How the research works
We review a defined set of estimates and trace what happened after each one. Here is what that review covers.
This page describes an agreed client record review. The published evidence that informs the method, including source selection and limitations, is in our research source register.
Define the population and the period
Written replacement estimates over a stated window, usually a recent quarter. The window, estimate types, and locations are agreed in writing before any record is read.
Review the records and the handoffs
Read estimate records and their attached activity. Distinguish a missing reason from one recorded somewhere unexpected, a vague note, or a stale update. Missing documentation is not proof that a conversation never happened.
Classify what happened after each estimate
Keep outcome, reason, and next action as separate fields. A customer can be undecided because of timing, have a later callback agreed, and still be unsold. Definitions, duplicate proposals, and the counting unit are agreed before counting starts.
Separate documented facts from assumptions
Separate the recorded event, the customer’s stated reason, the employee’s interpretation, and any status set by a system rule. Preserve the source and date. A customer statement is evidence of what they said, not independent proof of every underlying cause. An automatic status change is not a homeowner’s explanation.
Examine what reaches the next person
Within the agreed Diagnostic staff interviews, check whether the next responsible person can find the last customer position, prior contact, owner, and next action using their normal account and device. Distinguish missing information from access restrictions or a record stored elsewhere. Separate routine searches from decisions that need professional judgment.
Identify unresolved patterns and their value
Group supported reasons and keep unknowns visible. Look for repeated searches, duplicate contact, unclear responsibility, or questions that need an answer. Quote amounts provide context; alternatives for one decision are not separate potential jobs. Consider the cost of a change, contribution, and available crew time before assigning financial value.
Discuss the findings and decide what makes sense next
Review the findings with the owner and the people doing the work. Identify a responsible person, a practical change, and a way to measure it. Doing nothing further is a legitimate conclusion.
What the research will not do
- The research does not assume there is a problem. The existing process may already be working.
- No recovery rate, revenue figure, or job count is quoted in advance.
- Quoted unsold volume is not lost revenue and is not presented as revenue that would have been earned.
- The research does not assume any individual was at fault.
- Where the record is silent, the finding is that the record is silent.
What the research needs from a shop
Read access to estimate records for the agreed window, or an export, plus the activity attached to them. Gross margin figures where they exist. No system is replaced or migrated.
Customer records stay the shop's property, are used for the agreed research only, and are never shared with another shop. More detail is on the privacy and confidentiality page.
What the output looks like
An illustrative example of the shape of the findings, clearly labeled and built from constructed numbers, is on the what you receive page.