Performance and accessibility
Annexure E, Criterion 7 requires a live KPI dashboard in production covering task success, completion, satisfaction, effort and accessibility. This is that dashboard, running against the prototype.
Where these numbers come from
Prototype measurement, not production RUM. Figures come from scripted task runs and Lighthouse passes on this build, sampled 15–21 September 2026. In production the same panel reads from real-user monitoring and the service desk.
- Sampling window
- 15 to 21 September 2026
- Method
- Scripted task runs, Lighthouse and axe-core on this build
- Not shown
- Real-user monitoring, which only exists once the estate is live
The five service KPIs
Each figure is compared against the previous seven-day period, with the measurement method stated so the number can be argued with.
- Task success rate
94.2%of tasks
up 2.6 percentage points from 91.6% in the previous period
Target 90% or above
12 scripted tasks, 40 runs each, across five site groups
- Completion rate
88.7%of tasks
up 3.6 percentage points from 85.1% in the previous period
Target 85% or above
Share of started journeys reaching the confirmation screen
- Customer satisfaction
4.4out of 5
up 0.2 points from 4.2 in the previous period
Target 4.0 or above
Post-task rating, 1 to 5, 312 responses
- Customer effort score
2.1out of 7, lower is better
up 0.5 points from 2.6 in the previous period
Target 3.0 or below
'How much effort did that take?', 1 to 7, lower is better
- Accessibility score
98out of 100
up 2.0 points from 96 in the previous period
Target 95 or above, zero critical
Automated axe-core sweep plus a manual keyboard and screen-reader pass
Task success rate over eight weeks
One series, weekly, from the same twelve scripted tasks
| Week beginning | Task success rate |
|---|---|
| 28 Jul | 86.4% |
| 4 Aug | 87.9% |
| 11 Aug | 88.2% |
| 18 Aug | 90.1% |
| 25 Aug | 89.4% |
| 1 Sep | 91.6% |
| 8 Sep | 92.8% |
| 15 Sep | 94.2% |
Open accessibility violations by severity
37 open across all 90 sites
| Severity | Open violations |
|---|---|
| Critical | 0 |
| Serious | 2 |
| Moderate | 11 |
| Minor | 24 |
- Critical, 0None open across the estate
- Serious, 2Missing alt text on two sub-centre gallery images
- Moderate, 11Heading order breaks on legacy imported pages
- Minor, 24Redundant link titles in footer link sets
Core Web Vitals
Measured on the slowest page of each site group, not the fastest
Largest Contentful Paint
Target 2.5 s or less
1.8 s72% of the budget used
Slowest page: centre directory at 2.3 s
Within targetCumulative Layout Shift
Target 0.1 or less
0.0220% of the budget used
Worst page: news listing at 0.04
Within targetInteraction to Next Paint
Target 200 ms or less
142 ms71% of the budget used
Slowest interaction: tender filter at 186 ms
Within target
Availability against the SLA
Rolling 90 days
99.82%Within target
The contracted floor is 99.5%. Measured availability is 0.32 points above it, which is 233 minutes of downtime in ninety days.
Where the SLA sits on the scale
99.00%100.00%
The red mark is the 99.5% SLA floor; the filled band is measured availability.
Incidents by priority
14 raised in the same ninety days
| Priority | Open and closed | What it means | Response and resolution |
|---|---|---|---|
| P0 | 0 | Estate-wide outage | Respond in 15 min, resolve in 4 h |
| P1 | 1 | A site or a core service down | Respond in 30 min, resolve in 8 h |
| P2 | 4 | Feature degraded, a workaround exists | Respond in 4 h, resolve in 3 days |
| P3 | 9 | Cosmetic or content defect | Respond in 1 day, resolve in 10 days |
Accessibility score by site group
Automated axe-core sweep, averaged within each group, out of 100
National portal
1 site
99 out of 100
Directorate microsites
12 sites
98 out of 100
Sub-centre microsites
58 sites
96 out of 100
Centres of Entrepreneurship
12 sites
97 out of 100
Scheme portals
7 sites
95 out of 100
Shown as labelled meters rather than a bar chart: the scores sit between 95 and 99, and bars drawn from a zero baseline would be indistinguishable while bars drawn from 95 would exaggerate the differences.