How do you measure a B2B software redesign?
Measure a redesign by the tasks it was meant to improve: time to complete key workflows, errors and corrections, support requests per screen, onboarding time, and the share of work still done outside the product.
Collect a baseline before the redesign starts. Without it, results rest on opinion.
For example, choose one task and role, measure completion time and error rate over a defined sample before release, and repeat with comparable cases afterwards. Separate required exports from workarounds and record other product or process changes that may affect the result.
Key facts
- Measures follow the tasks the redesign was meant to improve.
- A baseline is collected before the redesign starts.
- Satisfaction alone is a weak measure.
- Periods with different volumes cannot be compared directly.
- Investigate the purpose of exports and spreadsheets before treating them as avoidable work outside the product.
- Business results such as new clients and revenue depend on more than the interface, and are reported alongside task measures.
Measures
| Measure | How to collect |
|---|---|
| Task time for key workflows | Session recordings or timed tests |
| Errors and corrections | Correction records, adjustment reasons |
| Support load | Tickets tagged by screen |
| Onboarding time | Time to the first completed workflow |
| Work outside the product | Exports and spreadsheets in use |
Product and business measures
Vendors also track measures that reflect the interface together with sales, pricing and the market: new clients, engagement, revenue, and how often prospects mention the interface as a reason to buy. They show the business effect and cannot isolate the redesign’s share of it.
The baseline comes from the same sources that will be used after release: the same task definitions, the same ticket tags, the same period length. Changing how a measure is collected between baseline and release makes the comparison meaningless.
Common mistakes
- Releasing without a baseline
- Measuring satisfaction only
- Comparing periods with different volumes
- Attributing all business growth to the redesign
Where we’ve worked on this