The most sophisticated CPD tracking tools available in 2026 are genuinely impressive pieces of software. AI-assisted evidence import pulls certificates and attendance records automatically. Zero-spreadsheet interfaces replace the annual scramble to reconstruct a year of CPD activity from email confirmations. Real-time compliance dashboards show, at a glance, exactly how many hours a professional has logged against their annual requirement, broken down by category, with gaps flagged before a renewal deadline arrives.
If CPD tracking has become this easy, here is the question worth sitting with: why does the evidence suggest so little CPD actually changes professional practice?
The answer is not that these tools are badly built. It is that they are solving the wrong problem with great precision.
The leverage-point problem
Donella Meadows, in her work on systems thinking, identified a hierarchy of places to intervene in a system, ranked by how much real change each produces. Near the bottom of that hierarchy sit parameters — the speed, intensity, or ease of flows within a system that is otherwise left untouched. Higher up sit the goals the system is organised around. Higher still sits the paradigm: the underlying, often unexamined belief about what the system is even for.
Frictionless CPD tracking operates at the parameter level. It changes how quickly and easily credit-hours accumulate. It does not touch the goal of the system, which remains exactly what it was before the tool existed: accumulate sufficient credit-hours to satisfy a professional body’s annual requirement. And it does not come anywhere near the paradigm — the largely unspoken assumption that credit-hours are a meaningful unit of competence development in the first place.
This is why a faster, more elegant version of CPD logging cannot, by itself, close the gap between attendance and practice change. The gap does not exist at the level the tool operates on. No amount of friction removed from the act of recording an hour changes what that hour is evidence of. A frictionless logging tool and a clunky spreadsheet are, in this specific sense, doing the same job — just at very different speeds.
CPD Impact and a CPD tracking tool are therefore not really in competition with each other, despite the surface similarity of “things that record CPD activity.” They are interventions at different levels of the same system, and they produce categorically different outputs as a result. One makes the existing measurement faster. The other asks whether the existing measurement was ever the right thing to be measuring.
The surrogate endpoint error
There is a precise term in statistical methodology for what credit-hours are functioning as here: a surrogate endpoint. A surrogate endpoint is a measurable, observable signal substituted for a true outcome that is harder, slower, or more expensive to observe directly. The concept comes from clinical medicine — Prentice (1989) established the formal criterion for when such a substitution is statistically valid: the true outcome must be fully explained by the surrogate, with nothing left over.
Medicine has a sobering case history of what happens when that criterion is not actually met. Rosiglitazone, an anti-diabetic drug, was approved by regulators in 1999–2000 on the strength of its measured effect on blood glucose and HbA1c — both well-established, validated surrogate markers for diabetes management. The surrogate moved in exactly the direction it was supposed to. A decade later, post-market data showed the drug increased the risk of heart failure and myocardial infarction. The surrogate had improved. The true outcome — patient health — had moved in the opposite direction.
Credit-hours are CPD’s surrogate endpoint, and the evidence that they fail Prentice’s criterion is not new or speculative. Baldwin and Ford’s (1988) foundational review of training transfer research identified the gap between training delivery and applied behaviour change as a structural feature of professional development, not an occasional failure. Beer, Finnström and Schrader’s widely cited 2016 critique — sometimes called “the great training robbery” — argued, from decades of organisational case research, that standalone training interventions reliably fail to produce lasting behaviour change regardless of how well-designed the session or how genuine the participant’s intent at the time. Their estimate at the time: global corporate training spend exceeded $350 billion annually, with most of it failing to produce durable change.
Making credit-hours faster and easier to log does not address any of this. It makes the surrogate more efficient to record. It does not make the surrogate a better predictor of the thing it was always meant to stand in for.
What the right lever looks like
This is not a case for abandoning measurement, or for treating administrative tracking as worthless. Removing friction from compliance logging solves a real, legitimate problem — nobody benefits from a profession spending hours reconstructing CPD records from old email confirmations.
The distinction is about what gets measured, not whether measurement happens at all. An intervention at the paradigm level does not ask “how much did this person attend?” It asks what the attendee intended to do differently as a result, and — critically — whether that intention was something they actually carried out. That is a different category of data entirely from a timestamped record of presence. Logging captures exposure. Capturing intent, and following up on it, is the only mechanism that has any chance of correlating with the outcome credit-hours have always been a proxy for, and have never reliably predicted.
The next time a CPD tracking tool tells you it makes compliance easier, ask what it measures after the session ends. If the answer is nothing, the tool is optimising the wrong thing — efficiently.
FAQs
CPD tracking records what a professional attended and when. CPD measurement asks whether attending changed how they practice. Tracking is administratively useful; measurement is professionally meaningful. Most tools available in 2026 do the first. CPD Impact does the second.
It removes administrative friction, which is genuinely useful. But it also entrenches the assumption that logging attendance is the right unit of CPD quality — an assumption the training transfer research literature has questioned for four decades. Easier logging makes the existing measurement system more efficient; it does not fix the measurement system.
In principle, yes. In practice, no CPD tracking tool has done this at scale in AEC, because the data required — verified practice-change evidence from the professional, not from the provider — does not exist in a tracking system. You cannot retrospectively add a check-in for a session whose window for honest behavioural reflection has already passed.
