Productise the platform. The value flow follows.

Why the ambition is the easy part

The slogan is short. The spec underneath it is not. Same-day delivery means that on the morning of day one, an engineer can describe a service they want. By the afternoon of the same day, that service runs in production. It observes itself. It costs nothing surprising. It passes controls the security team has already reviewed. It looks like a Nedap product rather than a one-off.

That ambition has a long list of preconditions. Every precondition is a thing the platform has to be, not a thing the team has to remember. If a precondition lives in a runbook the engineer has to open, the day is over before the deployment starts.

The honest version of the ambition is a paved road that picks up the idea where it lands and carries it through every checkpoint without negotiation. Building that paved road is a year of work. Productising it so other teams can use it without us is the harder year that follows.

Productise the platform, not just the building blocks

Most internal platforms we inherit are unproductised. They are a collection of building blocks. A Kubernetes cluster here, a Terraform module library there, a CI/CD pipeline pattern someone wrote in 2022 and three people still maintain. Each block is competent in isolation. None of them are a product.

A product is the same building block plus four things. An owner. A lifecycle. A contract. A roadmap. Without those, the block is a craft project. With them, it is something the rest of the organisation can use without asking permission and without inheriting your maintenance burden.

We started the work with Nedap by drawing the line between the two. Every existing block was classified as either productised or in transit to becoming productised. The unclassified ones were not allowed near the value creation flow until they had an owner, a stage, a contract, and a published consumption pattern.

Six principles that decide what gets in

The principles are not decorative. They are the entry test. A building block that fails one of them does not enter the paved road.

One. Lifecycle-based management. Every service has a declared stage (incubation, growth, mature, sunset, retired) and a roadmap that matches the stage.

Two. No pets, only cattle. Nothing is handcrafted. If a service cannot be redeployed from code at any hour, it is not a service yet.

Three. Compliant by default. Security, identity, observability and audit are baked into the template, not bolted on later. A team that adopts the golden path gets these for free.

Four. Batteries included. The thing arrives ready to use. Documentation, monitoring, alerting, ownership and a published consumption model come in the same shipment. No customisation tax.

Five. Fit for purpose, not a Swiss Army knife. Every service solves one well-defined problem cleanly. Generality is a maintenance bill nobody asked for.

Six. Standardisation is key. Shared patterns for architecture, CI/CD, observability and documentation. Exceptions are explicit and governed, not silently introduced.

The principles travel across engagements, but they took final form in the Nedap work because Foundry asked the harder question. Not what each block costs, but what the platform has to be.

The lifecycle is the contract

A productised service moves through five stages. Incubation is exploratory. Prove the value and the feasibility. Growth is the hardening stage. SLOs, baseline compliance, automated CI/CD, documentation, ownership. Mature is widely adopted and core to operations. Sunset is the explicit step toward retirement. Retired is gone.

Between the stages sit four gates. G0 opens the work. G1 declares the service ready for first users. G2 promotes it to a standard offering the rest of the organisation can consume self-service. G3 is the decision to optimise further or retire.

Each gate carries entry and exit criteria. Each gate produces deliverables visible in the catalogue. We do not have steering committees that argue about whether to promote a service. We have evidence the gate either holds or does not.

The reason this matters for same-day delivery is simple. The value creation flow can only pick up services from G2 or later. If a service has not earned its gate, it does not exist for the paved road. That sounds harsh until you remember the alternative. The paved road silently inherits unproductised debt and slows down.

"Same-day delivery is what the value creation flow looks like when it works. Building the flow is the slow year. Productising the platform underneath it is the slower one."

Sebastiaan van Parijs / Founder

OLAs make the contract operational

A lifecycle is a promise. An Operational Level Agreement is what makes the promise enforceable inside one organisation. Every service from G2 onward has an OLA that names the criticality tier, the on-call coverage, the response times, the SLO targets, the security commitments and the review cadence.

This is where most platforms quietly fail. They have SLAs facing external customers. They have nothing internal. The result is that when a payment service depends on a shared identity service depends on a shared logging service, there is no contract among the three. The team running the payment service inherits the failure modes of two others and discovers them at 3am.

The Nedap engagement made the OLA layer real. Tier 1 services run 24/7 with a 15-minute response and a 2-hour resolution target on P1. Tier 2 services run business-hours. Tier 3 services run best-effort. Every team consuming a productised building block reads the tier before they integrate. Surprises drop. The on-call burden shifts from heroics to coverage.

The value flow is the byproduct

With the platform productised and the OLAs in place, the value creation flow becomes possible without anyone redesigning it. The flow is the path a piece of value walks from Product Manager (strategic intent) to Product Owner (tactical refinement) to engineering team (continuous delivery).

The PM sets the epic, the priority, the success metric. The PO refines into stories with acceptance criteria. The engineering team builds against the paved road that the platform already provides. Each step takes its inputs from the previous one and produces outputs the next one can use without reformatting.

Foundry's line is what this looks like when it works. The idea enters at the PM stage. By the afternoon it has been refined, approved against gates visible in the catalogue, deployed through a pipeline that already exists, observed by a dashboard that already exists, and the team owning it has already signed the OLA.

We measure the flow through a single-pane insights and control layer that sits on top of the productised services. Lead time, deployment frequency, SLO attainment, cost per workload, security findings. Five numbers, one screen, every team can see them. If they cannot, the platform is not actually productised yet.

What this is not

This is not a slogan refit. It is not a workshop. We are not promising that the ambition is achieved by buying a tool. The productisation framework is the slow work. Twelve months to get the first paved road to G2. A year more to graduate the next three. The dividend is that after the second year the value creation flow stops requiring meetings to function. It just runs.

This is not unique to Nedap either. The same productisation work, the same lifecycle, the same six principles, the same OLA discipline applies to any engineering organisation that wants its platform to be a multiplier rather than a tax.

What is unique is the ambition Foundry set. Most leadership teams ask for faster delivery. Nedap asked for end of day. End of day is a much better spec, because it forces the productisation conversation early. We took the spec seriously. The value creation flow followed.