'The mutual learning process is still in its early stages' — what the Court of Audit already saw in 2013

In 2013, the Dutch Court of Audit (Algemene Rekenkamer) wrote one sentence in its report Aanpak van ICT door het Rijk 2012 that has barely been picked up since: "the mutual learning process based on general lessons from completed reviews is still in its early stages." Thirteen years later, many an archive cabinet still holds an evaluation report that was never opened again after delivery — and that is exactly the pattern that one sentence points to.

What the Court of Audit found exactly

At the time, the Court of Audit examined how central government approached IT projects, with specific attention to the role of Gateway Reviews: the interim, independent review moments during a project's lifetime. The conclusion was twofold. Learning from those reviews, structurally using lessons from one review for the next project, was still in its infancy. And the Court of Audit found that the realisation of benefits from business cases was still insufficiently monitored, and advised using business cases both at the start and throughout a project's entire duration as a steering instrument.

It is important to be precise about what the Court of Audit did not say. The report does not state that no ex-post evaluation takes place at all, that claim does not hold up on closer inspection. What the report does say is sharper and more interesting: government relies on Gateway Reviews and the Handboek Portfoliomanagement, but the learning capacity from those reviews, turning individual findings into organisation-wide improvement, was demonstrably not yet mature in 2013.

Has that improved since?

An honest answer: there is no recent, comparable independent research available that proves this pattern has since been broken. What we do know is that the Elias committee, a year later (2014), exposed a largely different gap: ex-ante testing was insufficient. The response to it (the creation of AcICT) focused entirely on the front end of projects. No comparable, mandatory institution has emerged for the ex-post learning process that the Court of Audit already pointed to in 2013.

That, in itself, is telling. Two independent bodies, a year apart, each signalled a different part of the same problem: one the testing beforehand, the other the learning afterwards. Only the first received a structural, institutionalised solution.

Why "early stages" is a precise diagnosis, not a reproach

The Court of Audit's phrasing is functional, not moralising. A learning process that is still in its early stages does not mean nothing happens. It means there is no repeatable, embedded mechanism to let lessons from one review or evaluation systematically land in the approach to the next project. That exact distinction, between incidentally writing something down and structurally changing something, is what the rest of this blog series keeps returning to, including in the piece on Williams (2007) and the difference between documenting and learning.

Recognisable beyond central government

Ask around in any organisation, commercial, public, non-profit, how evaluations of completed projects are followed up, and you rarely get a convincing answer. There is usually a report, a presentation, or at least minutes from the last project team meeting. What is rarely there: a demonstrable mechanism through which the organisation as a whole, not just the one project team, makes use of those lessons. That is exactly the "early stage" the Court of Audit pointed to in 2013, and it is a pattern that is at least as recognisable in 2026 as it was then.

What you can do with this

If your evaluation report does name lessons, the follow-up question is: are those lessons concrete and traceable to an event, or generic ("communication could be better")? And is there an explicit link to a follow-up action and an owner, or does it stop at simply stating the lesson? That distinction, between documentation and an embedded follow-up, is one of the seven categories on which EvaluatieScore assesses an evaluation report.

EvaluatieScore does not claim that a good evaluation is proven to lead to a better next project. That proof simply does not exist, and we are transparent about that. What the instrument does do: test whether your evaluation is built according to the frameworks that authoritative bodies (OECD, the British government, PRINCE2, the CIPP model) set for a good evaluation report, including the lessons-learned component the Court of Audit already flagged back in 2013.

Put that last criterion next to your own most recent evaluation report: does the follow-up action have a name attached, or does it end at establishing the finding? Check your report for €15 at dutchmind.com/producten/evaluatiescore.

Sources

1. Algemene Rekenkamer (2013), Aanpak van ICT door het Rijk 2012 — rekenkamer.nl.
2. Tijdelijke commissie ICT-projecten (commissie-Elias), Eindrapport "Grip op ICT" + kabinetsreactie (2014/2015) — pianoo.nl.
3. Williams, T. (2007), Post-Project Reviews to Gain Effective Lessons Learned — eprints.soton.ac.uk.