“Make it 2,000 words” — the most common and most wrong SEO brief ever written. Length follows intent: a definition answered in 400 words outranks a 3,000-word wander that never answers. Still, patterns exist across thousands of ranking pages. This is how long a blog post should be: intent-first ranges, what earns length, and the padding traps that sink wordy posts.
Part of the word counter guide. Measure drafts in the free word counter; check topical health in the density tool.
Length ranges by intent (heuristics, not requirements)
| Intent | Typical range | Why |
|---|---|---|
| What-is / definition | 800–1,300 | One concept, fully answered |
| How-to | 1,500–2,500 | Steps + examples + troubleshooting |
| Comparison / best-X | 2,000–3,000 | Multiple entities fairly covered |
| Pillar / ultimate guide | 2,500–4,000+ | Comprehensiveness earns links |
These describe what comprehensive coverage usually takes — Google ranks relevance and depth, never the number. A 900-word page that fully satisfies beats 3,000 words of wandering, every time.
What earns length (and what is padding)
- Earns: worked examples with numbers, comparison tables, troubleshooting sections, original test data, FAQ answering real PAA queries.
- Padding: restated introductions, generic background (“since the dawn of…”), synonym-stuffed repetition, stock anecdotes with no data.
- Test: delete any paragraph — if no sub-question goes unanswered, it was padding. Our 40-post blog gets this delete-test quarterly — intro throat-clearing is usually the first to go.
- Student parallel: examiners reward argument density per 100 words, not total words — same principle, different judge. Essay tactics in the pillar's India limits section.
Pillar vs how-to: different length jobs
Pillars earn links by mapping a topic (broad, 2,500+, hub of clusters); how-tos earn ranks by solving one task (1,500–2,500, steps first). Our own silos follow this: each 2,500-word pillar anchors 9 focused clusters. New blogs should publish one pillar plus 3–5 clusters minimum before expecting topical authority — single orphans rarely move. Track depth with counts plus readability (grade ≤8 general) via the readability checker.
Updating old posts: length decisions that compound
Posts decay: stats age, screenshots rot, competitors out-cover you. Quarterly, sort published posts by impressions-without-clicks (Search Console) — prime expansion candidates. Update playbook: add the missing sub-section competitors cover (+300–600 words of real coverage, never padding), refresh every number and screenshot, re-verify readability, and update the date honestly (republish notes beat silent date changes for trust). In one case I advised, merging 11 thin posts into 3 guides tripled that cluster's traffic in 4 months — single case, not typical, but maintenance compounds exactly like creation, and counts tell you which posts deserve the effort.
General writing guidance only. Test with real readers; formulas ignore tone and expertise. SEO outcomes vary — no count, score or density promises rankings.