adsoraLimit the variables you change at once
Choose a manageable combination of trade, market and receiving team. Testing a new source across every branch while replacing the dialer and changing the qualification rules makes the results difficult to explain. A narrower first group gives the people running the test a chance to inspect individual records.
Write the project definition, approved ZIP list, required fields, accepted delivery hours and purchase limit. Name one launch owner on each side. The buyer also needs a branch or contact-center owner who can confirm that accepted inquiries are reaching people who can work them. Keep expansion outside the initial scope until someone approves it.
Use separate delivery and sales review dates
| Example period | Work to complete |
|---|---|
| Before launch | Approve the source, terms, fields and test results |
| First business day | Trace accepted records into the assigned CRM queue |
| End of week one | Review rejects, duplicates, contact attempts and branch capacity |
| Weeks two and three | Inspect project fit and appointment progression by source and branch |
| Week four | Review mature outcomes, pending records and the next buying decision |
This is an example operating schedule, not a promise that four weeks will establish sales performance. Some projects will still be open. Keep later outcome reviews on the calendar, especially where the time from inquiry to installation is longer than the initial test. Delivery problems can be assessed much earlier than completed-sale results.
Prove the route before judging the source
Use agreed synthetic records to test an accepted lead, an out-of-area inquiry, a missing required field and a duplicate. Confirm the final response, the record in the CRM and the assigned queue. Also test a pause and an after-hours condition. Record expected and actual results, then resolve differences before opening intake.
On the first day of live delivery, compare accepted source IDs with received CRM records. Investigate an accepted lead that never reached the contact team immediately. A buying platform’s acceptance response does not show that a downstream assignment rule worked.
Keep pending work visible in the scorecard
Count accepted leads, assigned records, contact attempts, two-way conversations, appointments set and appointments held using written definitions. Keep credit requests and approved credits separate. For each stage, show both the numerator and denominator so a percentage cannot hide a small group.
Group records by their delivery date and allow comparable time to progress. A lead received yesterday should not be judged against one received three weeks ago as though both had the same chance to produce a held appointment. Show open opportunities and unworked records alongside closed outcomes. Avoid calling early differences statistically reliable when the sample and observation period cannot support that claim.
End the review with a recorded change
Give each issue an owner. A repair request in a replacement-only program may require a project filter. A lead waiting in an unassigned queue requires a routing fix. A pricing question belongs with the commercial owner. Send examples with source IDs through the agreed reporting channel, keeping homeowner information out of general email threads.
Record the decision, its reason and its effective date. If you expand, state which markets, caps or hours change. If you pause, define what evidence will allow a restart. The next report should be able to distinguish the original pilot from the revised program.