The first week of the month
Account managers spend the start of each month building slides. They export dispatch data from the WMS, filter by client, count orders dispatched after their promise, pull complaint logs to estimate pick accuracy, and paste charts into last month's template. Each client measures slightly differently: one counts dispatch against a two o'clock cut-off, another against same day for orders received by noon, a third excludes orders on hold for stock.
The numbers go out, and the client's operations manager challenges one of them. They have counted differently, and the next half hour of the review is spent reconciling instead of improving anything.
Why the pack is hard to automate by default
The WMS records the timestamps, but SLA definitions are written in contracts, not in the system. Exclusions matter as much as the base calculation: orders held at the client's request, stock not received, address problems, carrier failures after collection. Without those rules applied consistently, the figure is either unfairly bad or not trusted.
| KPI | The definition that has to be agreed |
|---|---|
| Dispatch on time | Cut-off per client, service type, working days, exclusions for holds |
| Pick accuracy | Counted from complaints, returns reasons or pack checks |
| Receipt to stock | From arrival, from booking, or from completed count |
| Returns processed | From receipt at the dock to graded and updated |
| Stock accuracy | From cycle counts, and at what tolerance |
What the manual pack costs
Account manager time every month, multiplied by the number of clients. Errors from copy and paste. And the bigger cost: service reviews that argue about the figure instead of discussing the causes, which means the same misses happen next month.
A late or inconsistent pack also signals a warehouse that does not have a grip on its own performance, even when it does.
Hand-built packs also hide trends. Each month is assembled in isolation, so a slow slide in receipt times over a quarter, or pick accuracy that dips every time a particular client runs a promotion, is only visible to someone who lines the old slides up side by side. Nobody has time to do that, so the pattern is spotted by the client first.
And when a contract comes up for renewal, you want a year of consistent evidence of the service you delivered. Twelve packs built by different people with slightly different filters are not that.
What we build
- A data feed from your WMS of order, receipt, return and count events with their timestamps.
- Per-client SLA definitions held as settings: cut-offs, working days, services, exclusions.
- Miss reasons recorded at the source where possible, such as stock short, address hold or client hold, and inferred from the data where not.
- The KPI pack generated each month per client, in your branded layout, with trend charts and the list of misses with reasons.
- A live version of the same figures for account managers during the month, so problems are seen before the review.
- Optional client access to their own dashboard, if you want them to see it continuously.
When a client changes their cut-off or adds a service, the definition is updated once and every future report follows it.
What reviews look like afterwards
The pack arrives on the first working day without anyone building it. Both sides agreed the definitions when they were set up, so the discussion starts with the misses and their causes. Account managers spend the time they saved on the clients who need attention.
Signs your KPI reporting is manual
- Account managers spend days building monthly reports.
- Each client's SLA is calculated in its own spreadsheet.
- Clients challenge the figures in service reviews.
- Misses are counted, but reasons are not recorded.
- You find out about a bad month when the report is built.