• Written by Azhar Aliyar, Director

Executives are not short of data. They are short of a view. The information needed for most decisions exists inside the organisation, distributed across functions that each report it in their own format on their own cycle, and assembling it takes long enough that the question has usually moved on.

The consequence is a familiar pattern: a decision is taken on the basis of the last pack plus the judgement of whoever is in the room, and the supporting analysis arrives afterwards to justify it.

The cost of the pack cycle

A management pack is a snapshot commissioned in advance, which means the questions it answers were chosen weeks before anyone knew what would matter.

The good ones are excellent at what they do. What none of them can do is answer the question that arises in the meeting itself, which is where decisions are actually made. Somebody asks how the production shortfall in one region relates to the inventory position and the revised forecast, three functions look at their own slide, and the honest answer is that nobody has put those three things together.

That question then becomes an action, and the decision is either deferred or taken without the answer. Both are expensive, and the second is more common.

Cross functional questions are the hard ones

Within a function, most questions are already answerable. The production team knows its output, finance knows its ledger, sales knows its pipeline. Difficulty concentrates at the boundaries, and so does decision value.

Consider: is the margin pressure in this category coming from input costs, from discounting, or from mix. Answering it requires production cost data, sales transaction data and finance's category structure at once, and in most organisations those three do not share a definition of category.

This is the specific gap generative AI for executive leadership addresses. Not producing a better chart, but making it possible to ask a question that crosses three functions and receive an answer while the meeting is still running.

Reconciliation is a feature, not an obstacle

The first serious attempt at a consolidated view always uncovers that functions disagree. Sales counts a booking at signature, finance at delivery. Production reports a month on a different calendar. Two systems hold a customer hierarchy that diverged years ago.

The instinct is to treat this as a data problem delaying the project. It is better understood as the project. These divergences have been distorting every cross-functional conversation the organisation has had, silently, because a pack presents each function's numbers on separate pages where they never have to agree.

Resolving them is not glamorous. It is also the thing that makes every subsequent answer trustworthy, and it has to happen once.

The second and third question

The most underrated benefit is not the first answer. It is the follow-up.

A leader who receives a number in a pack can accept it or query it. A leader who can ask why, and then ask which segment, and then ask whether it happened last year, is running an investigation rather than receiving a briefing. The quality of the resulting decision is different, and it is different specifically because the third question got asked.

Most of the value of shortening the question cycle shows up here, and it is invisible in any measure of how often the system is used.

What to remain careful about

There is a real failure mode where the measurable crowds out the important. Strategic decisions turn on competitive dynamics, regulatory direction, talent and timing, and none of that is in the data warehouse. A system that answers operational questions quickly can, without anyone intending it, make the operational the whole agenda.

The guard is to keep the unanswerable questions on the table explicitly. What would have to be true for this to work. What are we assuming about the competition. What would change our minds. These deserve the time freed up by not assembling facts, and if they do not get it the exercise has made the organisation faster at the wrong things.

Where to begin

Start with the recurring cross-functional question your leadership team asks most often and answers worst. Every organisation has one, and it usually spans exactly the boundaries where definitions disagree.

Connect the systems it depends on, resolve the definitional conflicts it exposes, and answer it live in the next leadership meeting. If the answer holds up to challenge from the function that owns each number, you have both a working capability and, more importantly, agreement on what the numbers mean.

What changes after six months

The visible changes are modest and the structural ones are not.

Meetings tend to get shorter, because the portion spent establishing what happened contracts. Actions of the form "come back with the numbers" largely disappear, and are replaced by actions about decisions. Functions stop presenting different versions of the same figure, not because anyone mandated consistency but because the discrepancies became visible and were resolved.

The less comfortable change is that assumptions become explicit. When a leader can check a claim in the meeting, claims get checked, including ones that had been accepted for years. Some of them will not survive. That is the value being delivered, and it does not always feel like it at the time.

The failure mode to watch for

The way this goes wrong is rarely technical. It is that the capability becomes the property of one function, usually finance, and everyone else experiences it as being measured by somebody else's definitions.

Once that happens, the other functions revert to maintaining their own numbers in parallel, and the organisation is back to reconciling in meetings with the added overhead of a system nobody trusts. Preventing it is a governance question rather than a product one: the definitions have to be agreed jointly and owned jointly, and every function needs to be able to ask questions rather than only to answer them.

To discuss this against your own reporting, schedule a demo.