Lead Routing and SLA Timers With a Replayable Log

The number that matters first: Not response time. Misrouting. In the routing audits I have run, between 8 and 20 percent of inbound leads reach an owner who was never going to work them, and almost none of those get corrected. The lead simply ages in a queue nobody reads. Fixing that is worth more than shaving minutes off the fast path, and it is the first thing I measure. Routing is a policy, not a canvas: The deliverable that outlives the build is a document: territory definitions, tie-breaks, capacity rules, what happens on a Friday at 17:30, and who signed each one. Your CRO signs it. Then it gets encoded. Rules that exist only inside a workflow builder are rules nobody reviews, and I have not seen a builder canvas survive two years of staff changes intact. Every routing decision leaves evidence: Rule version, the field values as they stood at decision time, the owner chosen, the owners rejected and why, timestamp. Written to your warehouse. When a rep says the system gave them a bad lead, you replay the decision instead of arguing about it. This is what vendors do least well and what actually ends territory disputes. Territory and assignment: Built on billing country, employee band and product line, which are attributes with an owner and a source, rather than on whoever claimed the account first. Round robin with real capacity: out of office, Irish public holidays, UK and US timezone coverage, ramping reps at partial weight. Reassignment when somebody leaves is a scripted operation rather than a weekend of manual work. SLA timers, started in the right place: The first-touch clock starts when the record is created, not when it is assigned. The gap between those two events is where most leads die and where most dashboards still look healthy. The escalation ladder is agreed with sales leadership: typically manager at 30 minutes, back to the pool at two hours, different rules outside business hours. Breaches go to a weekly review with names attached, or the timer is decoration. Build or buy: In week one I tell you whether Chili Piper, Default or LeanData expresses your ruleset without contortion, or whether it wants a small service running against the HubSpot or Salesforce API. The answer depends on how much of your logic the vendor's model can hold natively. I take no commission from any of them. Not included: Meeting scheduling widgets configured as an end in themselves. Speed-to-lead dashboards that measure assignment rather than human contact, because those flatter everybody. Round robin across reps whose territories are still politically unresolved. Commission and quota logic. Support ticket routing. Who this is not for: Teams that have not dealt with duplicates. A duplicate is two owners, two SLA clocks and a rep phoning a prospect a colleague called an hour earlier. And teams who want a rule that quietly favours one senior rep: such a rule can exist, but it goes into the document with that person's name against it, or it does not go in at all. If that is unacceptable, we are not going to enjoy working together. What ongoing looks like: Routing decays. Reps join, territories split, a product line moves. On retainer I hold the policy document, make each change through the same review, run a quarterly rebalance against actual load rather than headcount, and bring the breach review to your sales meeting. If nothing changed in a month, the report says so in two lines and I do not invent work to fill it.