← All work
2022-09 — 2023-06·marketing

Xiaohongshu playbook

A codifiable 3-second rule (cover + headline + opener) distilled from multiple 10k+ like hits, and a visual template library the team could ship from.

XiaohongshuContent operationsPlaybookVisual templatesDistribution modelA/B testing
01

The observation

On Xiaohongshu, the first three seconds decide everything.

I ran multiple Xiaohongshu accounts over a year. I watched the same accounts post the same kind of content, with wildly different results, and I could not point to a single reason. The team was solid. The topics were solid. The visual was solid.

What I noticed, over many hundreds of posts, was a pattern in the distribution model. The first three seconds of the cover-and-headline pair — the part the algorithm uses to decide whether to push the post into the first batch — was almost always where the difference lived.

A great post with a weak opener dies in the first 100 impressions. A weak post with a strong opener gets a second life.
02

The 3-second rule

Cover + headline + opener. Codified.

I codified the rule into a checklist, ran it against every post for the next quarter, and the hit rate climbed. I am not claiming causation — distribution models are noisy — but the team that followed the rule shipped more 10k+ posts than the team that did not.

01

Cover

The 1-second visual contract

Typeface hierarchy, color temperature, and a single visual hook. The cover has to read in a phone-sized thumbnail — which means typography matters more than illustration, and contrast matters more than decoration.

02

Headline

The 2-second promise

A short, specific claim. Not "How to do X" — "I tried X for 30 days, here is what happened". The headline either earns the click or it does not. Generic headlines earn nothing.

03

Opener

The 3-second hook

The first line of the body. Either a tension, a confession, or a number. Never a greeting, never a "let me tell you", never a recap of the headline. The opener has to do a job the headline did not.

03

The template library

A playbook is only useful if a non-expert can ship from it.

The rule is not enough. A rule is just a constraint. What makes a constraint usable is a template — a visual + headline + opener structure that a junior operator can copy, modify, and ship in under an hour.

I worked with the design team to ship a library of around 30 templates, organized by topic, by post type, and by emotional register (curious, urgent, contrarian, intimate). Each template came with two worked examples — one that hit, one that did not, with the postmortem on why.

The library was the difference between a playbook and a habit. The team stopped reinventing the cover every post. They started picking a template, customizing the headline, and shipping.

04

The outcome

A codifiable rule, a usable library, a team that ships.

The 3-second rule is now an internal standard, taught in onboarding. The template library is in its third iteration. The team ships more posts per week than they did a year ago, with a higher hit rate, and a faster feedback loop.

More importantly — the rule has survived being taught. New operators, picking it up cold, can ship posts that hit without having to internalize the entire distribution model first. That is the test of a good system: it works whether or not the person who built it is in the room.