Technician Closure — TGL vs Marketed

About this build. Close rate per technician, split by whether the work came from a tech-generated lead or a marketed one. A tech who converts well on their own finds but stalls on marketed calls needs different coaching than one who does the reverse. Figures are live: recomputed in Peaka every six hours over a rolling twelve months.

How each number is built

Which jobs count as a technician's

A job is attributed to a technician when that technician has payroll timesheet time recorded against it. The appointment-assignment tables in this dataset are truncated at 5,000 rows and stop in 2018, so timesheet records are the only reliable technician-to-job link available. Coverage is good — nearly every non-canceled job has at least one technician attached.

Close rate

Jobs closed divided by jobs where an estimate was presented. Jobs with no estimate at all are excluded from the denominator, so this measures closing, not option-giving. The estimate rate column shows how often a technician presents anything in the first place.

Selling tech vs everyone on the job

Selling tech credits the close only to the technician named on the sold estimate — clean for commission and coaching, but a helper on a two-person job gets no credit. Everyone on the job credits every technician who logged time. Revenue under that model is counted once per technician, so department and company totals will overstate; use the selling-tech model for any revenue that has to reconcile.

Sales and install are separate

They are different jobs, so pooling them is misleading: on an install the sale has usually already happened, and an install technician who logs time against a job with a sold estimate on it is not the person who closed it. Selecting Install shows technicians on their install work only; Sales shows the selling side. Sales + install together restores the pooled view if you want it. A technician who works both simply appears in both, with only the jobs from that side counted.

Drilling in

Click any technician — on the board or in the table — to open every job behind their numbers, filtered to the same dates, departments and attribution model showing above. Click a job row to expand its estimates. The drill-down covers jobs where an estimate was presented, which is the close-rate denominator; jobs with no estimate at all are counted in the Jobs column but have nothing to drill into.

ServiceTitan links

Job and estimate numbers open the record in ServiceTitan in a new tab. The URL patterns are editable at the top of the page. Job uses the confirmed pattern #/Job/Index/{id} — ServiceTitan is inconsistent here, so the /Index/ segment is required for jobs even though Invoice and Opportunity have no equivalent. The estimate pattern is still a guess. If a link misses, paste the working URL with its ID swapped for {id} and every link on the page updates.

Apprentices

Excluded by default. An apprentice is anyone carrying the Apprentice role in ServiceTitan. They log time on a small share of job-technician pairs but account for very little credited revenue. Under the everyone-on-the-job model they were mostly posting 0% close rates on work they assisted rather than sold, which dragged the board down. Switch the control to Included to put them back. This keys on the role, not the position field: a few people carry a Helper position under a full technician role and are counted as technicians here.

Tech-generated vs marketed

Set by the job the technician was standing on: tech-generated if that job traces back to an earlier job through Job Generated Lead Source, marketed otherwise. The two never overlap.

Lead generation

Credited to the employee recorded on the lead source, not to whoever later ran the job. Lead revenue is the sold value of the job the lead became, so it lands on the finder rather than the closer.