Marketing teams love a single number. Two thousand words. Fifteen hundred minimum. Long form wins. None of those rules survive contact with a SERP. Some queries rank 400-word definitions. Others need 3,000 words of worked examples. Length is an output of the brief, not a strategy.
Why word count targets fail
Volume-based targets encourage padding. Writers add background nobody asked for, repeat the same point in three sections, and bury the answer. Readers bounce. Editors cut. The post takes longer to produce and performs worse. Google does not reward length; it rewards satisfaction for the query.
The opposite mistake is equally common: cap everything at 800 words because people do not read. Some audiences do not read long posts. Many B2B buyers do, when the topic is complex and the alternatives on page one are thin. The SERP tells you which case you are in.
Read the SERP before you set length
Open an incognito window, search the target keyword, and read the top five organic results. Note format, depth, and what they skip. Export word counts if it helps, but read the pages. A 1,200-word post that answers the question beats a 2,500-word post that circles it. Our guide on reading a SERP for search intent walks through that process step by step.
Match the dominant format first. If every result is a numbered list with short blurbs, a narrative essay will fight the SERP even at the right word count. If results are long guides with tables and screenshots, a 600-word opinion piece will not compete on coverage.
Length by intent type
| Intent type | What the reader needs | Typical range | Risk if wrong |
|---|---|---|---|
| Definition / what is | Clear answer in the first screen | 400 to 900 words | Burying the definition below fluff |
| How-to | Steps, prerequisites, failure modes | 900 to 2,000 words | Skipping edge cases readers hit |
| Comparison | Criteria, tradeoffs, honest limits | 1,200 to 2,500 words | Feature lists without decision help |
| Thought leadership | Original argument and evidence | 800 to 1,800 words | Recycled takes with no thesis |
| Programme pillar | Comprehensive coverage and links | 2,000 to 4,000 words | Thin sections that should be separate posts |
Ranges are starting points. Your brief should state a target band after SERP review, plus permission to go shorter if the outline is complete or longer if a section needs worked examples. Rigid caps cause writers to stop mid-argument.
Outline depth beats a word count mandate
Build the outline before you set length. Each H2 should earn its place: answer a sub-question, handle an objection, or provide a worked example. If the outline has six thin sections, the post will feel long and empty at 2,000 words. If it has four dense sections, 1,200 words may be enough.
For blog writing programmes, we brief with section purposes, not padding quotas. Editors cut throat-clearing intros and duplicate H2s. Writers expand where the SERP shows gaps. The final count is a measurement, not a goal.
When longer posts make sense
Longer posts earn their length when the topic has real complexity: compliance, implementation, pricing models, or multi-stakeholder decisions. They also make sense for pillar pages that anchor a cluster, provided each section could stand alone as a useful answer.
- The SERP top five average depth is high and still misses a subtopic you can own.
- The sales team gets repeated questions that one URL should answer.
- You have original data, screenshots, or SME quotes that add information gain.
- Internal linking strategy needs a hub page with clear child links.
Do not lengthen a post to fix thin content. Split it. Two focused posts often outperform one bloated URL, especially when keywords represent different intents. See refresh versus new content for when to expand in place versus publish separately.
When shorter posts make sense
Shorter posts work for news reactions with a short shelf life, glossary entries, release notes with a single job, and answers where the SERP favors brevity. They also work when your angle is a sharp opinion with one thesis and three supporting points.
Short does not mean shallow. A 700-word post that names a problem, states a position, and gives three actionable steps can outperform a rambling guide. Edit aggressively. Cut the second intro paragraph. Merge overlapping H2s. Delete sentences that repeat the heading.
Measure satisfaction, not syllables
Track whether the URL ranks for the target cluster, whether scroll depth shows readers reach key sections, and whether the sales or support team still gets the same questions. Word count is a production metric for capacity planning. It is not a success metric for the page.
Brief writers with intent, outline, and a flexible range. Review drafts against the outline and the SERP, not against a number in a spreadsheet. That is how you get length that helps the reader instead of length that checks a box.
Word count in editorial review
Editors should flag posts that hit the word target but miss the outline. Padding often hides in generic intros, repeated H2s, and background sections the SERP does not reward. Cut first, then reassess whether the piece needs another example, not another hundred words of setup.
When a draft is short but complete, resist assigning arbitrary expansion. Ask whether an additional case study, table, or FAQ entry adds information gain. If the SERP leaders are concise and the question is answered, publish the shorter draft and monitor performance.
Programme dashboards can show word count bands by intent type without turning them into quotas. Track median length per cluster over time so outliers stand out. A sudden spike in average length often means briefs lost their outline discipline.
When stakeholders ask for more words, ask which outline section is thin. That reframes the conversation from length to reader value and keeps editors from approving padding to close a ticket.
Padding is obvious to readers and to algorithms
A 3,000 word article is often a 900 word article with repeated definitions, generic tips, and throat-clearing. Readers bounce. Editors should cut. Word count targets are ceilings and planning tools, not moral achievements.
Size to the SERP and the angle. If competitors rank with tight explainers, match the job and add information gain instead of matching their worst padding. If the query needs steps, examples, and edge cases, longer is honest.