Lead scoring ranks people and accounts by how well they fit and how ready they are to buy. Lead routing sends each one to the right owner, with context, within a set time. Together they decide whether a buyer who raises their hand gets a great first conversation or a slow, generic one.
1Why they belong on one page
Most teams treat scoring and routing as two projects. Marketing tunes the score, sales ops owns the assignment rules, and the handoff falls into the gap between them. A lead scores 80, sits in a queue for a day and gets a call from someone who doesn’t know why.
The point of a score is to trigger the right action. The point of routing is to act on it quickly and with context. Design them as one system and the first conversation gets a lot better for the buyer.
A score with no route is a number in a field. A route with no score is a round robin.
2Score a lead
Tick the signals for an imaginary lead and see where they land. Try a hand-raise, then try a disqualifier.
These weights are examples to show how it works, not a template. Set yours from your own conversion data: which signals actually show up before opportunities do.
3The four kinds of signals
Good models score fit and intent separately, so a perfect-fit company that’s never visited doesn’t look the same as a student who reads everything.
Fit
Who they are: industry, company size, region, role, tech stack, and whether they’re on your target account list.
Engagement
What they do with your marketing: pricing page visits, webinars, email replies, repeat visits from the same account.
Product usage
What they do in the product: activation, depth of use, teammates invited, limits reached. This is what makes a PQL.
Read it →Disqualifiers and decay
Students, competitors and bad data subtract. Old activity fades over time, so last spring’s webinar doesn’t keep someone hot forever.
4Scoring that sales actually trusts
- Hand-raisers skip the score. A demo request or a “contact sales” form goes straight to a person. Never make a buyer earn their way into a conversation they asked for.
- Score fit and engagement on two axes. A simple grid, such as A–D for fit and 1–4 for engagement, tells reps much more than one blended number.
- Set thresholds from real conversion data. Look back at which leads became opportunities, and set the MQL line where conversion actually jumps.
- Close the loop with dispositions. Every lead sales rejects should get a reason. Those reasons are how the model gets better each quarter.
- Add product signals if you have a product to try. I’ve built MQL and PQL scoring in HubSpot fed by Amplitude product signals, so adoption, not just form fills, drove routing and nurture.
5Agree on the stages first
Names vary by company. What matters is that every stage has one definition, one owner and one thing that moves it forward.
| Stage | What it means | Who owns it | What moves it forward |
|---|---|---|---|
| Lead | A known person who hasn’t shown enough fit or intent yet | Marketing | Score crosses the threshold, or they raise a hand |
| MQL | Marketing qualified: good fit and real engagement | Marketing | Routed to sales and accepted |
| PQL | Product qualified: usage shows value and buying signals | Marketing or growth | Routed for a product-led sales conversation |
| SAL | Sales accepted: a rep agrees it’s worth working | Sales development | A real conversation confirms need and timing |
| SQL | Sales qualified: need, authority and timing confirmed | Sales | An opportunity is created |
| SQO | Sales qualified opportunity in the pipeline | Sales | Moves through the deal stages |
6Routing, step by step
Each step removes a reason a good lead waits.
- Step 1
Match to the account
Connect each lead to its existing account so it goes to the person who already owns the relationship. Tools like LeanData do this well.
- Step 2
Enrich at handoff
Fill in missing company and role data the moment a lead qualifies, so assignment rules have what they need.
- Step 3
Assign by clear rules
Named accounts first, then territory or segment, then a fair round robin for the rest. Write the rules down.
- Step 4
Book or notify instantly
Let hand-raisers book a meeting right on the form with tools like Chili Piper, and alert the owner with the reasons the lead qualified.
- Step 5
Hold to an SLA
Agree on a first-touch time for each lead type and escalate when it’s missed, so nothing sits overnight unnoticed.
- Step 6
Capture the outcome
Require a disposition on every routed lead. It feeds the scoring model and shows where the process breaks.
7How to tell it’s working
Speed to lead
time from hand-raise or MQL to first human touchBuyers who reach out are often talking to more than one vendor. Responding first, with context, is a real advantage.
Watch it weeklySLA compliance
leads touched within the SLA ÷ leads routedShows whether the agreement is real or just a document.
Shared metricMQL acceptance rate
SALs ÷ MQLs routedHow often sales agrees the lead was worth working. A falling rate means the score needs tuning.
Score healthMQL to opportunity rate
opportunities ÷ MQLs, by cohortThe truest test of a score. Track by the month the lead qualified. Try the MQL to SQO calculator.
Bottom lineDisposition coverage
routed leads with a reason ÷ routed leadsWithout dispositions, nobody can improve the model.
Feedback loopUnassigned leads
qualified leads with no ownerShould be zero. Anything else is a buyer waiting for nobody.
Hygiene8Where it usually breaks
- Points for everything. When every click earns points, everyone eventually becomes an MQL.
- No decay. Old activity keeps people “hot” long after they’ve moved on.
- Leads that don’t match to accounts. A new contact at a customer or an open deal goes to a random rep instead of the account owner.
- A score nobody can explain. If reps can’t see why a lead qualified, they won’t trust it. Show the top reasons with the alert.
- Set it and forget it. Products, markets and campaigns change. Review the model at least every quarter.
Common advice vs. what I’d do
Keep going
Lead scoring and routing FAQ
How do you calculate an MQL?
Give points for fit (industry, company size, role, target account) and for engagement (pricing visits, webinars, repeat visits), subtract for disqualifiers, and set a threshold. A lead that crosses it becomes an MQL. Set the threshold from your own data on which leads became opportunities.
What is the difference between an MQL and a PQL?
An MQL qualifies through fit and engagement with marketing. A PQL qualifies through product usage, such as activation, inviting teammates or hitting limits, usually combined with fit.
What is lead routing?
Lead routing is the set of rules that assigns each qualified lead to the right owner, such as the account owner, a territory rep or a round robin, and notifies them with context.
What is a good speed-to-lead target?
For hand-raisers like demo requests, aim for minutes, not hours, during business hours. The right SLA depends on your team, but faster first responses consistently do better.
Should demo requests be scored?
No. Route them straight to a person. Scoring is for deciding who’s ready when they haven’t told you yet.
How often should a lead scoring model be updated?
Review it at least quarterly using conversion data and sales dispositions, and any time your product, market or main campaigns change.