How to write a listicle that still works
The list post did not die when AI arrived. The lazy version did. Here is the difference, and how to write the kind that survives.
100OUT OF 100
A listicle is a post built as a numbered list: "7 ways to…", "12 tools for…", "5 mistakes that…".
They have a bad reputation, and some of that is earned. A lot of list posts are padding with headings. But the format itself is sound. It promises the reader exactly what they will get, it is easy to scan, and it is easy to write in one sitting.
The question is not whether to write listicles. It is how to write one that is worth reading.
Short answer
A listicle works when each item earns its place and can be understood on its own. Pick a number you can fill honestly, order the items by usefulness, and put a real answer under every heading, not a sentence of filler.
What a listicle is actually for
A list post does one job well: it lets a reader take what they need and leave.
Almost nobody reads a "9 tools" post start to finish. Most people scan the nine headings, stop at the two they have not heard of, read those, and go. That is a success, not a failure. Design for it.
This is why the format survives. It matches how people read on the internet, and it matches how AI answer boxes pull information, as discrete self-contained chunks. Nielsen Norman Group's long-running research on how people read online is blunt about it: they scan, they do not read.
Pick the number honestly
Decide the number after you have the items, not before.
Writing "10 ways" and then hunting for a tenth is how filler gets in. If you have 7 genuinely good items, write 7. A tight 7 beats a padded 10 every time, and readers can feel the difference by item 8.
A few practical notes on numbers:
- Odd numbers are not magic. Ignore that advice.
- Bigger numbers set a bigger promise. A "37 tips" post has to actually deliver 37 useful tips.
- If your number is under 5, ask whether it needs to be a list at all. 3 points is usually just an article.
Order the items on purpose
Default to most useful first.
The temptation is to build up to the best item so people read the whole thing. Do not. Most readers will not reach the end, so putting the best material there means most readers never see it. Lead with your strongest item and let the tail be the tail.
2 exceptions:
- Process lists. If the items are steps, order them by sequence. "Step 4" before "step 2" is nonsense.
- Difficulty lists. If items build on each other, order them easy to hard.
Give every item the same shape
Each entry should answer 3 things:
- What it is. One sentence, plain.
- Why it matters. The reason someone would care.
- What to do. The concrete action, tool, or example.
Keeping the shape consistent makes the post scannable. A reader learns the pattern at item one and can then skim at speed. It also stops you from writing four paragraphs on your favourite item and one line on the rest.
Write headings that stand alone
Bad: 3. Consistency
Good: 3. Publish on a schedule you can hold when you are busy
The second one is useful even if the reader never opens the section. It also gives search engines something specific to match against a query, and gives answer engines something quotable.
Treat every heading as a mini-headline. If your table of contents reads like a list of one-word nouns, rewrite it.
What changed now that AI answers questions
Two real shifts, and neither one kills the format.
Thin lists lost their advantage. A post that lists nine tools with one sentence each is exactly what an AI summary produces instantly and for free. There is no reason for a person to click that. If your list contains only what a language model already knows, you are competing with the free version.
Specific, first-hand detail became the moat. What an AI cannot generate is what happened when you used the tool. The pricing that surprised you. The step the documentation leaves out. The one that stopped working last year. That is now the whole value of a list post.
So the practical rule: every item should contain at least one thing you could only know by doing it. Google's helpful content guidance asks the same question from the other direction, listing what first-hand expertise looks like on a page.
A worked example
Say you are writing "6 ways to get more people to read your blog."
Weak version: 1. Share on social media. Sharing your posts on social media helps more people see them.
That is true, useless, and free everywhere.
Strong version: 1. On feed platforms, post the thing itself, not a link to the thing. Feed networks want to keep people on the network, so a post whose whole payload is an outbound link is competing with the platform's own interests. On LinkedIn or X, rewrite the core point so it stands on its own and put the link in a reply or the first comment.
This one genuinely depends on the platform, which is worth saying out loud. On Reddit, a link post is the native format and rewriting it as a text post reads as evasive. On Pinterest the link is the entire mechanism. In a newsletter, nobody is throttling anything. The underlying test travels even when the tactic does not: would someone who never clicks still have got something?
Same item. One of them is worth clicking.
Common mistakes
- A long warm-up before item one. Readers came for the list. Two or three sentences of context is plenty.
- Uneven items. 500 words on item 1, 40 words on item 6.
- Overlapping items. If two items say the same thing in different words, merge them. That is how a real seven becomes a fake ten.
- No conclusion at all. A one-line close that says what to do first is enough, but say something.
- A number in the title that does not match the post. It happens more than you would think, usually after an edit.
A checklist before you publish
- Does every heading make sense on its own?
- Could any two items merge?
- Is the best item near the top?
- Does each item contain something specific that is not obvious?
- Does the number in the title match the number of items?
- Would you send this to someone who asked?
Frequently asked
- What is a listicle?
- A listicle is an article structured as a numbered list: "7 ways to…", "12 tools for…". Each item stands alone, which makes the post easy to scan and easy to quote.
- How many items should a listicle have?
- As many as you have genuinely good ones. Decide the number after gathering the items, not before, so you never pad the list to hit a headline figure.
- Do listicles still work for SEO?
- Yes, when the items contain specific first-hand detail. Thin lists that only restate common knowledge now compete directly with free AI summaries and lose.
- Should the best item go first or last?
- First, in most cases. Most readers do not reach the end, so saving your strongest item for last means most of your audience never sees it.
The list format is not the problem. Thin writing is.
If you write seven items that each contain something a reader could not have guessed, you have a post worth publishing. If you write ten that could have been generated in a second, the number will not save it.