B2B lead scoring is a points system that ranks leads by how well they fit your buyer and how much interest they show. In short: give points for fit, give points for actions, take points off for bad signs, and pick one score at which sales steps in. A small model built this way is often enough to start.
This guide shows how to build one, which tools support it, and when to keep it basic.
What B2B lead scoring is
B2B lead scoring is a way to rank leads with points, so sales talks to the best ones first. Each lead gets a number. A higher number means the lead looks more likely to buy soon.
Most models add points for two kinds of facts. The first is fit: does this company look like your best customers? The second is behavior: is this person showing interest right now? Most tools also take points away for bad signs. HubSpot's help pages describe a Positive section that adds points and a Negative section that removes them.
Sources: HubSpot Knowledge Base, "Set up score properties to qualify contacts, companies, and deals" (archived November 2022); Adobe Marketo Engage documentation, "Change Score" (archived July 2022)
Score fit and behavior separately
Fit and behavior answer different questions. Fit tells you who the lead is. Behavior tells you what the lead did. A student who reads ten blog posts can show lots of behavior and still be a poor fit. A perfect-fit buyer who never opens an email shows no behavior yet.
Salesforce's marketing tool for business buyers, then called Pardot and now also called Account Engagement, builds this split into the product. Its score is behavioral and is a number. Its grade is based on fit and is a letter from A to F. Scoring works from the start, while grading needs set up first because every company has its own idea of a model buyer.
| Signal type | Question it answers | Example inputs |
|---|---|---|
| Fit | Who is this lead? | Job title, company size, industry, annual revenue |
| Behavior | What did this lead do? | Form fills, email clicks, pricing page visits, demo requests |
Sources: Salesforce Ben, "Pardot Score" (archived February 2023); Salesforce Ben, "Guide to Grading (Part 1): What is Pardot Grading?" (archived October 2021)
Build a simple model in five steps
- Ask sales what a good lead looks like. Talk to the people who close deals. Pick four or five facts that mark a strong buyer.
- Pick a few behaviors. A form fill or demo request is a stronger sign than an email click. Salesforce's default activity points follow the same idea.
- Assign points. Use round numbers, such as 5, 10 and 20. Give the biggest points to the facts that best predict a sale.
- Add negative points. Take points off for a personal email address, a job title far from the buying team, or a request that is not about buying.
- Set a sales-ready line. Choose a total at which the lead goes to sales. Write the number down.
Here is a sample model. The facts and points are examples only, not a standard.
| Signal | Points |
|---|---|
| Job title is on the buying team | +15 |
| Company is in the target size range | +10 |
| Requested a demo | +25 |
| Visited the pricing page | +10 |
| Clicked a marketing email | +3 |
| Uses a free personal email address | -10 |
With a sales-ready line at 40, a buyer-team lead who requests a demo scores 40 and goes to sales. A lead who only clicks emails does not.
Sources: Salesforce Ben, "Guide to Grading (Part 1)" (archived October 2021); Adobe Marketo Engage documentation, "Change Score" (archived July 2022). The sample signals and points are hypothetical.
What the main tools let you do
The big marketing tools all work on the same idea, though the details differ. Check each vendor's own help pages before you build.
- HubSpot. Paid Professional and Enterprise accounts can edit the score property. You add criteria to the Positive or Negative section and set the points. A score does not add the same points twice if a contact meets the same rule again. You can click a test button to see which criteria a contact meets. After you save, records are checked again, and large databases can take up to two hours to update.
- Marketo. A Change Score flow step adds or subtracts points, or sets the score to an exact number. You can keep more than one score field.
- Salesforce Pardot. Scoring is on by default. Points change by activity type, by automation rules and by manual edits. Score categories let you track particular interests.
HubSpot also warned in its help page that Apple's iOS 15 privacy changes could make engagement scores look higher than usual. Email open data is less reliable now, so lean on clicks, form fills and page visits.
Sources: HubSpot Knowledge Base, "Set up score properties to qualify contacts, companies, and deals" (archived November 2022); Adobe Marketo Engage documentation, "Change Score" (archived July 2022); Salesforce Ben, "Pardot Score" (archived February 2023)
Keep the model simple at first
A simple model beats a complex one when you have little data. You may have few closed deals, a small team, or no tested guess about what predicts a sale. In that case, extra rules only add noise.
Salesforce Ben's advice for grading fits scoring too: do not over-build on the first try. Use four or five criteria, stick with one buyer type, and keep the first rules to match or no match.
Sources: Salesforce Ben, "Guide to Grading (Part 1): What is Pardot Grading?" (archived October 2021)
Check and adjust the model
A first model is a guess. Test it against real deals. Look at the leads that became customers and ask what they scored. Then look at the high scorers who never bought.
Adobe's Marketo help page makes the same point in plain words: get basic scoring in place quickly and tweak the results over time.
Also watch for old scores. A lead who clicked an email a year ago is not hot today. Salesforce Ben covers score decay for Pardot, which lowers a score as time passes without activity. Whatever tool you use, plan for scores to cool off.
Sources: Adobe Marketo Engage documentation, "Change Score" (archived July 2022); Salesforce Ben, "Pardot Score" (archived February 2023)
Sources
- HubSpot Knowledge Base, "Set up score properties to qualify contacts, companies, and deals." (archived November 2022) archive.org copy of knowledge.hubspot.com
- Adobe Marketo Engage documentation, "Change Score." (archived July 2022) archive.org copy of experienceleague.adobe.com
- Salesforce Ben, "Pardot Score." (archived February 2023) archive.org copy of salesforceben.com
- Salesforce Ben, "Guide to Grading (Part 1): What is Pardot Grading?" (archived October 2021) archive.org copy of salesforceben.com