A structured review that ends in a number, not an opinion. Two or three days in the room with the people who use the system, and you leave with what the gaps are costing you a year — on your own figures — and a phased plan to close them.
Every question asks for evidence — a number counted, a screen shown, a report run. "Were you happy with the data migration?" gets a shrug. "How many journals did you post to correct opening balances in month one?" gets a figure you can put in front of a board.
Nine hours a week correcting item data, at your own loaded rate, is £21,060 a year. Every line shows its arithmetic, so the conversation is about the rate rather than the total.
Scored separately. A system built for the wrong business needs money spent on software. A good system nobody has moved to needs money spent on people. Reading one as the other is the expensive mistake.
Each phase says what it takes, what it costs and what it removes. £11,900 of work against £35,100 a year of measured waste pays back in four months — and you can see how that was worked out.
Every one of these was diagnosed after go‑live, in the middle of the damage. Every one of them would have shown up in a review beforehand — the questions in this assessment exist because of jobs like them.
At go‑live the plant could not ship. Shipments would not consolidate by region, so nothing left the building and nothing was invoiced. Revenue stopped dead.
The implementation team had not understood the volumes, and had built manufacturing as though it were disconnected from sales and shipping. Reconnecting those three processes restored billing to expectation within weeks.
Constant shortages and a growing backlog of sales orders that could not be built. The system was capable; nobody trusted it, so everyone worked around it, which made it wrong, which proved them right.
The fix was alignment rather than software — organising the users so that what they did matched how the supply chain actually behaved. The backlog cleared, and the conversation moved to the sales team's door. Higher sales, and customers who got what they ordered.
Finite planning was in place and doing nothing useful. Capacity was planned on figures that did not describe the machines, so the schedule was ignored on the floor.
Correcting capacity planning and finite capacity scheduling raised throughput and labour efficiency, and released enough additional capacity to lift output by at least 30% — without buying a machine.
Clients are not named here. Ask on the call and I will tell you what I can.
Fixed, published, and inclusive of the report and the presentation of it. Travel and accommodation are charged at cost and agreed with you before anything is booked — UK, Europe or further; I have run these in Malaysia and China.
Priced and invoiced in pounds sterling by Oaksoft Services Limited, trading as Oaksoft Trading, a Hong Kong registered company — which is also why delivery in Asia is straightforward rather than exceptional. No VAT is charged on the invoice. VAT‑registered customers in the UK and EU account for it themselves under the reverse charge, in the normal way for a service bought from outside the territory. Payment by card or bank transfer; 30‑day terms on the full review, on receipt for the module check.
The work the review recommends is quoted separately, from the plan itself — priced in days, with your work and mine shown apart, so you can take on as much of it in-house as you want to.
Andrew Smith. Twenty-odd years of ERP delivery in manufacturing — Dynamics 365 Finance & Operations and SAP — on the floor as often as in the boardroom. Catch weight, batch and serial traceability, advanced warehousing, constraint-based scheduling.
The review exists because I kept being asked to fix implementations that nobody had properly diagnosed, and because "it's not working" is not a scope.
See the working demos on the main site → — the shop-floor tools are clickable, not screenshots.
Work delivered across the UK, Europe and Asia — foundries, machine shops, assembly, process and food. On site where it matters, remote where it does not.
No — that is close to the best moment. The people still remember what was promised, the workarounds have not yet hardened into habit, and the implementation partner is usually still reachable.
No. Older systems tend to show more recoverable cost, not less, because the workarounds have had time to become somebody's whole job.
The plan separates my work from yours explicitly, and prices only mine. Plenty of what comes out is work your own team should do, and it is labelled that way.
The review works on any ERP — the process questions are the same. The configuration-level questions are written for D365 Finance & Operations today; on another product that section is covered in discussion instead.
The people who use the system daily, plus whoever owns each process. Ninety minutes per module. It works badly with only the project team and well with the supervisors.
Thirty minutes, no charge, and you will know by the end of it whether this is worth doing at your site.