 30 days, then a linear fall to zero at one year. An undated signal counts as six months old. |
| Confidence, 5% | How well the evidence supports the scores above. High confidence needs a registry match, a clear website and at least one dated signal. |
Three rules that protect the ranking from itself
- Hiring counts only when the role relates to your offer. A company hiring warehouse staff has no reason, from that alone, to buy HR software. Hiring in general signals growth, not purchasing. In our scoring it lifts intent only when the role is the one you sell to, or the one your offer directly supports.
- No date is not neutral, it is a penalty. An undated signal could be two weeks old or three years old. If it counted as fresh, the list would reward stale news. So an undated signal gets a lower cap on intent and counts as six months old for freshness.
- Hard exclusions are not part of the score. Exclusions from your customer profile, legal forms, headcount range and voivodeships are checked by a separate, deterministic test. A company that fails it does not go out with the rest of the list, whatever its score. A high number cannot hide a condition the company does not meet.
How to read the score
A score of 80 or more is marked strong, 50 to 79 medium, below 50 weak. Most companies land in the middle, and that is intended. If half the list scored above 90, the scale would say nothing.
A number alone is not enough to write a first message. So every company comes with a reason it fits and signals with a date and a source address. The number tells you where to start. The reason tells you what to write. The source lets you check it is true.
The ranking is a starting point, not a verdict. You know your market better than we do. If company twenty matters more to you than company one, start there.
What Pursiva does with it
Pursiva reads your website, builds an ideal customer profile from it, finds matching companies in public Polish sources and scores each one with the formula above. You get up to 40 companies in score order, each with a reason it fits, its signals and links to their sources.
The run delivers itself within an hour of payment. You get no names and no contact details. People inside a company are described by role only.
Frequently asked questions
- How is prospect scoring different from classic lead scoring?
- Classic lead scoring counts the behaviour of people who already know you: opens, visits, downloads. In outbound prospecting that data does not exist, so you score the company: does it fit your offer, and does it have a dated reason to buy now.
- Why does fit weigh more than timing?
- Because a signal at a company that does not fit leads to a conversation that will not end in a sale. In Pursiva's score fit is 40% and intent 30%. Hard exclusions are checked by a separate test, outside the score.
- Can I change what Pursiva treats as a fit?
- Yes. Before you pay you see the ideal customer profile built from your offer and can correct it: industries, exclusions, legal forms, headcount range, voivodeships and PKD codes. The scoring formula is fixed, but what companies are compared against is yours to set.
Read next
See a ranking of companies built from your own offer
Paste your website address. You first see the ideal customer profile and can correct it before you pay. A run costs PLN 183.27 gross (PLN 149 net + 23% VAT) and lands in your inbox within an hour.
Paste your website