Read a Repair Analysis
Once a vehicle has been handed over, BayWise can produce an analysis of the visit: where the time actually went, what got in the way, and what would have made it faster.
It is an analysis, not a summary — it is worth reading on the jobs that went badly and on the ones that went unusually well.
Who can do this: Service Advisors, Workshop Controllers, Location Managers, Org Admins, Account Owners and Read Only users. Requires a plan that includes AI-assisted features.
Finding it
Open the vehicle from the board or from search, and find the visit. The analysis sits on the completed repair order.
It is available after handover. A job still in progress does not have one, because the story is not finished.
→ Find a vehicle and its history
What it contains
The numbers
| Metric | Means |
|---|---|
| Estimate accuracy | How the actual time compared with what was estimated |
| Time worked | Time actually spent on the vehicle |
| Time in the shop | Total elapsed time from start to handover |
| Time waiting | How long the vehicle sat paused |
| Drying and setting | Cure time, where the work involved it |
| Time spent working | Worked time as a share of elapsed time |
| Early / late | How the handover compared with the promise |
| Worst late start | The largest gap between a stage’s planned and actual start |
| Steps timed | How much of the job has reliable measurement behind it |
The gap between time worked and time in the shop is usually the most revealing number on the page. A vehicle that took four days and eleven hours of work spent three days waiting for something — and that is where the improvement is, not in the eleven hours.
The narrative
- Delays and issues — what actually held the job up, drawn from the reasons the floor recorded
- Biggest loss — the single largest contributor to the elapsed time
- What went well — worth reading; a job that recovered from a bad start usually did so because somebody did something specific
- Lessons learned and Recommendations — what to change
- Step insights — stage by stage
- Technician performance — for the people who worked it
- Data quality notes — where the record was incomplete enough to weaken the analysis
Why the data quality note matters
An analysis is only as good as what the floor recorded. If stages were started late in the app, delays went unreported, or pauses were never entered, the analysis says so rather than confidently analysing a fiction.
A run of analyses carrying data-quality notes is a signal about your check-in habits, not about the analysis. It is worth acting on before you act on anything the analyses say.
Using it
On a job that went badly. Read it before the conversation with the customer or the insurer. “The vehicle waited nine days for a panel that was ordered on day two” is a different conversation from “sorry it took so long”.
On a job that went unusually well. Rarer and more valuable. What made it fast is usually repeatable.
On a pattern. Several analyses of the same job type naming the same cause is a process problem you can fix once.
Santos Body Works, São Paulo read six analyses of medium collision repairs and found the same biggest loss in five of them: waiting for a fitment slot after parts arrived. Parts were being ordered promptly and then sitting. The fix was a scheduling habit, not a supplier change.
Common questions
There is no analysis for this job. It is produced after handover. A cancelled job does not get one.
It says no work was recorded on this repair order. Nothing was started or completed against it in BayWise — usually a job that was cancelled, or one where the floor never used the board.
Is it accurate? It works from what was recorded. Where the record is thin, it says so in the data-quality notes rather than filling the gap.
Can I share it with a customer or an insurer? It is written for internal use and names technicians. Use the facts from it rather than sending it as it stands.
Why do we need it when we have analytics? Analytics tells you what is happening across many jobs. The analysis tells you what happened in one. You need both, and they answer different questions.