Anonymised · Real estate and mega-development · Jeddah, Saudi Arabia · Multi-year engagement
A large-scale mixed-use development in Jeddah had grown past the point where its finance, procurement and facility operations could be run on spreadsheets and goodwill. It had never had an enterprise system.
The decision record
- What it looked like
An organisation with no enterprise system needed one.
- What it actually was
An organisation that had never been forced to agree what its own words meant.
- The decision
Establish the definitions before selecting the software.
Argued for before selection. Several people reasonably felt it was delay.
The situation
The organisation was operating a residential and commercial development with thousands of concurrent users across its assets, and no enterprise system underneath it. Finance, procurement and facilities each held their own version of the truth, reconciled by people rather than by systems.
What made it harder than it looked
Two things. There was no incumbent system, which sounds like an advantage and is not: without one, the organisation had never been forced to agree on definitions. And the operation could not pause. A development with residents does not have a quiet period in which to change how it runs.
So the problem arrived as a procurement question and was not one. Nobody disagreed that the organisation needed a system. What nobody had established was what the system would be asked to record: what an approval meant, when a job was complete, which entity carried which cost.
The decision
The decision I pushed for was to spend an uncomfortable amount of time before selection establishing what an approval meant, when a job was considered complete, and which entity carried which cost.
Several people reasonably felt this was delay. It was the opposite: those definitions were going to be set either by us deliberately or by the configuration accidentally, and the second is far more expensive to unwind.
What we deliberately did not do
We did not begin with software selection. The shortlist came after the definitions rather than before them, which reversed the order everyone expected and was the part that had to be defended.
We did not leave the open questions to the configuration. Every definition still unsettled at selection would have been answered anyway — by whoever configured that screen, under schedule pressure, with nobody in the room aware that a decision was being made.
We did not plan around a quiet period. There was not going to be one. A plan that assumed a pause would have failed on contact with the first occupied building.
The system that followed
Enterprise resource planning across finance, procurement and facility management — the first enterprise-wide system in the organisation’s history — together with a smart data centre, a redesigned network topology, an enterprise security estate including CCTV and visitor and tenant management, and compound-wide wireless coverage. A mobile facility-management platform put the maintenance workflow in the hands of the people actually doing it.
Built on
Microsoft Dynamics BC and Dynamics 365 CRM, IoT-integrated facility management, enterprise networking and security infrastructure.
What changed
The organisation moved from reconciliation to record: one place where a purchase order, a maintenance job and a cost centre meant the same thing to everyone. The reactive-maintenance load fell as the mobile platform closed the gap between a fault occurring and someone being told about it.
Specific performance figures are held internally and are not published here, as they have not been independently evidenced.
What this taught me
The absence of a legacy system is not a clean slate. It usually means the organisation has never had to make its definitions explicit, and the implementation becomes the first time anyone has asked. Budget for that argument, because it is the actual project.
The advisory work this belongs to is set out under advisory.