The Thumbtack response lag fieldbook

A practical workbook for measuring delay, finding the bottleneck, and handing leads off safely

For service team operators

Start with one clock

Record five moments for every lead

1. Lead received
2. Alert delivered
3. First review
4. First reply
5. Human owner assigned

Do not collapse these into one "response time" number. Each gap points to a different fix.

The lag ladder

Segment	Calculation	What it tests
Alert lag	alert delivered - lead received	Notification path
Review lag	first review - alert delivered	Coverage and attention
Reply lag	first reply - first review	Decision and drafting
Handoff lag	owner assigned - first reply	Ownership clarity

Use timestamps from one system of record. Keep time zones explicit.

The shift log

Copy this row for each lead

Field	Entry
Lead ID
Received at
Alert at
Reviewed at
Replied at
Assigned at
Route	automated / human / mixed
Exception
Outcome

Measure the workflow, not the person.

Separate normal work from exceptions

Tag before you compare

• New inquiry
• Duplicate or spam
• Outside service area
• Missing job details
• Safety or compliance concern
• Price or scope exception
• Existing customer follow-up

Averages become misleading when exceptions sit in the same bucket as routine leads.

Read the pattern

Fast alert, slow review

Coverage gap. Check who owns the notification window.

Fast review, slow reply

Decision gap. Check missing rules, prices, or job details.

Fast reply, slow assignment

Handoff gap. Check ownership and acceptance.

Every stage slow

System or staffing issue. Test delivery before changing scripts.

The safe handoff card

A handoff is complete only when the next owner has:

• Lead ID and contact channel
• Customer request in plain language
• What was already sent
• Missing facts and open questions
• Agreed next action
• Acceptance timestamp

No named owner means the handoff is still open.

Daily review, ten minutes

1. Count leads with all five timestamps.
2. Find the slowest segment by median.
3. Review the three longest delays.
4. Check whether exception tags explain them.
5. Pick one workflow change for tomorrow.
6. Set a date to judge the change.

Do not rewrite the whole process after one unusual lead.

Change log

Date	Bottleneck	Change	Expected signal	Review date

Stop rule

Keep the change only if the intended lag segment improves without creating missed handoffs or inaccurate replies.

One clean operating rule

Measure each gap. Tag exceptions. Require acceptance.

This workbook was prepared by the Auto-Respond Team. Auto-Respond supports configured lead-response workflows for Thumbtack and other supported lead channels.

For workflow context, see Thumbtack auto responder.