Your SDR has time for one more call. At the top of the queue: an operations director at a company you could serve well, and an independent consultant who watched every demo topic. Their combined lead scores are identical. Who gets the call?
A lead scoring matrix separates buying intent from customer fit, then assigns a response to each combination. Use recent behavior to assess interest, firmographics and requirements to assess suitability, and an explicit unknown state where evidence is missing. The score should help you choose the next action, not pretend qualification is finished.
What is a lead scoring matrix?
A lead scoring matrix compares how interested a prospect appears with how well you can serve them. It preserves a distinction that a single ranked list often hides: being interested and being a suitable customer are different things.
Intent: What recent actions suggest about a prospect's interest in evaluating your product.
Fit: How closely the prospect's company, role, use case, and requirements match the customers you can serve successfully.
The matrix is a working agreement between marketing and sales. One combination calls for a conversation. Another calls for a focused product tour. A third calls for checking a deployment requirement before borrowing a solution engineer's afternoon.
I like this approach because it makes disagreements specific. When a rep says, “These leads aren't good,” you can ask: wrong companies, weak interest, or the wrong follow-up? Those are different problems, and adding points won't fix all of them.
Keep the boundary clear, though. Watching a security walkthrough doesn't establish budget, purchasing authority, or an approved project. It tells you what to ask about.
Why can identical lead scores require different responses?
A combined score compresses different kinds of evidence into one number. That's useful for sorting a queue, but insufficient for deciding what to say.
Consider an illustrative model weighted 70% toward intent and 30% toward fit:
- A prospect with intent of 90 and fit of 20 scores 69.
- A prospect with intent of 60 and fit of 90 also scores 69.
The first deserves a quick suitability check. The second may deserve a conversation about whether the problem is active. Sending both a generic meeting request ignores the information you already have.
Dale, the demo automation platform, provides separate Intent and Fit scores, Hot / Warm / Cold bands, and the signals explaining them. Its default Lead Scoring model uses that weighting, with configurable weights, thresholds, decay, and signal points. The Lead Scoring controls let you adjust the model to your qualification policy.
My rule: use the total to organize attention, then inspect the components before deciding how to respond.
How do you build a useful intent-versus-fit matrix?
Define your fit requirements and meaningful intent signals, then give each combination an owner and a concrete action. Start with a simple operating table, not an elaborate formula.
This is a suggested team workflow, not a claim that your software supplies this exact matrix view.
| Fit | Intent | Next action | Question to resolve |
|---|---|---|---|
| High | High | Prioritize a contextual sales conversation | Is there an active project, and what remains unresolved? |
| High | Low | Offer focused education without assuming urgency | Is this problem a priority now? |
| Low | High | Check the apparent mismatch before assigning specialist time | Can we actually meet the requirement? |
| Low | Low | Keep outside the active sales queue | Has anything changed enough to warrant review? |
Unknown belongs outside the low category. An empty company-size field isn't poor fit. An unopened invitation isn't rejection. You can have a strong-fit account with insufficient evidence about intent.
- Separate constraints from preferences. An unsupported deployment requirement can prevent a sale. An unfamiliar job title usually warrants a question, not rejection.
- Name the behaviors that matter. Replace “engaged with content” with something observable, such as prioritizing an implementation topic and downloading its supporting document.
- Set thresholds your team can defend. Use past outcomes and current capacity. Another company's cutoff has little value without its sales context.
- Assign ownership and exceptions. Decide who checks mismatches, who handles quiet good-fit accounts, and when a rep can override the recommendation.
“Nurture” isn't a next action. “Send the approval-routing walkthrough and ask whether exception handling is on the current project list” is.
Which demo behaviors reveal meaningful buying intent?
Prioritize deliberate choices and evaluation steps over passive viewing. A long session can support an intent assessment, but duration alone shouldn't earn a technical meeting.
Explicit preferences give reps a better starting point. In a Linear Demo with buyer-selected topics, viewers rate topics Important / Somewhat Important / Not Important. Important topics play first, while Not Important topics are omitted.
Imagine a prospect marks approval controls Important, explores related Walkthroughs, and downloads an implementation document. That's a coherent evaluation sequence. Someone who leaves a video running while answering email may produce viewing time without the same evidence of interest.
I would organize the signals into a few practical groups:
- Problem relevance: The buyer deliberately prioritizes a topic related to a problem you solve.
- Evaluation depth: They explore related media, Walkthroughs, and documents rather than only opening the demo.
- Next-step interest: They click toward scheduling or another substantive action. Keep a scheduling-link click distinct from a completed booking.
- Internal circulation: They share the demo, potentially bringing other stakeholders into the evaluation.
These Buying Intent Signals from demo engagement are evidence, not verdicts. A rewatch could mean interest. It could also mean the explanation was confusing, or the viewer was interrupted.
Before assigning more weight to repeated activity, ask what a rep would do differently because of it. If the answer is “nothing,” you may be making the score more complicated without making it more useful.
How should website visits change a lead's score?
Website visits should influence intent when they add context to an evaluation, not improve fit merely because someone is active. A pricing visit can justify a closer look; it doesn't establish budget.
Dale Radar website visitor intelligence connects website activity with contacts and their demo history, including journeys where the site visit happens before the demo. Page views, time on page, return visits, and other website activity feed Lead Scoring automatically.
For example, a prospect watches an approval-workflow demo, returns to pricing, and then reads implementation material. A useful follow-up might offer help scoping rollout requirements. It wouldn't open with a detailed account of their browsing history.
Now consider a pricing visit with no supporting context. The visitor might be comparing vendors for a future project, researching the market, or checking an existing subscription. The page alone can't tell you which.
Keep fit anchored to suitability. More visits cannot resolve a confirmed use-case mismatch or make an unsupported requirement disappear.
Should unopened demos count as Cold leads?
No. In the product's scoring model, contacts who haven't opened a demo are New, never Cold. Missing behavioral evidence is different from observed interest that has faded.
This distinction matters most when teams are under pressure to tidy the pipeline. An invited buyer might be on leave, waiting for a colleague, or simply unable to give the material attention yet. None of those possibilities proves interest, but neither does an unopened link prove disinterest.
For a good-fit New contact, use the context you actually have. If they requested information about routing exceptions, send a concise reminder about that workflow. Don't write “Following up on what you watched.”
A previously active buyer whose engagement has aged is a different case. Intent decay reduces the influence of old activity so yesterday's evaluation doesn't retain permanent urgency. Fit may remain strong even as current intent weakens.
And don't let the model become a gatekeeper. If a suitable prospect asks for a technical conversation, they shouldn't have to watch more content to qualify for a reply.
What should sales say to each type of lead?
Match the message to what remains uncertain: timing, relevance, or suitability. Use the evidence to make the offer useful rather than to demonstrate how much you tracked.
The following are illustrative messages, not customer quotes.
High fit, high intent: offer a working conversation
“If approval controls are part of the evaluation, we can use the next conversation to map your exception process and identify any gaps.”
This offers a reason to meet. It also leaves room for the buyer to correct your interpretation instead of assuming they're ready to purchase.
High fit, low intent: reduce the commitment
“If routing exceptions is on your team's agenda, here's a focused walkthrough. If it isn't a priority yet, no need to book a call.”
A relevant product tour is often a better next step than another calendar request. For New contacts, ground the message in a prior conversation or stated need, not unobserved viewing behavior.
Low fit, high intent: check the constraint
“Before we arrange a technical session, can you confirm whether self-hosting is a requirement?”
Ask about whatever might actually prevent a successful deployment. Firmographic data can be incomplete, and a consultant may be evaluating on behalf of a suitable client. Verify the mismatch before dismissing the interest.
Once these responses work manually, Segments and Campaigns for behavioral follow-up can support repeatable outreach. Segments group contacts by behavior, and automated Campaigns can send when a contact joins a segment. Agree on the response before automating its delivery.
How do you test a lead scoring matrix before rollout?
Review recent contacts using only the information available when the decision would have been made. Then check whether the proposed action would have helped the buyer and used sales time sensibly.
Include successful opportunities, stalled evaluations, unsuitable inquiries, and unopened invitations. Studying only won deals hides the false positives that keep pre-sales busy without moving an evaluation forward.
Use this short checklist before enabling automated follow-up:
- Intent and fit are visible separately.
- Unknown information isn't treated as a confirmed mismatch.
- New contacts aren't labeled Cold.
- Old activity loses influence.
- Every recommended action has an owner.
- Reps can record why a recommendation was wrong.
Measure whether the policy produces more useful conversations and less unnecessary qualification work. A longer list of Hot contacts isn't an improvement if reps spend their mornings rejecting them.
If you're evaluating interactive demo software, bring those test cases to the Dale interactive demo, then check Dale pricing for the plans covering the scoring and website intelligence capabilities you need.
For your next sales-marketing review, bring a small sample of recent leads. Write down intent, fit, the biggest unknown, and one next action for each. Resolve the cases where your team disagrees before turning the rules on.