Content pillars are the three to five topics you post about on repeat, chosen because your audience has already shown they want them from you. A useful pillar is a bet backed by evidence from your own posting history. A topic bucket you picked in a brainstorm is a guess wearing a strategy costume. So if you're asking what are content pillars and how do I pick mine, the honest answer is that yours are already sitting in your last 90 days of posts.
What are content pillars in plain terms?
Think of pillars as the lanes your content drives in. If a stranger watched ten of your videos back to back, the lanes are whatever they'd say you talk about. For a fractional CFO, that might be pricing, cash flow, and what to fix before you hire. For a career coach, it might be salary negotiation, difficult managers, and quitting well.
One clarification that saves people a lot of confusion: pillars are topics, and formats are containers. "Talking-head video" is a format. "Pricing mistakes" is a pillar. You can pour any pillar into any format, and the pillar is what tells your audience who you are.
Pillars do two jobs at once. They teach the platform who to show your content to, because repeated topics establish your niche faster than anything you write in your bio. And they teach your audience what to expect from you, which is the difference between someone tapping follow and someone thinking "that was nice" and scrolling on.
Notice what that definition leaves out: your opinion about what you should post. That part trips up almost everyone, so let's deal with it directly.
Why topic buckets fail
Most advice tells you to pick pillars by brainstorming. Write down what you know, what you enjoy, and what your customers ask about, then find the overlap. It sounds reasonable. I did a version of this when I started posting, and I'd bet you did too.
The problem is that a whiteboard cannot tell you which topics your audience sends to their friends. Only your numbers can. I've watched smart people grind for months in a lane they picked at a kitchen table while one "off-topic" post outperformed everything else they made. They called it a fluke. It was a vote.
A pillar is a topic your audience picked for you, with receipts.
You still get a say, because you have to like a lane enough to stay in it for a year. But your audience votes first, and their ballot is made of shares and saves rather than applause.
How do you find content pillars from your own data?
You need your post history and about 45 minutes. I call this the top 20 method:
- Pull your last 60 to 90 days of posts from TikTok or Instagram analytics. A spreadsheet helps but isn't required.
- Rank every post by shares plus saves. Skip likes and views for now. Shares and saves track much closer to reach because they cost the viewer something, and I wrote the full argument in shares and saves vs likes.
- Take the top 20 posts. Set the rest aside.
- Label each one with its topic in five words or fewer. "Pricing mistakes." "How I fired a client." "My morning before a launch."
- Cluster the labels. Any topic that appears three or more times is a pillar candidate. Treat anything that appears once as an experiment still waiting on evidence.
- Keep three to five clusters and cut the rest, even the ones you love.
This is the exact math Batchly runs when you upload your analytics export. It scores every post with shares and saves weighted heaviest, then clusters your winners into pillars, so the 45 minutes becomes about two.
What this looks like with real numbers
A leadership coach I'll call Dana had four buckets a course handed her: mindset, leadership, productivity, and behind the scenes. She posted 68 videos in three months. When she ranked them by shares plus saves, her top 20 told a different story. Nine were about handling hard conversations at work. Six were about first-time manager mistakes. Three were about how she runs her week. Two were one-offs.
Her mindset bucket, a third of her total output, put zero posts in the top 20. That's roughly 22 videos, scripted, filmed, captioned, and posted, for an audience that shrugged. Dana rebuilt around the three pillars her data chose: hard conversations, new manager survival, and how I run my week. She didn't post more. She stopped spending a third of her energy in a lane nobody wanted, and her average reach climbed over the next two months because more of her output lived where demand already existed.
How many content pillars should you have?
Three to five. Under three, you'll get bored and your feed turns repetitive in the bad way. Over five, your data gets too thin to read. If you post three times a week across six pillars, each pillar gets two posts a month, and two data points prove nothing.
Pillars also expire. Audiences shift, your offer changes, a lane runs dry. Re-run the top 20 method once a quarter and let the rankings argue with you.
Weight your lanes unevenly, too. If one pillar produces half your top 20, let it carry half your calendar. Equal airtime across pillars feels tidy and wastes reach. A workable split for a three-pillar account is 50 percent to the strongest lane, 30 to the second, and 20 to the one you're still testing. Your data sets the ratio, and the quarterly re-run adjusts it.
Two related decisions sit next to this one. If you're still choosing the broader territory, start with how to pick a content niche, since pillars live one level below that call.
And when a pillar starts to feel empty, the refill is your own back catalog, which I cover in never run out of content ideas.
Do this next
Open your analytics tonight, rank your last 90 days of posts by shares plus saves, and label the top 20. You'll know your pillars before you finish your tea, and you'll probably retire a bucket you've been forcing. If you'd rather skip the spreadsheet, Batchly does the scoring, the clustering, and the next round of scripts from one analytics upload.