The situation
An urban co-operative bank with ninety active vendors ran its annual third-party assessment out of a shared workbook. Two people owned it. The cycle ran eleven weeks, most of which was spent chasing attachments over email.
What changed
| Before | After | |
|---|---|---|
| Cycle time | 11 weeks | 19 days |
| Vendor response rate | 54% | 91% |
| Evidence chased by email | Every vendor | Exceptions only |
| Findings with a named owner | Unknown | 100% |
The response-rate change did the real work. Vendors were not refusing to answer; they were losing the request in an inbox. A portal with its own reminder schedule removed the step that was failing.
We did not make the assessment shorter. We stopped losing it in email.
What we would do differently
Tiering should have come first. The first cycle sent the full questionnaire to all ninety vendors, including the ones whose failure would have been an inconvenience rather than an incident. The second cycle sent a nine-question version to forty of them and finished a week earlier.
Facts last verified 18 Aug 2026
