Maestro Maestro for Marketing Maestro for Sales Maestro for Operations Playbooks Content Schedule Call
← Playbooks

The Lead Routing Playbook

Sales

Works with

HubSpot Salesforce Any form or chat tool Slack or Teams

Implementation kit

  • The lead source inventory, with the count each source keeps for itself
  • Qualification rules that score fit and intent apart, and the one-screen lead brief
  • Routing rules, with the assignment order and a named fallback owner
  • Review queue criteria, and the streak a lane needs before it runs unattended
  • The landed lead check, read daily against the source count
  • The routing read, weekly by lane and monthly by outcome
  • Five sample leads with an answer key, and a sample day where the check finds three misses

How do you know your lead routing worked yesterday? List every way a lead comes in, write the rules that pick an owner, and check each day that every lead reached a person.

The full method is below. An inventory of every lead source, fit and intent scored apart, routes and an assignment order with a fallback, a review queue for the close calls, a daily check against the source count, and a weekly read of whether routing is working.

A demo request sits in a spam folder for two weeks, and the dashboard shows a quiet month.

The failure this playbook fixes

How it works

01List every way a lead comes in

Routing rules only act on leads that reach them. Walk the site and the tools and list every source, including demo forms, resource forms, chat, meeting bookings, product sign-ups, shared inboxes, event lists and referrals.

For each source, write where the lead is first counted, how it travels to the CRM hop by hop, and who owns that connection. The count at the source is what every later check compares against.

Watch for a source with no count of its own. It cannot be checked, so fix that before writing a rule.

02Score fit and intent apart

Fit scores who the lead is, from role, company shape and a real identity. Intent scores what they did, from what they asked for, how often they returned and what they said in their own words. The two scores are never added into one number.

A signal you could not verify scores null, never zero. Zero means you checked and the answer was bad. Two or more nulls on the fit side send the lead to a person.

Every lead gets a one-screen brief that lists what was verified, what could not be, and the source of each fact.

03Write the routes and the assignment order

Four routes come out of the bands, and each one does a fixed thing.

  • Qualified, a named owner, a dated next step and an alert with the brief attached
  • Nurture, marketing emails and no sales outreach
  • Monitor, high activity from a company that does not buy, watched and never sold to
  • No fit, they get what they asked for and nothing else

Assignment rules are written in the order they are checked, and the first match wins. Existing account owner, then open deal owner, then named accounts, then territory, then round robin.

Watch for the last rule. Name a fallback owner for the lead that matches nothing, or it gets no owner and no alert.

If an AI model helps, it reads the form text and proposes a score. The route and the owner are set by the rules.

04Send the close calls to a person

Review is checked first, before any band. Conflicting evidence, an unclear duplicate, a score near a band edge and an identity that cannot be verified all go to a person, with the specific reason attached.

Every lane starts supervised. A lane runs unattended only after a streak of routes a person confirmed without changing, and one changed route sets the streak back to zero. No fit graduates last, because a buyer wrongly marked no fit never shows up in a report.

05Check every day that each lead landed

A routing failure rarely shows an error. Count the leads where they were first recorded, count them again in the CRM, and compare. A lead can stop in five places.

  • At the source, held or filed as spam by the form tool
  • Between tools, the connector reports success and the CRM has no record
  • In the CRM, a record that matched no rule
  • Routed, with an owner who is blank or inactive
  • Owned, and untouched past the time limit

Every lead in the gap gets a name and a reason before the read is closed. A source that normally sends leads and sent none gets a test submission the same day.

Watch for a check that starts from the CRM's own count. That compares the CRM with itself.

06Read weekly whether routing is working

Write down what "working" means before the first read. Then each week, read time from arrival to first touch, the share of qualified leads touched inside the limit, leads with no active owner, the age of the review queue and the routes a person changed, lane by lane.

Once a month, join each route to what happened next. A meeting that came from nurture, monitor or no fit is a lead the rules ranked too low.

Rules change at this read and nowhere else, each with the evidence and a date to check again.

Every lead counted at the source, and every one accounted for the next morning.

The implementation kit above packages the source inventory, the qualification rules and lead brief, the routing rules, the review queue criteria, the landed lead check, the routing read and a sample company with a day where the check finds three misses.

Maestro for Sales

Want This Running in Your Stack?

Magnetiz builds workflows like this one inside your tools, with your team trained to run them. Your first system is live in 30 days.

Schedule Call