|8 min read

Reviewed by LeadPilot Editorial

Infusionsoft Lead Scoring in Keap Max: How It Works and How to Design Rules

This guide explains how Keap Max lead scoring assigns, expires, and subtracts points, and how to design and test scoring rules.


If you are researching Infusionsoft lead scoring, Keap's documentation describes lead scoring in Keap Max. In Keap Max, a lead score is a set of rules that automatically assign points to a contact record based on tags applied. Points can also be subtracted, and the number of flames on a contact record is based on the percentage of the ideal score accumulated. Lead scoring is only available in Keap Max [1].

That short description hides several decisions that determine whether the score helps your sales team or just decorates the contact list. This guide explains the mechanics the documentation confirms, then offers a recommended way to design rules, check them against your own outcomes, and spot the signals a points system tends to miss. It is a design and testing guide, not a click-by-click setup walkthrough, so follow Keap's current documentation for exact menu steps.

What Keap's lead scoring does

The documentation describes a lead score as a set of rules that automatically assign points to a contact record based on tags applied [1]. Confirm the full list of other rule inputs in Keap's current documentation before you build around them. The details that matter most for planning are these:

  • Rules, not a manual field. Points are assigned automatically by the rules, not by someone typing a number into the contact [1].
  • Points can expire. You can set points to expire after a specified period [1].
  • Points can be subtracted. The documentation's example is a contact who has not opened an email within a defined period [1].
  • It runs continuously. Once enabled, scoring runs in the background and updates in real time, adding and subtracting points as the rules dictate [1].
  • Flames show progress toward an ideal score. The number of flames on a contact record reflects the percentage of the ideal score the contact has accumulated [1].

Two practical points follow. First, the flame display is relative to an ideal score, so it is only as meaningful as the target you set. Second, because scoring is continuous, a contact's flames today may not match what they were last week. If you want to learn from past outcomes, you have to record the score at the moments that matter.

Confirm your plan before you design anything

The documentation states that lead scoring is only available in Keap Max [1]. If your account is on a different plan, you will not be able to build automatic scoring there. Check this first, because it changes the decision. You may need to evaluate whether a plan upgrade is justified, or whether a simpler approach (for example, tagging contacts by hand when your team judges them ready) covers your volume.

The following framework is the author's recommendation. It is built around the mechanics above: rule-based points, expiration, and subtraction.

1. Start from the action you want sales to take

Before you create any rule, write down what a "high" score should cause. For example: a sales rep calls the contact promptly, or the contact enters a nurture sequence. If a score does not change anyone's behavior, it does not need to exist. Work backward from that action to the evidence that would make a rep confident in making the call.

2. Use tags as your evidence layer

Because the documentation confirms tags can drive points [1], your tagging discipline becomes your scoring discipline. A tag applied by a form, a purchase, a campaign step, or a rep's manual judgment can all become a scoring input. Two habits help:

  • Give each tag one meaning. A tag that means "downloaded a guide" in one campaign and "attended a webinar" in another produces noise.
  • Separate tags by strength of evidence. A contact who requested a quote has told you something far more direct than a contact who opened a newsletter. Your rules should reflect that gap, with larger point values for explicit requests and smaller ones for passive activity. Choose the actual numbers from your own sales history rather than copying anyone else's.

3. Use expiration so old interest does not look current

Because points can expire after a period you choose [1], you can make the score represent recent interest. A contact who clicked a pricing page months ago and has done nothing since should not look identical to one who did so this week. Pick expiration windows that match your sales cycle. A business that closes quickly can use shorter windows. A business with a long evaluation process needs longer ones, or active prospects will cool off on paper while they are still deciding.

4. Use subtraction for silence and disqualifiers

The documentation says points can be subtracted, for example when a contact has not opened an email within a defined period, so a score can fall as well as rise [1]. That is useful in two ways:

  • Silence. If engagement stops, the score should drop so reps stop spending time there.
  • Disqualifiers. If you tag contacts who are clearly not a fit (for instance, outside your service area or a student researching a project), consider a rule that subtracts enough to keep them from ever reaching your ideal score. This only works if the disqualifying information reaches a tag somehow, which is a data-capture problem rather than a scoring one.

5. Set the ideal score deliberately

Since flames express the percentage of the ideal score a contact has accumulated [1], the ideal score is effectively your definition of "fully ready." If you set it too low, many contacts will look nearly maxed out. If you set it too high, even strong prospects will show a modest number of flames and reps may learn to ignore the display. Treat the ideal score as a setting you revisit, not a number you pick once.

A reader-run check: does the score match what sales actually sees?

You do not need a data science project to find out whether your scoring is useful. Here is a simple check you can run with your own records. It is a recommended method, and the results will reflect your data, not a benchmark.

  1. Pull two small groups of contacts. One group should be contacts your team turned into customers or booked meetings with. The other should be contacts who went nowhere after reasonable follow-up.
  2. Compare their flames. Look at the current display for each group. Because scoring updates in real time [1], current flames on older contacts reflect their present state, not their state when sales first reached out. Keep that limitation in mind, and favor recent contacts so the comparison is less distorted.
  3. Look for patterns in the misses. Which converting contacts had few flames? Which dead-end contacts had many? The misses tell you which rules to change. If converting contacts commonly lack a certain tag, you may be missing a scoring input. If dead-end contacts rack up points from low-value activity, reduce those point values.
  4. Ask the reps. Have the people who work the leads say which flames they trust. A rep's skepticism is data about how the score reads in practice.
  5. Record the score at handoff going forward. If you want a clean comparison later, note the score when a contact is first passed to sales, for example in a note or a tag your team applies. Since the live score keeps changing, this gives you a fixed reference point.

Repeat the check after you change rules, so you can tell whether an adjustment helped.

What a points system cannot see

Rule-based scoring is good at counting recorded events. It is weaker at capturing things a visitor says but never clicks. As a hypothetical example, consider a visitor who lands on your site, reads a services page, and asks a specific question about timeline and budget. Unless that exchange becomes a tag or other recorded input, your scoring rules never register it. Meanwhile, a long-time subscriber who opens every email may score highly without any plan to buy.

This is a structural limit of any system that infers interest from tags and activity, and it is worth designing around. Two options:

  • Add explicit questions to your forms and map the answers to tags, so stated needs feed the score.
  • Capture the conversation itself. If many of your valuable visitors never fill out a form, you need a way to learn what they said and when they would like to talk.

Where a website chat agent fits

If your gap is visitors who never trigger a tag, an on-site conversational layer can be a complement to scoring. LeadPilot is an AI sales agent that embeds on a website, learns from your approved site content, answers buyer questions, asks relevant qualifying questions one at a time, captures contact details, and can offer a booking link when intent is clear. Its handoff saves the contact details together with intent, fit, timing, score, and the conversation behind them, so your team sees context instead of a bare form submission.

Be realistic about scope. Confirm how conversation data would reach your Keap contact records before assuming any tag-based workflow. A chat agent also does not replace the need for well-designed scoring rules on the contacts already in your database. It addresses a different problem: collecting stated intent from visitors who would otherwise leave without a trace.

When Keap's scoring may be enough

Keap's built-in scoring may be sufficient if most of your leads arrive through forms, campaigns, and purchases that already apply tags, your sales cycle makes expiration windows easy to choose, and reps are willing to use the flame display. Consider adding other qualification inputs if your own records show that many of your best prospects come through unstructured channels, such as website visitors who ask questions without submitting anything.

Whatever you choose, the principle holds: a score is a hypothesis about who is ready to buy. Write it down, tie it to an action, and test it against what your sales team actually experiences.

If your biggest gap is website visitors who leave before they ever become a contact, you can Build your LeadPilot agent and see how conversation-based qualification works on your own site. The free plan requires no credit card.

Sources

  1. Set Up and Use Lead Scoring in Keap Max

Put your website conversations to work

Build a LeadPilot agent from your website and review it before you publish.

Start Free