Operating Model Design: The Six-Tile Blueprint

If it does not fit on one page, no one carries it

An operating model that does not fit on one page does not live in any executive's head, and a model no one carries is a document rather than an operating model. The discipline of compressing the model to six tiles is the discipline of deciding what actually matters; everything else is supporting detail that belongs beneath the tiles, not on them. The six tiles are not a creative interpretation of the firm. They are the durable categories that have organized operational decisions across engagements for two decades, refined into the form that survives the pressure of execution.

The six tiles are strategy and choices, customer and offer, processes and flows, data and systems, people and operating rhythm, and governance and decision rights. Each tile is one paragraph at most, three bullets at most, one diagram at most. The constraint is the point. When a tile will not compress to that level, the firm has usually not decided the underlying question, and the overflow is evidence of an unmade decision rather than a case for a bigger page. A model held to one page can be reviewed in a single sitting, argued about honestly, and updated without a project, which is exactly what a long-form model cannot.

The six tiles in sequence

Each tile answers one question and hands the next tile its premise. Read in order, the tiles form the argument for how the operation runs.

Strategy and choices states what the firm is doing and, harder, what it has decided not to do, because a strategy that refuses nothing resources nothing. Customer and offer fixes the contract with the customer that everything below must serve. Processes and flows are drawn at the resolution of decisions, not tasks. Data and systems name authoritative sources rather than application names, because applications change and sources persist. People and operating rhythm name the cadence that keeps the model alive rather than an artifact, and governance and decision rights name who decides what, on what evidence, by when.

Consider a worked case. A specialty lender compresses its model. Strategy and choices names that it underwrites two segments and has decided not to enter consumer lending, freeing the resource it kept spreading thin. Customer and offer fixes the offer as approval inside forty-eight hours. Processes and flows diagrams the underwriting decision, not the twenty tasks around it. Data and systems names the authoritative credit source rather than the four applications that touch it. People and operating rhythm names the weekly credit committee. Governance and decision rights names the credit officer as the approver above a threshold. Six tiles, one page, and the contradiction between forty-eight-hour approval and a weekly committee surfaces immediately, which is the real operating decision.

Draft, reconcile, and keep it alive

The value is in the compression and the reconciliation, not the drafting. Write the tiles fast, then spend the time surfacing where they contradict each other.

The reconciliation session is where the model is actually decided. Drafted tile by tile with different owners, the six tiles will disagree, and those disagreements are not editing errors to smooth over. They are the real operating choices the firm has been avoiding, surfaced at last on a single page where they cannot hide. Resolve them in the room, record the resolution, and the operating model is made rather than merely written.

• Draft each tile in one sitting with the executive who owns it, holding hard to the one-paragraph, three-bullet, one-diagram limit.
• Push any content that will not compress into a named sub-document rather than expanding the tile, and note the sub-document reference on the page.
• Reconcile the tiles in a single half-day session whose only job is to surface the contradictions between them.
• Resolve the contradictions in the room, because those contradictions are the actual operating model decisions, and record the resolution.
• Commit the six tiles to the executive team's standing operating cadence and review them every quarter against operating evidence, not against last quarter's tiles.

Where the blueprint drifts back into a deck

• Letting a tile overflow the limit. The overflow usually hides an unmade decision. The fix: treat the overflow as a signal, decide the underlying question, and push the rest into a sub-document.
• Mapping processes at the resolution of tasks. Task lists belong in process documentation and drown the operating decisions. The fix: diagram decisions only and send tasks to the SOP.
• Naming applications in the data tile instead of authoritative sources. Applications change and the map goes stale. The fix: name the source of record and let the tooling beneath it vary.
• Omitting the operating rhythm. Without a named cadence the model is a static artifact no one runs. The fix: name the specific meetings and their frequency in the people tile.
• Skipping the governance tile, leaving an org chart in its place. An org chart shows reporting, not decision rights. The fix: name who decides what, on what evidence, by when. Reporting lines answer to whom; decision rights answer who decides, and only the second keeps the operation moving.

Before the model is committed

• All six tiles fit on one page, each within the one-paragraph, three-bullet, one-diagram limit.
• The strategy tile names at least one thing the firm has decided not to do.
• The process tile is drawn at the resolution of decisions and the data tile names authoritative sources.
• The people tile names a specific operating cadence and the governance tile names decision rights.
• The tiles were reconciled in a session that surfaced and resolved the cross-tile contradictions.