Every Lima engagement starts by measuring where you are, so the change can be shown afterwards. This page collects the results so far: case tiles from customer engagements, anonymised until each customer agrees to be named; figures from the showcase estate; the method used to measure; and samples of the two reports every engagement produces, the quarterly delta report and the reconciliation report.
Anonymised until a customer agrees to be named; specific enough to be checked with the account owner. Adding a name is a content edit.
The bridge from the ERP into the planning application carries more than a thousand tables. Integration cost came out about 70% lower than a custom build on the same program.
Eight hundred tables from SAP into the planning application, kept current on the planning calendar.
Planning data delivered on the calendar since 2022; the loop returns approved plans to the ERP.
The longest-running planning integration on the platform.
Re-routing decisions made on live order, stock and carrier data: 23% higher container utilisation and 11% shorter lead times on the routes in scope.
An agent that proposes slotting and capacity moves for people to approve, saving about €300,000 a year in the warehouses in scope.
Excess and slow-moving stock identified against live demand and disposed of through approved actions.
Allocation proposals written back to the ERP automatically once the rule had earned that right.
ERP data into Kinaxis and approved plans back into the ERP, each write confirmed, in production since the fourth quarter of 2021.
A result is only worth showing if the number before it is written down. This is how every engagement does that.
| KPI | Plain meaning | Read from |
|---|---|---|
| OTIF | Order lines delivered on the promised date, in full | sales orders, deliveries |
| Days of inventory | How many days the stock on hand would last | stock, consumption |
| MAPE | How far the forecast was from what happened, as a percentage | forecast, actuals |
| Perfect order | Orders with no error in date, quantity, damage or paperwork | orders, deliveries, claims |
| Time to resolve | From an anomaly firing to the fix in place | incident record |
| Run-window adherence | Loads that met their deadline | run reports |
| Reconciliation variance | Rows and totals that differed between source and target | reconciliation reports |
| Hours to first load | From kick-off to the first full planning dataset | engagement record |
The sample below is redacted and uses the showcase estate's figures. The structure is the one every customer receives.
| KPI | Baseline (4 wks pre go-live) | Latest quarter | Change | Engagement | Attribution |
|---|---|---|---|---|---|
| OTIF, key accounts | 87.2% | 89.6% | +2.4 pts | control tower · carrier exceptions | high |
| Time to resolve a data incident | ~15 h | 44 min | −14 h | orchestration · auto-hold and replay | high |
| Run-window adherence, planning loads | 91% | 100% | +9 pts | Backbone · run windows | high |
| Reconciliation variance, planning loads | n/a | 0 rows · 3 held | reported | Backbone · checks | n/a |
| Compute cost, cache savings | £0 | £1,840 / month | saved | Data Fabric · cache | high |
| Forecast accuracy (MAPE), 7-day | n/a | 88% accuracy | tracked | ChatLima forecasts | medium |
Attribution says how confident the team is that the engagement caused the change: high when the KPI is read from the pipeline itself, medium when other changes were in flight. The baseline 'about 15 hours' for incident resolution is the industry figure used where no internal baseline existed.
The report a planner, an auditor or a migration lead uses to accept a load. Rows in, rows out, control totals and variance per table, and the exceptions with their owners.
| Dataset | Source rows | Loaded | Variance | Control total | Status |
|---|---|---|---|---|---|
| Item masterMARA · MARC | 2,914,220 | 2,914,220 | 0 | — | Reconciled |
| BOM componentsSTPO · STKO | 18,406,113 | 18,406,113 | 0 | qty 1.20 × 10&sup9; | Reconciled |
| Open purchase ordersEKKO · EKPO | 1,204,881 | 1,204,878 | 3 → owners | value matched | 3 exceptions |
| Sales ordersVBAK · VBAP | 6,102,447 | 6,102,447 | 0 | qty matched | Reconciled |
| Forecastexternal file | 412,006 | 412,006 | 0 | — | Reconciled |
| Planned orders returnedBAPI_PLANNEDORDER_CREATE | 41,207 | 41,207 confirmed | 0 | — | Confirmed |
Each aggregate is traceable to the engagement reports behind it; the account owners confirm each program before a figure is quoted outside this page.
Named stories are added here as customers agree to be quoted. The section is designed so that adding one is a content edit.
Planning bridge and control tower. Name and quote pending the customer's approval.
Planning integration since 2020. Name and quote pending approval.
Dynamic network routing agents. Name and quote pending approval.
Review badges appear on this page when ten or more verified reviews exist on the public review sites. Every first-load and pilot customer is asked for one at go-live.
Because their approval to be named is pending. Each tile is specific enough to be checked with the account owner, and a name is added as a content edit when agreed.
Lima does, from your own systems, with definitions your team agreed. No survey and no estimate.
Yes. Both reports are data products in Lima; they can be read through the API or landed in your warehouse.
A demo that starts with your KPIs and shows where the baseline would come from.
