How to build lead routing that assigns in seconds

What matters here

• Match before you route. A lead that belongs to an open opportunity already has an owner, and routing it by territory hands a live deal to a stranger.
• Rules evaluate top down and stop at first match. A rule set where two rules can both fire is a rule set nobody can debug.
• The fallback queue should be empty. When it is not, that is a rule waiting to be written, and it is the most useful number on the page.
• Measure arrival to first touch, not assignment to first touch. Queue time is the part you can actually fix.
• Stamp which rule fired on every lead. Without it nobody can answer why a lead went where it went, and the rules never get corrected.

Who this is for

You run revenue operations, marketing operations or sales operations. You already have rules somewhere, probably in a Salesforce assignment rule nobody wants to touch. What you want is routing you can read, change without a deployment, and defend when a rep asks why they did not get a lead.

How it works in practice

What happens between a form being submitted and a rep picking up the phone.

1. 1

   A form fill lands and enrichment runs

   Company, size and industry from the email domain. If enrichment times out, routing proceeds on what is there instead of holding the lead.

2. 2

   The lead is matched to an account before any rule fires

   Domain, then fuzzy name, then parent hierarchy. An open opportunity on that account short-circuits everything below.

3. 3

   Rules evaluate top down and stop at first match

   Named account, open opportunity, territory, round robin. Which rule fired is stamped on the lead.

4. 4

   Assignment posts to the rep where they already work

   With the enrichment and the matched account attached, so the first touch does not start with research.

5. 5

   Anything unmatched lands in the fallback and alerts

   Not a null owner, not a silent drop. An empty fallback is the goal and a non-empty one is the signal.

6. 6

   Ops changes the rule table without a deployment

   The same workflow opens in Tray Build, so a territory change is a visual edit rather than a ticket.

Paste these into Claude Code or Codex with the Tray Headless plugin installed. Each stage runs on its own. The systems named in them are the worked example rather than a requirement, and every prompt says so.

Once per project, run /tray-workflows:set-workspace to pick the workspace these build in. Point it at a sandbox first.

1. 1

   Set up, and see what is actually connected

   Before anything is built, confirm the ground truth.

   Headless skills build-workflow

   Copy

   Use build-workflow. The systems in play are Salesforce, Marketo,
   Clearbit, Slack, Outreach, Workday and Snowflake, or whatever we run
   in those seats. Before you plan anything, tell me which of them are
   already authenticated in the workspace, because I do not want a
   connector stubbed that I have not authenticated. I expect Salesforce,
   Marketo and Slack.
   
   Tell me what is missing before we build anything, and do not stub a
   connector I have not authenticated.

2. 2

   Match the lead to an account first

   The step that decides whether routing is right instead of merely fast.

   Headless skills build-workflow

   Copy

   Work against Salesforce so the query uses the real operations and
   field names, then build the matcher in this order:
   
     1. Exact match on email domain against Account.Website
     2. Strip generic domains first: gmail, outlook, yahoo, hotmail, icloud
     3. Fuzzy match Lead.Company against Account.Name at 0.85 similarity
     4. Walk Account.ParentId up two levels for subsidiaries
     5. Check for an Opportunity on that Account where StageName is not in
        Closed Won or Closed Lost
   
   If there is an open opportunity, the lead goes to that opportunity owner
   and skips every rule below. Write the matched AccountId back to the lead
   so nobody has to redo this later.

   Stripping generic domains first is not a nicety. Without it every Gmail signup matches whichever account happens to have gmail.com in a website field.

3. 3

   Build the rule chain, in order, stopping at first match

   The order is the design. Write it so somebody can read it.

   Headless skills build-workflow

   Copy

   Use build-workflow. Trigger on Lead created where LeadSource is not null.
   Rules evaluate top down, first match wins:
   
     1. Named account list, owner from Account.OwnerId
     2. Open opportunity, owner from the matcher above
     3. Territory: Country then State, from a rules table I can edit
     4. Segment round robin: under 200 employees to the SMB pool, 200 to
        2000 to mid-market, above that to enterprise
     5. Fallback queue
   
   Set Lead.OwnerId, stamp Routed_Rule__c with which rule fired, and post to
   the rep in Slack with the enrichment and the matched account.
   
   Never leave the fallback out. The queue that should be empty is the one
   that tells you a rule is wrong.

4. 4

   Handle the five things that actually break it

   Every one of these has quietly lost somebody a lead.

   Headless skills tray-gotchas tray-patterns

   Copy

   Use tray-gotchas, then handle these explicitly:
   
   Rep is out of office. Read the out of office flag and route to their
   deputy rather than assigning to somebody who will not look until Monday.
   
   Same lead submitted twice inside a minute. Dedupe on lower(email) before
   routing, not after.
   
   Enrichment times out. Route on what you have. Do not hold a lead in a
   queue waiting for a firmographic field.
   
   Territory with nobody assigned. Fall through to the fallback and alert.
   Never assign to a null owner.
   
   Lead arrives before its account exists. Create it, leave it unmatched,
   and re-run the matcher nightly.

5. 5

   Time it from arrival

   Because the queue is the part you can fix.

   Copy

   Stamp three fields: Arrived__c on creation, Assigned__c when OwnerId is
   set, and First_Touch__c on the first logged Task or Event.
   
   Speed to first contact is arrival to first touch, not assignment to first
   touch. The gap between those two is the queue, and the queue is the part
   you own.
   
   Report median and 90th percentile by rule, by segment and by rep, plus
   the count that hit the fallback. If the fallback is not empty, that
   number is the most useful one on the page.

6. 6

   Validate it, then hand the rules to ops

   The step that decides whether this survives its first territory change.

   Copy

   Run the per-step schema checks and the whole-workflow audit before this
   touches production.
   
   Then open the same workflow in Tray Build so revenue operations can change
   a territory or add a named account in the visual canvas, without coming
   back to a coding assistant. Same workflow, same governance, same audit
   trail underneath.

   This is the step that decides whether the thing you built is still yours in six months, or whether every territory change becomes a ticket.

What it connects to

Routing sits between where leads arrive and where reps work, so it reads from both ends.

Named systems are the ones most teams run, not the only ones that work. Each is an authentication in your Tray workspace, referenced by name, so the workflow never holds a credential. Where we have a connector page, the name links to it.

Questions people ask

Further reading

Background on the same subject, for the case rather than the build.

• How to automate lead routing to win more deals
• Lead routing and scoring simplified: From data to ROI

Related guides