WeChat's short video surface — the one Chinese users call 视频号 — is not a separate application. It sits inside WeChat, next to the conversations and the official accounts, and that position is the whole of its interest for a company. A video there can be forwarded into a group chat, linked from an article, attached to a shop page and opened from a mini-program without the viewer leaving the application they are already in. Nothing about the format is remarkable; what is remarkable is that it shares an identity, a payment method and a contact list with everything else you have built.
What it adds to an official account that already works
An official account is read by people who chose to follow it, and an article reaches roughly the people who are already there. That is its strength — a known audience, reached directly — and also its ceiling: growth happens one deliberate follow at a time. Video on this surface behaves differently, because it circulates. It is recommended to people who follow nobody you know, and it is passed between contacts in a way an article rarely is. The pairing is the point: video widens the top, the account holds what the video brought.
The second thing it adds is a face. A B2B official account is usually anonymous prose, competent and faceless. A video shows who answers the phone, what the workshop looks like, how the person speaking holds the product — and for a foreign company in China, where the first question is whether anyone is really there, that is not decoration. It is the cheapest proof of existence available, and it is the kind a reader cannot get from a page of text however well written.
A different logic from Douyin
Douyin is a discovery machine. The feed is assembled for a stranger by a recommendation system that cares about whether the next second is watched, and a video either earns its audience in the opening moment or does not get one. Social connection plays almost no part. That rewards a particular craft: a hook that works on someone who has never heard of you, a pace that never settles, and an edit that assumes the viewer is one flick away from leaving.
Inside WeChat, recommendation exists but it is not alone: what your contacts have watched and liked weighs on what you are shown, and forwarding into a conversation or a group is a normal act rather than an exception. A video therefore travels along relationships, which changes what makes a good one. It can open more slowly, because the person watching often arrived through someone they know. It can be longer and more explanatory, because it is being watched by someone with a reason to watch. And it is judged partly on whether it is worth sending to a colleague — which is a very different test from whether it survives the first second.
Video, shop and mini-program: the path after the view
The reason this surface matters commercially is that the view does not have to end in a view. A video can carry a link to a mini-program, a shop page, a product, a form or a live session, and the viewer crosses into it without an account to create, a password to remember or a payment method to enter. For a B2B company the destination is rarely a purchase — it is a catalogue, a technical sheet, a sample request or a conversation with a salesperson — but the mechanism is the same, and the absence of friction is the thing being bought.
- One destination per video. A video that offers a catalogue, a sample request and a newsletter offers nothing. Decide what the next step is before filming, and build the closing moment around it.
- The landing page is a mini-program page, not your website. Sending someone out to an external site breaks the one advantage you had, and that is before the question of whether the site loads quickly from where they are.
- The account and the video surface are the same identity. Linked properly, a viewer who liked a video finds the articles, and a reader of the articles sees the videos. Two unconnected presences do half the work each.
- A live session is the same surface, not a new project. It uses the audience the videos assembled, it can open the same mini-program, and for B2B it works best as a technical session answering real questions rather than as a selling event.
- Measure the step, not the view. How many people opened the destination and what they did there is the only figure that tells you whether to film another one. View counts tell you about the platform, not about your business.
A rhythm two people can hold
The reason most company video efforts stop is not quality, it is cadence. A first video is made carefully, takes a week of everyone's attention, and sets a standard the team cannot repeat; the second one is late, the third never happens, and the account sits there with three videos and a date that is visibly old. A small team should design the rhythm first and let the rhythm decide the ambition.
- Choose a frequency you could hold in a bad month, not a good one. A video every two weeks, published on the same day, beats four in a burst and then nothing.
- Fix two or three recurring formats and stop inventing. A workshop sequence, an answer to a customer question, a product detail filmed close up: once the format is settled, the work is filming rather than deciding.
- Film several at once, publish them one by one. A single half-day in the workshop with the right light produces several weeks of publication, and the cost per video falls to almost nothing.
- Keep the production deliberately plain. A phone, a stable support, usable light and clean sound are enough on this surface, and over-produced video reads as advertising — the one thing you are trying not to be.
- Write the subtitles every time. They are read with the sound off, they carry the terms your audience searches for, and they are the difference between a video that is understood and one that is scrolled past in a noisy office.
One more thing belongs in the rhythm: answering. Comments and the messages that follow a video arrive in the same place as everything else in WeChat, and a question left unanswered for days is more damaging than a week without publishing. Budget the answering time alongside the filming time, or publish less often.
Republishing a video made for another network
The temptation is obvious and the result is recognisable. A video cut for a European network arrives with the wrong aspect ratio, subtitles in the wrong language burned into the image, a watermark from the platform it was exported from, music nobody here recognises, and a call to action pointing at a site that may not load. Beyond the mechanics, the editing itself is wrong: it was built for an audience arriving cold, so it shouts in the first second at a viewer who arrived through a colleague. The honest version of reuse is to keep the footage and redo everything else — reframe vertically, rewrite the subtitles in Chinese, replace the opening, change the destination. Reusing the raw material is sensible; reusing the finished video is visible, and what it makes visible is that this market was handled with leftovers.
Before filming anything, decide which page the viewer lands on and make sure that page exists and works. Most of the value of this surface is the short distance between a video and a mini-program page, and a video that ends by asking people to search for you has given that away.