How to Identify Your Knowledge Pillars

Every business knows more than it can say. The expertise is real, but it's scattered — in the founder's head, in a hundred sales calls, in the answers you give so often you've stopped noticing them. Before you can organize what you know, you have to find it. Identifying your knowledge pillars is that first step: naming the handful of subjects you have genuine authority on, so everything you build afterward stands on something solid.

What a knowledge pillar actually is

A knowledge pillar is a core subject your business has genuine, hard-won authority on. Not a topic you could research and write a competent post about — a subject you already understand deeply, from doing the work, that you could cover thoroughly and keep covering without running dry. It's the difference between something you know about and something you simply know.

Three things make a subject a pillar rather than a passing interest. It has depth: you have enough real understanding to go far past the surface. It has relevance: clients actually care about it, because it touches a decision they're trying to make. And it has a point of view: you don't just relay the standard take, you have a specific, defensible position shaped by experience. A subject that clears all three is a pillar. One that clears only one or two is a topic — worth a mention, maybe, but not a foundation.

A useful way to feel the difference: explain the subject to someone who already knows the basics. On a topic you run out in ten minutes. On a pillar, ten minutes in you're only reaching the interesting part — the exceptions, the cases where the standard advice fails, the thing you'd only know from having got it wrong once. Depth isn't length. It's how far past the obvious you can go before you start guessing.

It's worth saying what a pillar isn't, because the word gets borrowed a lot. A pillar isn't a keyword or a category on your blog — those are labels applied afterward, and they can be renamed on a Tuesday without anything else moving. It isn't a service either, though the two get confused constantly, and it isn't a promise about the future. It's a description of what your business already knows, written down honestly, which makes the exercise closer to an audit than a strategy session. The most useful thing a pillar does isn't tell you what to write. It tells you what to decline — every subject that isn't one, however good the traffic on it looks.

Why you identify before you organize

It's tempting to skip this and go straight to organizing — building the folders, the content calendar, the clusters. Resist it. Organizing knowledge you haven't identified is like building shelves before you know what you own; you end up with a tidy structure around the wrong things. Identification comes first because it decides what everything else is for.

Get your pillars right and the rest of knowledge architecture becomes almost obvious — you know what to organize, what to write, what to be known for. Get them wrong, or skip naming them, and you scatter effort across topics that were never going to build authority, however neatly you file them. This is the discovery step that the wider knowledge architecture framework depends on: you can't structure knowledge you haven't surfaced. Spend the time here, and everything downstream gets easier.

You can see the skipped step from the outside. A business that organized before it identified has a content calendar that reads like a list of chores: sensible subjects, no thread between them, and a small dread every time the next slot comes round. Nothing is wrong with any single item; there's just nothing being built. When the pillars are named first, the calendar stops being a list of things to produce and becomes a sequence of pieces covering a subject you intend to be known for. Same amount of writing, entirely different feeling, because the work accumulates instead of resetting every week.

It also saves the more expensive mistake. Organizing takes a weekend; producing against the wrong pillars takes a year — and the structure hides the error, because a tidy system looks like progress right up to the moment somebody asks what it got you.

Where pillars actually come from

The mistake most people make is trying to invent their pillars — brainstorming an impressive list from scratch. Don't. Your pillars already exist; they're just unnamed. The work is discovery, not invention, and there are four reliable places your expertise is already visible if you look.

The first is the questions you answer over and over. Every business has them — the same handful of questions that come up in nearly every conversation. You've answered them so many times they feel obvious, which is exactly why you overlook them. They're obvious to you because you're an expert; they're gold because they mark the subjects clients keep needing your help with. The second is your sales conversations. What do prospects ask before they trust you? What objections come up? What do you find yourself explaining to move a deal forward? Those topics are pillars, because they're the knowledge that actually earns business.

The third is the decisions clients hire you to get right — the high-stakes choices where your judgment is the thing of value. Whatever sits under those decisions is a pillar, because it's where your expertise matters most. The fourth is the simplest: what could you talk about for an hour, with no notes and no research, and still have more to say? That test cuts straight through pretension. It doesn't find what you wish you were expert in. It finds what you actually are.

The four sources working together, briefly. A firm that does bookkeeping for restaurants ran this exercise expecting "bookkeeping" to be the answer. What actually surfaced was narrower and far more useful: getting a new restaurant's books set up cleanly in the first month, reading a profit-and-loss statement to catch food-cost creep, and staying audit-ready for a surprise sales-tax review. Those were the questions clients kept asking, the topics that moved deals forward, and the decisions owners hired them to get right. "Bookkeeping" was the service. Those three narrower subjects were the pillars.

There's an obvious fifth place people look, and it's the one to avoid: what your competitors publish. Their blog tells you what the market has already covered and nothing about what you know that they don't. Keyword tools have the same flaw in a better costume — demand is real information, but it's a filter you apply to a list you already have, not a way to make the list.

Three tests for a real pillar

Once you've got a list of candidates from those four sources, you'll have more than you need, and some won't hold up. Three tests separate the real pillars from the topics that snuck onto the list. The depth test: could you cover this thoroughly, from many angles, over a long time — or would you exhaust it in one post? If it runs dry fast, it's a topic.

The relevance test: does this connect to a decision your clients actually make, or is it just interesting to you? Expertise nobody needs isn't a pillar, however deep it goes. The distinctiveness test: do you have a genuine point of view here, or would your take be indistinguishable from everyone else's? A subject where you'd only repeat the standard advice can't carry authority, because there's nothing in it that's yours. A candidate that passes all three — deep, relevant, and distinctly yours — is a pillar. Be honest applying them; the temptation is to wave through subjects you like rather than ones you own.

There's a fourth signal that isn't quite a test but decides how long a pillar is worth building on: durability. Will this still be true in five years? A subject tied to one platform's current behaviour, or one tool's current feature list, needs rewriting before it ever compounds. The subject underneath it — why the platform behaves that way, what decision the tool is helping someone make — moves far more slowly. When two candidates are close, take the one that ages better.

The distinction between a topic and a pillar is easier to feel than to define, so here it is laid out signal by signal.

Read down the right-hand column and the five signals turn out to be one fact from different angles: you know it deeply because you've done it repeatedly, you've done it repeatedly because clients keep needing it, and your view is distinct because the repetition taught you what the general advice leaves out. That's why a genuine pillar passes all five at once, and why a candidate that passes two isn't a pillar with a fixable weakness. It's a topic.

How many pillars you should have

The right number is a handful — roughly three to six. That range isn't arbitrary. Too few, and you've under-described what you know, leaving real expertise off the map. Too many, and you've stopped identifying pillars and started listing topics, which spreads your effort so thin that no single subject ever gets deep enough to build authority.

If your list has twenty entries, the problem isn't that you're unusually knowledgeable — it's that you're mistaking topics for pillars. The fix is to group and abstract: those twenty are almost always five or six real pillars with their sub-topics mixed in. Find the few themes the long list clusters around, and name those. The sub-topics don't disappear; they become the pieces inside each pillar. Pillars are how you cover a subject deeply, not a catalogue of what you could mention.

When you group, group by the argument rather than by the category. Write each candidate on its own line, then look for the sentence that covers three of them at once — not a heading like "operations", but a claim: "getting the first month right decides the next year". If that's a sentence you'd defend, it's the pillar and the three lines under it are its pieces. If the only sentence you can find is a category name, those three lines probably belong to different pillars. Categories are for filing. Pillars are for arguing.

Too few is the rarer failure, but it's real: a business that sounds narrower than it is, publishing on one subject for three years while two others were doing just as much work in the sales calls. If you can only name one or two, go back to the four sources and read the raw list again without editing as you go. Under-describing what you know costs you the client who needed the second pillar and never found out you had it.

Which pillar to build first

You won't build them all at once, so identification ends with a decision about sequence. Don't start with the pillar that's most impressive or the one you personally find most interesting. Start with the one that sits closest to a buying decision — where being known as the authority most directly earns you business. Depth plus commercial relevance is where the first pillar should be.

A useful way to choose is to weigh two things for each pillar: how much genuine authority you have in it, and how directly it connects to work clients pay for. The pillar that's high on both is where you begin — it's where your expertise converts most quickly into trust, and trust into conversations. The others follow in their own time. What matters is sequencing deliberately instead of spreading attention evenly and building depth in none. Once you've chosen, turning that pillar into a connected body of published work is its own craft, which we cover in content clusters: the bridge between knowledge and content.

In practice, "first" means one pillar gets the next quarter and the others get a folder and a list: when a good idea arrives for pillar three, you write the line down in pillar three's folder and go back to pillar one. Be strict about it, because authority is a threshold rather than a slope. Five pieces on one subject reads as a point of view; one piece on five subjects reads as a blog.

Common mistakes

A few predictable errors derail this step. The first is aspirational pillars — naming subjects you wish you were known for instead of ones you genuinely own. They collapse the moment you try to produce depth, because the depth was never there. The second is mistaking your services for your pillars. What you sell and what you're an authority on overlap, but they're not the same; your pillars are the knowledge beneath your services, not the services themselves.

The third is going too granular, listing every sub-topic as its own pillar until you've got a sprawling list and no real structure. The fourth is skipping the honest tests — letting subjects onto the list because they sound good rather than because they're deep, relevant, and distinctly yours. Every one of these comes from the same place: describing the business you'd like to have instead of discovering the one you do. Pillars are found, not invented, and the method is only a way of looking honestly.

The fifth mistake is subtler: treating the list as permanent. Pillars do change, only slowly — a business doing genuinely different work in three years should have at least one different pillar, and keeping the old list out of loyalty is its own kind of dishonesty. Six months on, the way to tell an aspirational pillar from a real one is to open the folder rather than reread the list. A real pillar accumulates; an aspirational one holds exactly what you put in it the day you named it.

A worked example: counting what you already answer

The four sources are a thinking exercise, but the first of them — the questions you answer over and over — leaves a trail you can actually read. Most businesses already have the raw material: recorded sales calls, a support inbox, a shared channel where the same question arrives from a new person every few weeks. What they don't have is a count. Here's how we do it, with the tools we build on.

Start with material you can read in one sitting. Export the transcripts of your last thirty calls into a folder, open that folder in Claude Code, and ask for the questions the client asked — verbatim, one per line, no summarising. Verbatim matters more than it sounds. A summary quietly merges "how long until my books are clean?" with "when do I get to stop worrying about this?", and those two questions have different pillars underneath them. You'll get a few hundred lines. Ask for them grouped by what they're really asking about, then read the groups yourself: the grouping is a suggestion, not a finding.

Then make it standing rather than one-off, which is where n8n or Make.com earns its place. A board that fires after each call, pulls the transcript, asks for the questions it contained, and appends them to a single file with a count beside each line will tell you in two months what a workshop can only guess at: which questions your business answers most, and how far ahead of the next one they are. That gap is where your first pillar is. Zapier does the same job perfectly well if that's what you already pay for — the shape of the automation matters more than which board it runs on.

Be clear about what the machine did and didn't do. It counted; it didn't decide. It has no idea which of those questions you answer well, which one bores you, or which one sits under a decision worth real money to the client. A count is evidence, and it's evidence you didn't have before, but a pillar is a judgement about depth and distinctiveness and that lives in the people who did the work. The useful moment is when the count and the intuition disagree — that's the conversation worth having, and it's the one you can't have without the count.

Setup is about an afternoon and it keeps working afterward, which is the general shape of everything we build. It doesn't produce your pillars; it produces the evidence that turns naming them into an argument about facts rather than taste.

What this honestly costs

Identifying pillars is one of the cheapest things in this discipline, which is exactly why it gets skipped — cheap work feels optional. Here's the honest accounting, because the hours are not where the cost lands.

The first pass is an hour, on your own, writing badly. The list will be too long and partly embarrassing, and both are fine. The expensive part isn't the writing, it's the week after: you go back to the list with the four sources in your head, having noticed what actually came up in that week's conversations. A list made in one sitting is a list of what you remember about your business. A list revised a week later is a list of what your business does.

Then find someone who knows the work and argue about it for two hours. This is the step people skip, and it's the one that does the most, because the tests only bite when someone else is applying them to you. "Could you talk for an hour on that?" is easy to answer generously on your own and hard to answer generously with a colleague looking at you. Expect to lose two or three candidates in that conversation, and to gain one you'd never have listed because it was too obvious to you to be worth saying.

After that it's a yearly review — an hour, same as the first pass, asking whether the list still describes the work. The real cost isn't hours at all, though. It's what you give up: naming three to six pillars means agreeing not to publish on the other twelve subjects on your list, including a couple you're fond of. That's the part that stings, and it's also the entire benefit. A business known for three things is known. A business publishing on fifteen is publishing.

Frequently asked questions

What is a knowledge pillar?

A knowledge pillar is a core subject your business has genuine, hard-won authority on — deep enough that you could cover it thoroughly, relevant enough that clients actually care, and distinct enough to carry your specific point of view. Pillars are the foundation your content and your whole knowledge architecture are built on, so getting them right matters more than almost anything downstream.

How many knowledge pillars should a business have?

Usually a handful — roughly three to six. Few enough that each one is genuinely deep and you can cover it thoroughly; enough to represent the real range of what you know. If you've listed twenty, you've listed topics, not pillars, and you'll spread yourself too thin to build authority on any of them.

How do I find my knowledge pillars?

Look where your expertise already shows up rather than inventing a list from scratch: the questions you answer over and over, the topics that fill your sales conversations, the decisions clients hire you to get right, and the subjects you could talk about for an hour without notes. The patterns across those four sources are your pillars — you're discovering what's already there, not manufacturing it.

What's the difference between a pillar and a topic?

A topic is anything you could write about; a pillar is a subject you have real depth and a point of view on, that clients care about, and that stays relevant over time. Every pillar is a topic, but most topics aren't pillars. The test is whether you could cover it thoroughly and distinctively, or only touch it once before running out of anything original to say.

What's the difference between a knowledge pillar and a service?

A service is what a client buys; a pillar is the knowledge underneath it. They overlap, but they're rarely the same shape — a firm that sells bookkeeping to restaurants has pillars like setting a new kitchen's books up cleanly, reading a P&L for food-cost creep, and staying ready for a surprise sales-tax review. Naming the service as the pillar is the most common version of this mistake, and it produces content that reads like a brochure because that is what it is.

How often should knowledge pillars change?

Review them once a year and expect them to move slowly. A list that flips every quarter was a list of topics; a list that hasn't moved in five years while the work has changed is loyalty rather than accuracy. The usual pattern is one pillar in three shifting over a couple of years, as a business gets clearer about what it is actually good at.

What if a pillar feels too broad?

If you can't cover it in a finite set of pieces, it's a category rather than a pillar — split it by the decision underneath it. "Marketing" is a category; "deciding what to publish before you have an audience" is a pillar. The test is whether you can state a sentence about it you'd defend against your own industry. If the only sentence you can make is a definition, go narrower.

Where to start

Don't overthink the first pass. Set aside an hour and write down every subject that shows up in the four places your expertise already lives — the questions you keep answering, your sales conversations, the decisions clients trust you with, and the things you could talk about endlessly. The list will be messy and too long, which is exactly right. Then run it through the three tests, group the survivors into the few themes underneath, and you'll have three to six pillars that genuinely describe what your business knows.

If it helps, set the bar at one sentence: "I'd be comfortable being introduced as the people who know about this." That does the work of all three tests at once and is much harder to fool yourself with than a checklist. Whatever survives it is your foundation, and everything else in your knowledge architecture — the folders, the clusters, the calendar — is downstream of that list being honest.

Keep reading

• The knowledge architecture framework — the system your pillars sit inside.
• How to organize business knowledge — what to do once you've named your pillars.
• Content clusters: the bridge between knowledge and content — turning a pillar into published work.