How to Build Social Media Content Pillars.

A content pillar is a durable theme a brand publishes against month after month, specific enough to rule ideas out and broad enough to produce work every week.

Most brands need three or four, each with a stated job and a rough share of the monthly output. Together they decide what goes on the calendar before anyone opens the calendar.

Most pillar sets never get that far. Education, inspiration, community, promotion: those are moods, and a mood cannot reject anything, so the calendar still fills with whatever turned up that week.

A pillar that works is a filter. It tells the team what to make, what to repeat and what to stop, and it holds that line when a stakeholder arrives with a favour to ask.

What a content pillar actually is

A pillar is a theme the brand commits to for at least a year, held closely enough that the audience would eventually associate it with you rather than your category.

Three properties make it real. It is specific to your brand. It can be produced repeatedly with the people and access you already have. And it can reject an idea.

That last one is the test worth keeping. Name three good ideas a pillar excludes. If you cannot, you have written a heading.

A heading looks fine in a strategy document and changes nothing about what gets made on Monday morning, which is why brands with a documented pillar set still publish at random.

Why three or four pillars, and what breaks at seven

Three arguments, all pointing the same way.

The first is recognition. An audience only registers a theme it meets repeatedly. Imagine a brand publishing twelve times a month: across three pillars that is four posts each, which reads as a pattern. Across seven it is under two, which reads as a shuffle.

The second is judgement. You cannot review a pillar on five posts in a quarter. At that sample every reading is noise, so nobody can say which theme is working and the set never changes.

The third is production. Every pillar carries its own research habit, its own sources and its own formats. Seven pillars means seven of those, run by a team that usually has room for three.

Fewer than three brings its own problem: one note, repeated, with nowhere to go when the brand needs to say something else. Three or four, plus deliberate space for reactive posts that belong to no pillar at all.

A pillar is not a format, and it is not a campaign

Teams conflate all three, and the calendar pays for it.

• A pillar is a theme with no end date. It survives the quarter, the year and the person who wrote it down.
• A format is the shape a pillar takes: the sixty-second explainer, the myth-buster carousel, the weekly customer question. One pillar carries several.
• A campaign is a dated push with a start, an end and a target. It borrows from the pillars while it runs.

Name a pillar after a format and the theme dies the month that shape stops working. Name it after a campaign and it leaves a hole in the mix the week the campaign closes.

The clean split: pillars decide what you are known for, formats decide what you make, campaigns decide what you push this month.

Step 1: Collect the raw material before you name anything

Pillars invented in a workshop tend to describe the brand the team wishes it worked for. Start with what already exists, in three places.

• What the business actually knows. The expertise already in the building, the arguments the founder has had a hundred times, the things your delivery team can explain without preparation.
• What customers actually ask. Sales calls, the inbox, comments, the same three questions in every onboarding, and the places your buyers talk to each other. Listen to unfiltered community conversation before planning content, because people describe their problem there in language no survey returns.
• What the team can sustain. The access, subject time, footage and edit capacity you have in an ordinary week, not in a good one.

A pillar lives where all three overlap. A theme with demand and no capacity is a wish. A theme with capacity and no demand is a hobby.

Step 2: Name a pillar so it constrains a decision

The name is not a label. It is the filter the team applies fifty times a month, so it has to carry a decision.

“Education” rejects nothing, because everything teaches somebody something. Written as a decision, it becomes something a brand can own: imagine a skincare brand running “skin decisions made simpler”, a gym running “real member progress”, a services firm running “what the work actually looks like”.

The test takes ten minutes. Hand ten rough ideas to someone outside the team and ask them to sort each one in or out. If two people sort the same ten differently, the name is not finished.

One more rule. Name the pillar in the audience’s language, not the org chart. Pillars named after departments describe who makes the content rather than who it is for, and they age with your structure instead of your market.

Step 3: Give every pillar one job

The name says what the pillar is about. The job says what it is meant to change.

Most jobs are one of four: reach people who do not know you, make the offer legible, prove you can actually do the work, or give the brand a position worth remembering.

Write the job next to the name and keep it to one sentence. Two pillars with the same job are one pillar with two names, so delete one and take the space back.

The job is also what you will judge the pillar on later. A trust pillar measured on reach will always look like a failure, and good pillars get retired that way every quarter.

Step 4: Test a pillar before you commit to it

Run each candidate for four to six weeks before it goes into a strategy document. You are testing production, not performance: at that sample the numbers say nothing useful, and the failure that kills pillars is the one where a theme cannot produce.

Four questions to answer at the end of it:

• Could the team brief it without calling a meeting?
• Did it produce posts that do not look like each other?
• What blocked it? Access, approval, a subject who was never free.
• Can someone outside the team describe the theme without being shown the label?

A candidate that needed three chases to produce two posts is not a pillar yet. It is a project, and it will stop quietly the first week the calendar gets busy.

Step 5: Size the mix before you open the calendar

Pillars are not equal, and an even split is a decision not to decide.

Give each one a rough share of the month, set against its job rather than how much the team enjoys making it. Imagine a founder-led services brand with three pillars: buyer education taking about half the output, proof about a third, and the offer pillar the rest. Those numbers are a starting position, not a formula, and the first quarter of real data should move them.

The mix is a budget, not a schedule. You check it at the end of the month against what actually published, which is usually a different set of posts from what was planned.

Write it down. Without a stated mix, the loudest stakeholder wins the calendar.

Turning pillars into a calendar without building a rota

The common mistake is turning the mix into weekday slots. Education on Monday, proof on Wednesday, culture on Friday.

Now the team is feeding slots. The weakest post of the month goes out because Wednesday needed something, and the strongest idea waits nine days for its allotted lane.

Plan the month against the mix instead, then order by what is ready and what is timely. Leave some slots deliberately open, because a calendar with no slack turns every opportunity into an argument.

One rule keeps the whole thing honest: a post serves one pillar, because a post that serves three serves none.

What clear pillars do to feedback

The quiet benefit is not planning. It is the end of subjective feedback.

Without pillars, one person wants the post more commercial, another wants it warmer, a third asks why it exists at all, and the strongest opinion in the room wins.

With pillars, the question becomes whether this post does the job of its pillar, rather than whether the room likes it. That is a question a junior can answer and a director can be wrong about.

Which is why pillars belong in the brief template and the approval process, not only in the strategy deck. A standard nobody applies at the point of review is not a standard.

The retirement test: when a pillar has stopped earning its slot

Pillars should be stable, not permanent. The failure in most teams is not changing them too often, it is never retiring one. A pillar stays in the mix because it is in the document, taking a share of the calendar it stopped deserving two quarters ago.

Run five questions per pillar, once a quarter:

1. Did it produce on schedule, or did every post need chasing?
2. Judged against its own job, is it doing that job?
3. Does it still match what customers ask, or is it answering a question they stopped having?
4. Can the team still name three ideas it excludes?
5. Building the mix from scratch today, would you defend its share of next month?

Question five decides it. A pillar you would not choose again is finished, whatever its engagement numbers say.

Dead pillar, or badly executed pillar?

Most pillars declared dead were never properly tried. Telling the two apart is the difference between fixing a craft problem and throwing away a theme the audience actually wanted.

A badly executed pillar is uneven. Some posts land properly, the best one is genuinely good, and the failures share a craft problem: a weak opening, the wrong length, the wrong platform, a format the theme never suited.

A dead pillar is flat. Nothing spikes. The best post is barely distinguishable from the median post, and it fails the same way regardless of who makes it or what shape it takes.

The separating test is cheap. Keep the theme, change the format, run it for a month. A live theme moves when it gets a new shape. A theme that fails in three formats, on the platform that should suit it, made by people who understand it, is the problem itself.

Two things look like death and are not. A pillar that collapsed when the person carrying it left is unstaffed, which is a resourcing decision. And a pillar pulling good numbers from the wrong audience is worse than a quiet one, because the numbers hide the failure.

What to do with a pillar that fails

Retirement is the last option, not the first. In order of cost:

1. Sharpen the name. Loose definitions cause more pillar failures than bad themes do.
2. Change the format. The cheapest real test, and it settles the question above.
3. Merge it. Two half-working pillars with the same job usually make one that works.
4. Retire it, and reassign its share of the mix on purpose. Freed slots refill with whatever is easiest to produce unless somebody decides otherwise.

Retire it in the document, not just in practice. An undocumented retirement stays arguable, and the pillar returns the first time someone opens last year’s strategy.

One retirement a quarter is a healthy rate. A team replacing half its pillars every quarter does not have a pillar system, it has a rolling opinion.

What good looks like

• Three or four pillars, each with one written job.
• A stated monthly mix that survives contact with the calendar.
• Briefs that name the pillar and its job before they describe the post.
• A quarterly review by pillar, not only by post.
• At least one retirement decision made on evidence rather than boredom.

None of that makes a brand more controlled. It makes the route from idea to published shorter, because most of the arguing has already happened.

A pillar also holds its meaning across channels while the execution changes by platform, so a good set of themes does not turn into the same post published four times.

How NBK thinks about content pillars

Pillars are an operations artefact, not a slide. They belong in the brief template, the approval standard and the monthly review, because those are the three places a decision actually gets made.

That view comes from running social at publisher scale, where the team has shipped more than 46,000 posts. At that volume a pillar set is the thing that stops a calendar becoming a queue. It decides what earns a slot before the week starts, and it makes the review afterwards about the theme rather than the individual post.

The brands that get this right are rarely the ones with the best ideas. They are the ones whose good ideas do not have to be re-argued from scratch every Monday.

Next step

Pull last quarter’s posts into a list and sort them into your current pillars. Count what will not sort.

If most of it will not sort, the pillars are not running the calendar and something else is. If it all sorts but one pillar is thin, you have a capacity problem rather than a strategy problem.

If your social output feels busy but not effective, an NBK audit is a sensible place to start. It looks for the constraint in the system instead of adding more content on top of it.