Nobody knew who to call

The margin problem is not a pricing problem

Facilities management outsourcing has grown for years. Industry EBITDA margin has sat around 7 per cent through that growth, which is the whole reason the question gets asked. Over the same period buyers have widened what they expect: compliance risk transferred to the provider, an active role in the customer strategy, more than hands on site.

That combination usually gets answered with a product. Build an energy analytics service, sell a monitoring subscription, attach software to the contract. The sprint was set up to decide which of those to build, and in what order.

The answer changed shape as soon as the current workflow was drawn end to end.

What happens today when a building misbehaves

High energy use in a building raises a notification on a third party analytics platform. That platform tells the building manager. The building manager tells the property manager, who is typically responsible for five to thirty buildings and who does not know the cause. Because the cause is unknown, the description passed on is partial. The provider then sends hardware specialists one after another until one of them identifies it.

The alarms do not reach the enterprise system that raises work orders, so none of that sequence exists as a job until a person types it in.

The reason it stays this way is structural rather than careless. Not every facility is connected to building automation at all. Where it is, the automation parameters were set at design and are not updated, and get touched only when something goes badly wrong. Manual inspection of the automation system happens about twice a month. And there is no standardised communication protocol, so a new solution cannot simply be plugged in.

The evidence for digital FM belongs to somebody else

Fourteen solutions were profiled end to end, and thirteen carried a written deployment with numbers attached. Reading them together made an uncomfortable pattern obvious. Every benefit figure in this market belongs to a software vendor describing its own customer, not to a facilities provider describing its own contract.

The reported effects are real and they are wide. Energy consumption reductions across the deployments reviewed ran from 5 per cent to 64 per cent depending on what was being measured. Energy cost savings were quoted as ranges of 20 to 35 per cent. One work order platform reported a 33 per cent reduction in the number of service providers a retailer used, and another reported a payback period of 18 months. Every one of those is a vendor reported result from another company deployment. None of it is a result the client had achieved.

Three capabilities that looked decisive on a slide did not survive checking. One vendor markets asset level tracking and benchmarking that is not yet in the market. Another depends on a gateway that is incompatible with older buildings because they do not run the standard protocol. A third offers real time monitoring the building management system is already doing, so it is redundant rather than additive.

The sequencing that came out of it

If the workflow cannot tell you what is wrong, no analytics product sitting on top of it will pay. So the roadmap was ordered by dependency rather than by attractiveness.

Horizon one is not a product. It is integration. Hardware joined to the analytics platform and the alarm system so a work order raises itself rather than arriving after three phone calls, and a new way of categorising a complaint into either energy or condition, because that is the only distinction a customer can make reliably. A cold office and a jump in electricity consumption are two different dispatches, and today they are the same phone call.

Horizon two is the user centric building, where zoning and occupant feedback drive condition monitoring, with the level of control bounded by what the contract actually covers. Horizon three is a single source of truth across the whole portfolio.

The occupancy prediction service, the one everybody was most excited about, was deliberately held back. Customers want it, there are few commercially viable solutions, and the sensor requirement puts significant capital cost on the building owner. It is a stated need without an established market on either side of it yet.

What the client was left holding

A profiled market of fourteen solutions with the capability gaps marked, including the three that do not work as advertised. A workflow map of the current dispatch path with the points where time is lost. And a three horizon sequence with the dependency stated, so the cheap integration work does not get skipped in favour of the interesting product work.

This was a workshop engagement and the roadmap is a proposal. Nothing in it had been built, funded or piloted at handover, and no adoption or benefit is claimed here.