Accelerating Discovery

Most large organisations know this cycle. A new executive arrives, or a new programme is funded, and the first task is a diagnostic. Workshops are booked. Process owners are interviewed. Spreadsheets are requested from finance. Twelve weeks later a slide deck describes the current state.

Eighteen months on, a different sponsor commissions a different team to do it again. The new deck rarely agrees with the old one, and nobody can say with confidence which is right.

This is not diligence. It is a structural failure to keep what the organisation has already learned.

The time cost is already large

Decision making consumes a striking share of management time. In a 2018 McKinsey Global Survey of 1,259 executives, respondents reported spending on average 37 per cent of their working time making decisions. More than half of that time was judged to be used ineffectively [1]. McKinsey estimated that, for managers at an average Fortune 500 company, this could mean more than 530,000 days of lost working time and roughly $250 million of wasted labour cost a year [1].

The same survey found that only 48 per cent of respondents agreed their organisations make decisions quickly, and just 37 per cent said decisions were both high in quality and fast [1]. Speed and quality were not in tension: organisations that decided quickly were twice as likely to decide well [1].

This matters commercially. Bain research led by Marcia Blenko, Michael Mankins and Paul Rogers found that decision effectiveness and financial results correlated at a 95 per cent confidence level or higher across every country, industry and company size in their sample [2].

Every repeated diagnostic adds weeks to the front of a decision that was already slow.

Why diagnostics repeat

Three conditions make repetition almost inevitable.

Knowledge leaves with people. Organisational memory, as James Walsh and Gerardo Ungson described it, is held in individuals, culture, routines, structures and records [3]. When the record is thin, memory depends on who is still in the room. The UK civil service shows the effect clearly. The Institute for Government found that managers across Whitehall stay in post for less than two years, and estimated the direct cost of excessive turnover at up to £74 million a year [4]. A separate Institute for Government study found that policies and organisations were recreated with alarming regularity, citing poor institutional memory as a persistent cause [5].

Outputs are documents, not models. A diagnostic usually ends as a presentation. A presentation cannot be updated when a process changes, rerun with a new assumption, or queried by the next team. It begins to age the day it is presented.

Evidence comes from opinion. Workshops and interviews capture how people believe work happens. Process-mining research has long shown that event data recorded by information systems can reveal how processes actually run, and can be checked against the documented model to expose deviations [6]. Where that evidence is not used, each diagnostic starts again from recollection.

Repetition dilutes certainty

Repeating a diagnostic does not simply waste time. It can leave leaders less sure than before.

Daniel Kahneman, Olivier Sibony and Cass Sunstein call unwanted variability in professional judgement “noise” [7]. In one study at a large insurer, 48 underwriters priced the same realistic cases. Executives expected the typical difference between two underwriters to be around 10 per cent. The median difference was 55 per cent of the average premium [7][8].

Organisational diagnostics are exposed to the same effect. Different teams, using different frameworks and interviewing different people, will produce different baselines of the same organisation. When two credible reports disagree, the rational response is to commission a third, or to wait. Either way, the decision slows and confidence falls.

What accelerating discovery looks like

The aim is not to stop diagnosing. It is to stop starting from zero.

1. Build the baseline once, as a model. Capture capabilities and the processes that deliver them in a structure that can be edited, not a deck that must be redrawn.
2. Make it the shared reference. The next programme should begin by opening the existing model and correcting it, not by booking new workshops.
3. Anchor it in evidence. Durations, volumes and costs should come from recorded data where it exists, with workshop estimates labelled as estimates.
4. Keep it queryable. A baseline earns its keep when someone can ask a new question of it without commissioning new work.
5. Let more people read it than change it. Wide read access builds shared understanding. Controlled write access protects the reference.

When discovery compounds in this way, the second diagnostic takes days rather than months, and leaders spend their time on the decision rather than on reconciling competing versions of the facts.

Where MIRAI fits

MIRAI holds the current state as a model rather than a document. Capabilities are drawn as a map. The processes beneath them are drawn as swimlanes with decisions, sub-processes and variations, and they can be saved, reloaded and revised. Simulation runs on the same model, so a revised process produces revised time and cost figures without a separate analysis.

Builder licences create and change the model. Viewer licences give executives read-only access to maps, results and the portfolio, so the reference can be shared widely without being altered. The simulation features have been built and tested on sample processes.

Questions to ask your own organisation

• How many current-state diagnostics have we paid for in the last three years, and how many are still in use?
• If two of them disagree, do we know which is right?
• Where is our process knowledge held: in a model, in documents, or in people who might leave?
• How long would it take a new programme lead to understand how a capability works today?