Build, buy or adapt: choosing a learning platform

We build and operate our own learning platform, which makes us an unusual source of advice on this: we have an obvious interest in you commissioning one, and we still talk most enquiries out of it. Here is the reasoning we use.

Buy when your process is ordinary

If you deliver courses, enrol learners, run assessments and issue certificates in roughly the way everyone else does, an off-the-shelf platform will do it today for a subscription. Building the same thing is months of work to arrive at parity with something you could have had on Monday.

Adapt when one part is genuinely yours

The common case is an organisation whose delivery is standard but whose enrolment, fee collection or reporting is not — because of how their institution, regulator or funding works. Here the answer is usually a platform plus integration: keep the commodity parts, build the part that is actually specific to you, and connect them.

Build when the platform is the product

—You are selling the learning experience itself, and the interface is your differentiator rather than a cost centre.

—Your access rules, cohorts or accreditation logic cannot be expressed in an off-the-shelf permission model without fighting it every term.

—You need to own the data outright — for procurement, residency or safeguarding reasons that will not bend.

The costs people forget

A custom platform is not finished when it launches. It needs hosting, upgrades, support during enrolment week, and someone to answer at 8am when a term starts and nobody can log in. When we quote a build we quote this too, because a platform nobody maintains becomes a liability within about eighteen months.

Things to check before you decide

—What happens at term end — reconciliation and reporting are where most platforms disappoint.

—Whether your learners are on shared devices and metered data, which constrains page weight far more than feature count.

—Whether the academic calendar you need, including resits, can actually be modelled in the tool you are considering.

If two of those come back badly, buying will not save you. If none of them do, building probably will not pay.