From Data Sprawl to a Single Source of Truth

02Know who decides what

Listen|Step 2

Know Who Decides What

Everyone works from the same truth, at the depth their decisions need.

Design for decisions

Every operational dashboard eventually disappoints someone. Executives find it too detailed. Engineers find it too shallow. That’s what happens when you design for “users” instead of for decisions.

One shared experience doesn’t mean everyone sees the same screen. It means everyone works from the same truth, at the depth their decisions need. An executive needs to know in seconds whether operations are healthy and where they aren’t. A reliability engineer needs to follow that signal down to a single asset.

How to apply it

01 / 04
  1. Roles

    Find the roles that keep coming up in your interviews.

    Most organizations have three to six that matter.

  2. Decisions

    Define each role by its decisions, not its job title.

    Write down the decisions it makes most often, the information each decision needs, and how fresh and how detailed that information has to be.

  3. Signal

    Separate signal from noise.

    Ask what each role looks at first and what it never looks at. Both answers shape the design.

  4. Habits

    Respect existing habits.

    Note the tools each role already relies on. Whatever you build has to be better than the spreadsheet it replaces, or people will go back to the spreadsheet.

In the field

At the oil and gas company, five roles kept appearing, each with a clear need and its own kind of friction.

Roles
05
KPIs in use
1000

[W]e have 1000 KPIs and they’re all over the place. That’s not realistic to do a quick eye shot and say, all right, are we doing good?

From the interviews

Five roles, one truth

Each role’s primary need, its typical friction, and how deep into the same truth it needs to see.

Most of the time people tend to gravitate back towards existing tools, legacy systems spreadsheets that they’re already using.

From the interviews
RoleNeedsSlowed bySees
Operations executiveNeedsA concise, trustworthy view of performance and exceptions.Slowed byToo many measures with too little context.SeesHealth at a glanceAre operations healthy, and where aren’t they? Exceptions first, in seconds, with the detail one step away.
Project managerNeedsVisibility into progress, risks and dependencies.Slowed byScattered project information.SeesProgress across every workstreamMilestones, risks and the dependencies between them, all in one place rather than spread across trackers.
Maintenance leadNeedsFast access to operational signals and the detail behind them.Slowed byDelayed reporting and disconnected tools.SeesThe signal, then the detailAn alert first, then the readings and work orders behind it, fresh enough to plan the next shift.
Reliability specialistNeedsOne reliable view of equipment health.Slowed byConflicting information and duplicate data entry.SeesAsset by assetThe condition of each piece of equipment from one trusted record, entered once and used everywhere.
Engineering managerNeedsTimely information for proactive decisions.Slowed byLimited access to detail they could depend on.SeesA single asset’s historyReadings, changes and failures over time, detailed and dependable enough to act before something breaks.