The first morning of a big-room planning event runs on a schedule you could set a watch by. Two hundred people in a hotel ballroom, sticky notes bought by the case, string stretched across one wall to show what depends on what. An executive opens the day with fifteen minutes on why this quarter matters, poses for a photograph with the room behind her, and leaves for another meeting. I have run enough of these to know that the door closing behind her is where the event actually begins. What follows is a group of capable people reconstructing, from a slide deck and their own judgment, what the business wants built. They usually do a decent job of it. They are also guessing.

Digital.ai's 18th State of Agile Report puts a number on the guessing. Fifteen percent of respondents said business and executive leaders actively shape their organization's Agile practices.1 Seventy-six percent reported increased scrutiny on the business impact and return of that same work.1 Two things are worth saying plainly before going further. Digital.ai sells enterprise planning software, so it has a commercial interest in the finding that companies cannot connect delivery to value. And the survey ran to 349 practitioners, which is a modest sample drawn from inside the practice.1 I would not build a business case on the decimal places. The direction of it matches what I have watched happen inside delivery organizations for a long time.

The value question arrives at the wrong door

Take those two figures as one sentence and you have the arrangement most delivery organizations live inside. The people who decide what gets built are absent when it gets planned, and the people who are present are being asked to prove the worth of decisions they had no part in making.

Whether a piece of work was worth doing gets settled at the moment somebody puts it at the top of a list. Everything downstream of that is craft. A team can build the wrong thing beautifully, ship it on schedule, pass every quality gate, and hold a retrospective that genuinely improves how they work together. When the return question comes around, it comes to them, because they are the ones holding a burndown chart and a demo environment. Whoever chose the item eighteen months earlier has a different title by now, or a different employer.

An audit of delivery cannot recover value that was never selected. What you are grading is the second half of a decision whose first half nobody in the room witnessed.

Which is why the report's own recommendation reads a little strangely for a vendor document. It advises building forums where business leaders regularly take part in prioritization and review outcomes.1 That is an admission, printed inside a marketing asset, that the leaders are not currently in those conversations.

What a team measures when nobody tells it otherwise

With no business voice at intake, teams do not stop measuring. They measure whatever sits in front of them. Cycle time. Throughput. Escaped defects. Percentage of committed work completed. Each of those is an honest report about the machine, and none of them speaks to whether the machine was pointed at anything worth building. Under half of respondents said their product managers can manage the full delivery pipeline and measure business or customer value.1 In most places the instrument for the harder question was never installed.

A few weeks ago I wrote about why a measure only changes behavior when the person being measured can watch it move. That constraint runs upward too. A number informs an executive decision only when the executive had some hand in deciding what the number was meant to stand for. Hand a delivery metric to somebody who was elsewhere when the work got chosen and you have given them a very precise answer to a question they were not asking.

More than half of respondents said they struggle to prioritize the right work. A similar share said they struggle to track business impact.1 Those are one problem written down twice. Nobody can trace the impact of work whose intended impact was never stated in a form anyone could check later.

Fifteen minutes is not participation

Only 13 percent of respondents described Agile as deeply embedded across their business and technology functions, while 22 percent said it is scaled across IT with limited business involvement.1 That second figure is the structural picture of everything above. A practice living entirely on the technology side of the house, doing work chosen on the business side of the house, and answering for results measured somewhere in between.

The remedy is unglamorous and it involves no new framework. Sit in the conversation where forty reasonable ideas get sorted into an order, and say out loud what each of the top few is supposed to buy. Then write that sentence down. It is the only thing an honest return review can be measured against later. It also decays, which is something I have described before as the half-life of strategic clarity, so it needs restating far more often than feels reasonable.

Fifteen minutes at the podium and a photograph is attendance. The room can tell the difference, and so can the quarter. If your delivery organization is under a return review right now, the question I would put to whoever commissioned it is a simple one. What did the last quarter of work buy, and who wrote that down before the work started?