Skip to main content

Matrix Thinking

Ideas hide what's missing. A matrix makes the gaps visible. Connect dots. Fill cells. Develop strengths. Mitigate weaknesses. Delegate ownership.

"The representation is part of the cognition." - Judy Fan

Core Idea​

A story hides its gaps. A matrix shows them.

Narrative says: "I have an idea."

Matrix says: "I have a gap at these coordinates."

That is the whole move.

Pick two lists. Cross them. Every intersection becomes a cell. The cell is known, unknown, missing, or owned by someone else. The value is not the cell. The value is the shape of the grid.

A row full of unknowns means you own a tool you have not tested. A column full of gaps means a job is unserved. A cluster of delegated cells means the work wants a contract, a workflow, or a product.

Explore the space to understand what matters most. An empty cell is not nothing. It is potential with coordinates.

Worked Example​

Take two subjects: tools and jobs.

Tools are the instruments you can use: spreadsheet, LinkedIn, AI agent, specialist contractor, payment rails.

Jobs are the progress a customer needs: find prospects, qualify fit, write proposal, deliver work, collect payment.

Cross them. Then ask one question for every cell: can this tool do this job?

Some cells are proven. The spreadsheet can help find prospects. Payment rails can collect payment. Mark those as known.

Some cells are guesses. LinkedIn may qualify fit. An AI agent may find prospects. Mark those as unknown and test them.

Some cells are gaps. A spreadsheet will not collect payment. Payment rails will not write a proposal. Mark those as gaps.

Some cells should leave your hands. A specialist contractor may deliver work better than you. Mark those as delegated.

Now stop reading the cells one by one. Read the shape.

If one job has no good tool, the job is an opportunity. If one tool is full of unknowns, it needs experiments. If one job is mostly delegated, it may be ready for a handoff system. If one row is full of proof, document it and reuse it.

That is matrix thinking.

tip

Naps Fill Gaps

Read The Shape​

Use four marks.

Known means you have proof. Reuse it. Document it. Standardize it.

Unknown means you have a belief without a receipt. Run a small experiment.

Gap means no tool in the kit serves the job. Find one, buy one, build one, or leave the gap on purpose.

Delegate means another agent does the work better. Hand it off with a contract and a receipt.

The marks matter less than the pattern. A grid is a map of attention. It shows where to spend time, energy, capital, and thought.

Hunt the Void​

Edge hides where no one has coordinates.

Known unknowns are locatable. They have a ? cell waiting. Unknown unknowns have no cell, no row, no column, no axis. You cannot ask about them because the grid has no place for the question.

The hunt is not for answers. The hunt is for the meta that creates new coordinates. Name a new dimension and a new axis of void appears. That axis is where edge lives, because no competitor has drawn it.

Start with the missing axis. What dimension would name what the grid cannot see?

Then place the thing. Where does it sit?

Then test the cell. What lives there?

Then move first. Private time in a new cell is edge.

Cells age. A proof today is the answer everyone knows tomorrow. When a column fills with proof, the grid is not finished. It is dying. The dimension has stopped making new questions.

Hunt the meta, not the cell. The instrument is the axis you name, not the square you tick. Hoard the path. The route to a cell compounds even when the cell commoditizes. Watch for flood. A full column means grow the grid.

Edge is not ownership of a cell. Edge is resolution - seeing a dimension others have not drawn.

Name The Axes​

Naming is the resolution of the grid.

Taxonomy defines the dimensions. Nomenclature defines what goes in each cell. Ontology defines how cells relate. Better names give you a finer grid. A finer grid gives you more void to explore.

The data model is the grid. Understanding the data model is understanding the domain.

This is why naming standards matter. A weak name smears the cell. A strong name locates the work.

Walk The Pipe​

When learning a system, walk the pipe.

Trace the flow from input to output. Record what you see. Map each item to a standardized name. Arrange those names into a matrix. Read the empty cells.

This is how P&ID engineers survey a plant before drawing the diagram. The same method works for software, operations, sales, research, and learning.

The steps are simple.

  1. Walk the flow end to end.
  2. Name each item.
  3. Arrange the names into rows, columns, or dimensions.
  4. Read the gaps, strengths, risks, and handoffs.

Once the names are right, cartesian products generate the matrix mechanically. Every row crossed with every column gives you a cell. The cell either has proof or reveals a gap.

Price The Future​

A matrix is a prediction instrument.

You build one to forecast what time, energy, capital, and attention will return before you spend them. Each cell is a bet with a known shape: input, effort, risk, yield. Fill one cell with a receipt and the neighboring cells sharpen.

The grid turns guesswork into a schedule.

It asks:

  • How many cells can close this cycle?
  • Which rows drain energy?
  • Which columns compound?
  • What does each gap cost to close?
  • Which cell returns the most per unit of attention?

That is why the matrix exists. Not to describe the past. To price the future.

Discovery And Density​

The same grid has two modes.

Discovery asks where to look. Empty cells glow. Unknowns become prompts.

Density asks how strong the proof is. A known cell may still be shallow. A gap may be a threat, a wedge, or a decision to abstain.

Discovery finds the cells. Density grades them.

A GitHub heatmap works because it lets you see the shape before you read the detail. A good matrix does the same for work.

Excel​

See cash-flow projection for a worked example of a grid that tells the story before the numbers do.

Excel may be the most valuable thinking tool ever shipped. It won because the grid is universal and the setup cost is zero. Type headers. Fill cells. Read the story.

Its strength is also its weakness.

Anyone can read rows crossed with columns. Empty cells become visible prompts. Sparse rows expose weak coverage. Dense columns expose over-investment. A spreadsheet sells conviction because the shape of the grid speaks before the formulas do.

But each sheet invents its own meaning. Column meaning dies at the file boundary. A filled cell can be a claim with no source. Two sheets can share a word and mean different things.

The fix is not a better spreadsheet. The fix is the meta layer Excel never got: named rows, typed columns, cells bound to sources, empty cells treated as verifiable intents.

Keep the grid. Add provenance. Every cell becomes a question waiting for a receipt.

Applications​

Any two subjects whose meta you can define will produce a useful space.

Skills crossed with roles show which roles you can staff and which gaps to hire for.

Industries crossed with forces show where transformation wedges may exist. Start with Industries crossed with AI, blockchain, crypto, devices, and energy.

Principles crossed with decisions show which principle governs which move.

Agents crossed with jobs show who owns the work and who is idle.

Standards crossed with artifacts show which artifacts conform and which drift.

Define the meta before you cross the subjects. Without the meta, the grid is a doodle. With the meta, the grid is a thinking instrument that sells the story, exposes the unknowns, and routes the work.

Opportunity​

Convert one promising cell into a build thesis.

Name the industry. Name the force. Name the friction. Name the wedge. Name the first proof.

For AI data, the force may be AI. The friction may be data quality and provenance. The wedge may be verification and pricing protocols. The first proof may be one paid pilot.

For manufacturing, the force may be devices plus AI. The friction may be quality bottlenecks and rework. The wedge may be a closed-loop quality system. The first proof may be a measurable drop in defect rate.

For robotics, the force may be devices plus cloud coordination. The friction may be coordination across actors. The wedge may be shared task and settlement protocols. The first proof may be one live deployment.

If you cannot express the move in one row, keep mapping.

Question The Void​

All you can know is all you can experience. The grid gives the unknown coordinates.

A known known is a filled cell. Without the grid it is scattered knowledge.

A known unknown is the empty cell. Without the grid you sense the gap but cannot locate it.

An unknown unknown is the row, column, or axis you have not named. Without the grid it stays invisible.

Each dimension you add is a new axis of void. Industries crossed with forces gives you one surface. Add maturity stage and the surface becomes a volume. Most cells will be empty. That is the point.

More meta in the grid means more space to explore.

The empty cell is not absence. It is where you help someone see who they could become.

tip

What are the best questions you can place in each space?

Context​

  • pairs-with Knowledge Schema - How to fill the grid fast using an anchor domain you already own
  • depends-on Gates of Knowledge - Rhetoric x Senses: knowledge bounded by what we can experience
  • applies-to Inner Space - Matrix thinking applied inward: the modes, loops, and states within you are dimensions waiting for coordinates
  • applies-to Inner Game / Outer Game - What becomes possible when matrix thinking makes the invisible inner game navigable
  • depends-on Naming Standards - The resolution mechanism: better names, finer grid, more void
  • applies-to Persuasion - Matrices are assets of persuasion; the grid sells the story before the numbers do
  • instance-of A&ID Template - P&ID nomenclature applied to AI + Crypto systems
  • pairs-with Performance - proof that value is real
  • pairs-with The North Star - the fixed reference that tells you which gaps matter
  • instance-of Software Jobs To Be Done - Applied matrix: software categories x essentiality x verdict x coverage x ROI
  • depends-on Potential - why gaps matter
  • applies-to Zeigarnik Effect - The mechanism that keeps you inside the empty cell until you invent something
  • pairs-with Intelligent Hyperlinks - The pipe that carries what the matrix reveals: information to value to intent

Questions​

If all you can know is all you can experience, what dimensions are missing from your grid?

  • What row or column would you add if you had a sense you don't have?
  • When the grid reveals a gap, is the gap in reality or in your perception?
  • What's the difference between a matrix that extends your cognition and one that merely confirms what you already believe?
  • How do you know whether your grid's dimensions are named well enough to see what's actually there?