Pillars aren't brainstormed from a blank page — they're mined from the brand profile, tested, and selected. This is the method.
The five sources (mine each for candidates)
Read brand-profile.md and pull candidate themes from each source. Aim wide first — 8–12
candidates — then cut.
1. Positioning + point of view → the signature pillar
The brand's differentiator and its strong beliefs are the richest source. The POV ("most tools are bloated because they're built for enterprises") becomes the signature theme ("software that respects small businesses"). This pillar is what the brand wants to be known for.
2. Audience jobs-to-be-done + pains → audience pillars
Look at what the audience is trying to do and what frustrates them. Their problems are pillars; your features are not. "New founders terrified of payroll" → a pillar about making the scary thing simple. These pillars earn relevance because they're about the audience, not the brand.
3. Proof + expertise → credibility pillars
What can the brand teach authoritatively? Where's the track record, the data, the hard-won lessons? That expertise is a pillar the brand can own because few others can teach it as credibly.
4. The human / brand story → a connection pillar
The people, the process, the journey, the values. This is what builds loyalty and makes a brand followable rather than just useful. For personal brands this is often central.
5. Offers / product → at most one conversion pillar
The product and offers can anchor one pillar, used sparingly. Frame it around the transformation it delivers, not a feature list. Keep it the minority of the mix.
From candidates to pillars
- Generate 8–12 candidates across the five sources, phrased as subjects ("small-business money mistakes"), never as intents or formats.
- Test each with the four-part pillar test (ownable / relevant / strategic / sustainable —
see
what-makes-a-pillar.md). - Cluster overlapping candidates into bigger themes (three narrow topics may be one pillar).
- Select 3–5, ensuring:
- exactly one signature pillar,
- a spread across audience problems, credibility, and connection,
- at most one conversion-linked pillar,
- a balance across goals (not all-teaching, not all-selling).
A fast sanity check on the final set
- Could a stranger read the pillars and immediately understand what this brand is about?
- Does the set include the brand's differentiator (the signature pillar)?
- Is every pillar about something the audience cares about (not just the brand)?
- Could you fill a month's calendar from these without straining? (If not, a pillar is too narrow — or you have too few.)
- Is each pillar phrased in language the audience actually searches? TikTok, YouTube, and
Pinterest are search-driven; a pillar named in the audience's own query language (see
audience-research's search-query evidence) compounds there, while internal jargon doesn't. - Is promotion the minority?
If the set fails the "what is this brand about?" test, you have intents/formats masquerading as pillars — go back to the sources and re-derive real themes.
