A disappointing quarter is usually the first moment someone on the board asks the question that has been lingering for a while: does our assumption about AI still hold, or have we been calculating for four quarters with something that now works differently. The figure itself explains nothing. It only forces a look.
As long as results are running according to plan, there is no reason to bring back an assumption and test it. A disappointing quarter does not break that calm because it proves an assumption was wrong, but because it is the one occasion on which a board collectively looks at it. It is the moment when the budget, the forecast, and the underlying assumptions about who does the work — human, AI, or a combination with human oversight — are back on the table.
The temptation is to explain the figure right away: the market, the price, the competitor. All of that may well be true. But if one of the assumptions underlying the plan was a claim about what AI would take over — the speed of a process, the capacity of a team, the turnaround time of a service — then the disappointing quarter is also the moment to ask whether that assumption still describes today's factual level, or the factual level of the quarter in which it was written down.
AI takes over work in parts, not all at once. Some tasks can by now be done entirely by AI. Others run partly through AI, with a human approving or rejecting with reason. Others still remain human work, whatever a vendor claims about it. This division changes continuously, and not at the same pace everywhere in the company: one team already works with AI that handles a task entirely, while another team still uses the same technology only as an aid alongside its own work.
The difference rarely lies in the technology. It lies in who within the company first let go of the assumption. A team lead who noticed three months ago that a task went faster adjusts their planning without reporting it. The assumption in the multi-year plan then stays as it was written down, while practice has already moved on. A disappointing quarter sometimes exposes that gap without naming it: the result deviates, but no one can immediately say whether that is due to the market or to an assumption that has no longer held within the company for months.
At this point it can be established which assumptions in the plan say something about AI and work — explicitly, or hidden in a productivity figure. And it can be established when that assumption was last confirmed: by whom, on the basis of what information, and whether that was still this year. That is a factual exercise. It does not tell you whether the quarter fell short because of it, but it does tell you whether there is an assumption in the plan that no one has checked in the past six months.
It can also be established how the board's self-image relates to the outside view: what the team itself believes AI is already doing, versus what sector data, regulation, and competitive signals show about the pace at which that is happening elsewhere. Differences here are not a problem in themselves — but they are a signal about which assumption deserves attention first.
A disappointing quarter is not proof that an AI assumption was the cause, and it is also not a moment to draw conclusions from it about personnel. If the outcome of this review touches on decisions about employees, separate legal requirements apply to those — they do not follow from a quarterly figure or from this review.
Nor can it be established at this point whether a task that is still human work now will still be so in two quarters' time. That division shifts, and the pace at which it does so differs by task and by sector. Anyone who makes a claim about that now without having looked at the underlying task is, in effect, repeating a new assumption in place of the old one.
A disappointing quarter is not the only moment at which a board asks this question, but it is the most direct one: the result forces a look, whereas an investment decision or an acquisition already carries the occasion within it beforehand. Midway through a multi-year plan is also a planned occasion; here it is unplanned and therefore sharper. Anyone wondering whether this is an incident or a pattern can broaden the question to what to do if the market changes faster than your own plan.
The underlying question — which work in this company can genuinely be taken over by AI right now, which part with human oversight, and which part remains human work — is answered per task with the work scan from FTE TO AI. That is not an answer to the question of why the quarter fell short, but it is an answer to the question of whether the plan is still calculating with something that still holds true.
Anyone who wants to start without setting up the full stress test right away can do the free assumption check: a short round in which you name your key assumptions and see, per assumption, when it was last confirmed. The full stress test, with the board's self-assessment set alongside the outside view, is under construction. Anyone who would rather first go through their own strategy as a whole again will find a starting point at the question of whether the strategy still holds.
Stel uw vraag. Vaak zit de echte vraag een laag dieper — daar mag ik naar vragen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.