Your Executive Education Programme Is Being Judged on a Happy Sheet. Its Buyers Are Judging Something Else.

Bottom line up front: Executive education and continuing professional development (CPD) are evaluated with the wrong instrument. The end-of-course satisfaction survey — the "happy sheet" — measures whether participants enjoyed the room. But the people who pay for executive education, usually a sponsoring employer, are asking a different question: did behaviour change back at work, and did it move a business outcome? That is Kirkpatrick Level 3 and Level 4, and almost nobody measures it. In one widely cited breakdown of training-evaluation practice, 94% of evaluation happens at Level 1 (reaction), 34% at Level 2 (learning), just 13% at Level 3 (behaviour) and only 3% at Level 4 (results) (Devlin Peck, summarising the Kirkpatrick evaluation literature). For a product whose entire value proposition is impact, that is a measurement failure hiding in plain sight.

The buyer and the beneficiary are not the same person

In a taught degree, the student is broadly both the customer and the beneficiary, and end-of-module evaluation — for all its well-documented flaws — at least asks the right person. Executive education breaks that symmetry. On a custom programme, an employer commissions the course, defines the outcomes, and writes the cheque; the participant is a delegate, not a purchaser. On an open-enrolment programme, the participant may self-fund, but they are typically buying a change in their own capability and career, not a pleasant week on a campus.

So when you hand a cohort of senior managers a five-point scale asking whether the facilitator was clear and the catering was good, you are measuring the reaction of the beneficiary while ignoring the question of the buyer. The sponsoring organisation did not commission the programme so that its people would rate it 4.6 out of 5. It commissioned it because it believed the training would change how those people lead, sell, decide, or manage risk — and that this change would show up in performance. None of that is on the happy sheet.

Even the prestige rankings run on satisfaction

You might expect the top of the market to have solved this. It has not. The Financial Times executive-education rankings — the sector's most-watched quality signal, drawn from a survey base spanning schools in the UNICON consortium of leading business schools and beyond — are built substantially on participant and client survey responses. In the 2026 methodology the most heavily weighted measures include faculty quality, new skills and learning, course design, and "aims achieved," each around 10% of the score (Poets&Quants, on the FT 2026 methodology). "Aims achieved" and "new skills and learning" are the closest the ranking gets to impact — and they are still self-reported perceptions gathered at or shortly after the programme, closer to Kirkpatrick Level 2 than to demonstrated behaviour change at Level 3.

This is not a criticism of the schools, who are doing what is measurable at scale. It is an observation that the whole sector's quality vocabulary is anchored in reaction and perceived learning, precisely the levels the evidence says are easiest and least predictive. If the market leaders are ranked on sophisticated satisfaction data, it is unsurprising that a mid-sized CPD provider defaults to a one-page feedback form.

What executive-education evaluation should actually capture

The Kirkpatrick model is old, and its critics are right that Levels 3 and 4 are hard, expensive, and confounded by everything else happening in a business. But "hard to measure" is not the same as "not worth measuring," and the alternative — measuring only what is easy — is how you end up optimising for enjoyment. A serious executive-education evaluation strategy looks different from a module survey in four ways:

• It defines success with the sponsor, before the programme. The CIPP evaluation model — context, input, process, product — starts by asking what the commissioning organisation is actually trying to change. That conversation, held up front, is the single biggest determinant of whether evaluation will be meaningful.
• It measures behaviour, not just reaction. A short structured follow-up at 60–90 days — with the participant and, ideally, their manager — asking what they have actually done differently is worth more than any number of exit-survey stars. This is the same logic that makes work-integrated learning resist the end-of-module survey: the outcome happens after the room empties.
• It treats short and modular formats as their own problem. Executive education is increasingly delivered as stackable short courses and micro-credentials, which break the end-of-semester survey entirely. You cannot run a semester-end instrument on a three-day programme.
• It measures capability, not happiness. Beyond satisfaction, the question is whether specific competences developed — and for leadership and enterprise programmes, frameworks like the EU EntreComp competence model give you a defensible vocabulary for what "new skills" concretely means.

Some providers push all the way to Level 5 — Jack Phillips' return-on-investment extension of the Kirkpatrick model, which puts a monetary value on business results. That is appropriate for a large custom contract with a defined performance target, and overkill for a professional-update seminar. The point is not that every programme needs an ROI study; it is that the level of evaluation should match the buyer's actual definition of value, not the provider's convenience.

"But behaviour change is confounded — you can never attribute it to the course"

This is the strongest objection, and it is partly correct. If a manager becomes more effective six months after a leadership programme, was it the programme, a new boss, a reorganisation, or simple maturation? You cannot cleanly attribute a single business outcome to a single course, and any provider claiming to "prove ROI" with a satisfaction survey and a spreadsheet is selling a story.

But the honest response is not to retreat to Level 1 because Level 3 is confounded. It is to gather better evidence while being candid about proof. A structured behavioural follow-up that asks specific, concrete questions — "describe a decision you made differently," "what did you stop doing" — produces qualitative evidence a manager and sponsor can weigh, even without a control group. Triangulating participant self-report with manager observation reduces the single-source bias that makes happy sheets so weak. The goal is defensible evidence of change, framed as such — not a false claim of causal proof. Providers who are transparent about that distinction earn more trust than those who inflate a reaction score into an impact claim.

A second objection: senior executives are busy and will not complete a follow-up survey. Often true — which is exactly why the format matters. A conversational follow-up that takes the shape of a short interview, not a 40-item questionnaire, is far likelier to get a considered answer from someone who would never open a grid of Likert items.

Where Koji fits

This is the gap Koji for Education is built to close. Instead of a static end-of-course form, Koji runs AI-moderated conversational interviews that can probe what changed — following up on a vague answer, asking for a concrete example, surfacing the behaviour behind the rating. Because the moderation is standardised, a 60-day follow-up with 200 alumni is as consistent as one with 20, without the human-facilitator variability that makes qualitative follow-up expensive to do at scale. Automatic thematic analysis turns hundreds of open-text and interview responses into the themes a programme director and a corporate sponsor can act on, and programme-level reporting rolls it up for the commissioning organisation.

Koji does not claim to prove ROI — no honest instrument can from feedback alone. What it does is move evaluation off the reaction level and toward the behavioural evidence your buyers actually want, in a format busy professionals will engage with. The same conversational interview engine powers koji.so for teams running general customer and market research, so the method is battle-tested well beyond the classroom.

If your executive-education or CPD portfolio is still being judged on a happy sheet while its buyers judge it on impact, the instrument is the problem. See how conversational evaluation works for executive education.

Frequently asked questions

Why is a satisfaction survey the wrong tool for executive education?

Because the buyer and the beneficiary differ. A sponsoring employer commissions executive education for behaviour change and business impact (Kirkpatrick Levels 3-4), but a happy sheet measures only the participant's reaction (Level 1). The instrument answers the beneficiary's comfort, not the buyer's question.

What do the Financial Times executive-education rankings actually measure?

They are built substantially on participant and client survey responses. In the 2026 methodology the most heavily weighted measures include faculty quality, new skills and learning, course design and "aims achieved" — each around 10%. These are self-reported perceptions closer to Kirkpatrick Level 2 than to demonstrated behaviour change at Level 3.

How can you evaluate behaviour change if you cannot prove causation?

You gather defensible evidence without claiming causal proof. A structured behavioural follow-up at 60-90 days asking for concrete examples of what participants did differently, triangulated with manager observation, produces evidence a sponsor can weigh — even without a control group. The key is being transparent that this is evidence of change, not proof of ROI.

What is Kirkpatrick Level 5?

Level 5 is Jack Phillips' return-on-investment extension of the four-level model, placing a monetary value on business results. It suits large custom contracts with defined performance targets and is overkill for a professional-update seminar. The evaluation level should match the buyer's definition of value.

Why do short and modular formats need different evaluation?

Executive education is increasingly delivered as stackable short courses and micro-credentials. A semester-end instrument does not fit a three-day programme, and the outcome — applied behaviour change — happens after the course ends, so evaluation has to reach beyond the room and the closing survey.