Points for activity anyone can rack up
How you'll know: Leads cross the line on newsletter opens and blog visits, and sales finds nobody ready to talk.
Fix: Weight buying pages and direct requests far above general reading, and stop scoring opens.
Lead scoring gives every lead a number, so sales calls the likeliest buyers first instead of working a list from the top. Here's how a lead scoring model works, how to set one up on any platform, and what it looks like in Marketo.
Fit points for who the lead is: job title, company size, industry. Behavior points for what they do: pricing page visits, form fills, webinar attendance.
A threshold marks the score where marketing hands the lead to sales, often called the MQL line.
The score only earns its keep if crossing that line makes something happen, such as an owner assigned, a task created or a rep alerted.
Pull last year's closed-won customers and write down what they had in common before they bought: roles, company size, industry, and the pages or emails they touched. That list is your model's first draft.
Give points for the traits on that list. Weight the ones that matter most, so a decision-maker at a company your size outscores an intern at a huge one.
Give more points for buying signals, like the pricing page, a demo request or a case study download, and fewer for general interest like a blog visit.
Watch for: email opens. Mail privacy features load emails automatically, which counts as an open, so opens inflate scores. Clicks and visits are more honest.
Subtract for signals that a lead is not a buyer: competitors, students, job seekers on the careers page, unsubscribes, a free email domain if you only sell to companies.
Remove behavior points on a schedule when a lead goes quiet, so the number shows current interest rather than a peak from six months ago.
Pick the score where a lead gets handed over, and decide together what happens at that moment and how fast a rep follows up.
Every month or quarter, compare high-scoring leads with the deals that were won. Raise the points that predicted wins and cut the ones that did not.
| Title: VP of Operations (fit) | +15 |
|---|---|
| Company of 50 to 200 staff (fit) | +10 |
| Industry you sell to (fit) | +10 |
| Visited the pricing page twice (behavior) | +20 |
| Attended a webinar (behavior) | +10 |
| Viewed the careers page (could be a job seeker) | − 10 |
| No activity in the last 30 days (decay) | − 10 |
| Lead score | 45 |
With a handoff line at 50, this lead waits. One more pricing visit puts it over. The model only helps if that moment reaches a rep the same day.
Marketo has no separate scoring module to switch on. You build scoring from smart campaigns and score fields. A trigger watches for an action, such as filling out a form or visiting a web page, and a Change Score flow step adds or subtracts points. The total sits on the person record, where other smart campaigns, alerts and reports can use it.
Most Marketo instances keep two scores: one for fit, built from firmographic data, and one for behavior. A lead can look perfect on paper and still score low on behavior because they've gone quiet, and seeing both tells sales more than one blended number.
Decay is a batch campaign that runs on a schedule and subtracts points from people with no recent activity. The handoff is another campaign that fires when the score crosses your line, and usually changes a status field and passes the lead to your CRM. If that CRM is Salesforce, see our page on the Marketo and Salesforce integration.
How you'll know: Leads cross the line on newsletter opens and blog visits, and sales finds nobody ready to talk.
Fix: Weight buying pages and direct requests far above general reading, and stop scoring opens.
How you'll know: Your highest scores belong to competitors, students and people who last visited a year ago.
Fix: Add negative scoring and decay. A score should be able to go down.
How you'll know: Sales calls MQLs "marketing's leads" and works its own list instead.
Fix: Set the line with sales, then review it together against won deals.
How you'll know: Hot leads sit in a smart list or a report until someone remembers to look.
Fix: Make crossing the line assign an owner, create a task and send an alert at that moment.
How you'll know: An account with an angry support ticket or falling product usage still scores as hot.
Fix: Pull CRM, help desk and product data onto the record so the model can score it. Marketo can't see it otherwise.
A way of ranking leads with points. Fit points reflect who the lead is, behavior points reflect what they've done, and the total tells sales who to call first. Most marketing automation and CRM tools include some form of lead scoring software.
One that keeps fit and behavior as two separate scores instead of one blended number, takes points away as well as adding them, and gets checked against won deals. Two scores let sales tell a good-fit lead who has gone quiet from an active lead who isn't qualified.
In tools that use both terms, such as Salesforce's Account Engagement (formerly Pardot), the grade is fit, usually a letter, and the score is interest from behavior. An A-grade lead with a low score is a good buyer who isn't active yet.
Yes, but not as a module you switch on. You build it from smart campaigns, triggers, the Change Score flow step and score fields. That gives full control over the rules, and means someone has to design and maintain them.
Salesforce's Einstein lead scoring and HubSpot's scoring tools, including its predictive option, are more packaged and need less setup. Marketo is more flexible and entirely built by you. None of the three can score on data held outside the platform until an integration brings it in.
Only if you build it. The usual pattern is a scheduled batch campaign that subtracts points from leads with no recent activity, so the score reflects what they're doing now rather than a peak months ago.
Not on its own. The outside data has to land on the Marketo record first, through an integration, before a smart campaign can score it. That integration is the gap we close.