Gaming and metaverse trailers: the beats that sell a world

Games have spent thirty years working out how a trailer holds attention. Most web3 game and metaverse trailers ignore all of it. The six beats, and the three cuts to build from one shoot.

September 9, 2026 · 9 min read

Game trailers are one of the few marketing formats with a thirty year track record. The conventions are settled, the audience is trained on them, and a viewer knows within two seconds whether they are watching something made by people who play games. Web3 game and metaverse projects mostly ignore all of it.

What gets published instead comes in two shapes. The first is a cinematic: a rendered character walks through a canyon, a logo lands, and at no point does anything suggest a game exists behind it. The second is a feature reel: eight seconds of a map, eight of an inventory, eight of a marketplace, captions announcing that the project has land, crafting and a token.

Both get skipped, for the same underlying reason. Neither one tells the viewer what they would be doing. A trailer is not a summary of what a project contains. It is a promise about an activity, and the audience for games is unusually good at telling the difference.

Why the cinematic and the feature reel both fail

The cinematic fails because it is unfalsifiable. Nothing in it can be checked against a build, so an experienced viewer reads it as a statement about the budget rather than about the game. That reading is often correct, and it is why the comments under a pure cinematic reveal are so predictably hostile.

The feature reel fails for the opposite reason. It is all evidence and no argument. A list of systems tells someone what the project has, never why they would want an evening of it. Systems are the answer to a question nobody asked until they already cared.

Metaverse trailers tend to fail harder than game trailers, and the cause is specific: there is often genuinely nothing to do in the footage. Avatars walk through architecture. A crowd stands in a plaza. The camera does the moving because the player has nothing to. Empty space rendered beautifully still reads as empty space.

The six beats a trailer actually runs on

This is the structure the format converged on, and it holds whether the piece is twenty seconds or two minutes. The beats compress, they do not disappear. Skipping one is what produces a trailer that feels off without anyone being able to say why.

Note how much of that is ordering rather than production value. A trailer cut in this order from modest footage outperforms an expensive one that opens on a logo and ends on a partner wall. The beats are free. The footage is what costs money.

The verb is the product

Every game that markets well can be described by a verb that a person could mime at a table. You drive. You build. You trade. You survive the night. The verb is not the genre and it is not the feature list, it is the physical thing the player repeats for hours.

Finding it is the first working session, before any shot list exists. If a team cannot state the verb in one clause, the trailer is not the problem yet. Three or four candidate verbs usually surface, and choosing between them is a positioning decision that outlives the video.

Once it is chosen, the verb sets the edit. It appears in the first ten seconds, it comes back at the escalation, and it is the last thing on screen before the card. Everything that does not serve it becomes cuttable, which is exactly the discipline missing from a feature reel.

For a metaverse or a social world the verb is harder and more important. Walking is not a verb anyone chooses to spend an evening on. Hosting, performing, building, selling, competing: those are. Show one of them happening between people, and the empty plaza problem solves itself.

Three trailers, one build

A launch does not need three productions. It needs one capture effort and three edits, planned together so the same material can be cut to three different jobs. Deciding this upfront is the difference between a three week schedule and a three month one.

The gameplay cut is the one teams try to skip and the one that does the work. A held, uncut sequence is the only asset in the set that a sceptical viewer accepts as evidence. Thirty seconds of unbroken play beats two minutes of cross fades, and it is cheaper to produce.

The sequencing around these cuts follows the same logic as any other launch video schedule: the announcement runs while the build is still moving, the gameplay cut lands when there is something to defend, and the launch film goes out against a fixed date rather than a hope.

What you can show before the game is playable

Most teams commission a trailer at the worst possible moment: the build exists but it is ugly, the UI is placeholder, and the temptation is to render something instead. There is more shootable material at that stage than it looks.

• One vertical slice. A single environment finished to final quality, used for every shot. Cheaper than a finished game and indistinguishable in a trailer.
• Greybox with final interface. Untextured geometry reads as deliberate style far more often than teams expect, and the interface is what proves a build exists.
• The character doing one thing well. A single animation cycle, shot from three angles, is a character. Twelve unfinished cycles are a bug reel.
• Real capture at low settings. Honest footage from a running build, clearly labelled, outperforms a render that viewers assume is a lie.
• Say which is which. If a shot is pre rendered and not in engine, mark it. Games that hide this get found out in a day and spend the launch apologising instead of selling.

That last rule is not just ethics, it is efficiency. A trailer that labels its own footage can use both kinds freely. A trailer that blurs the line has to avoid ever cutting between them, which removes the escalation beat and flattens the whole piece.

The feed is not a storefront page

A trailer built for a store listing is watched by someone who already navigated to it. The same file dropped into a timeline is watched by nobody in particular, muted, at a third of the screen, with a thumb already moving. The version that works there is a different edit, not a different export.

That means a vertical cut with the verb inside the first three seconds, legible motion at small sizes, and no reliance on audio for the hook. It also means the card is redundant: the account name is already on screen, so the last two seconds can be spent on one more moment of play.

There is a narrow case where the feed slot has to be filled before any of this footage exists, usually a listing or a page going live ahead of the build. We would rather point at our own tool than a competitor, so to be explicit about who owns it: ShipTeaser is a product from the same founder as this studio. It takes a URL and returns a 15 second 1080p motion graphics teaser sized for muted feeds, with an optional music bed and the first video free without a card. For a game it has an obvious limit: it reads a web page, so it contains no footage of your world, no character and no verb, and there is no voice-over or timeline to edit. It fills a slot on the day. It is not a trailer.

How we build one

The first pass is not a shot list, it is the verb and the rules of the world. Those two decisions govern everything downstream, including which environments get finished first, which is why the conversation is worth having with whoever owns the build and not only with marketing.

From there the shape is the same as our Launch Sprint: the narrative spine and a visual micro system first, then the film, then the art direction for the page it lands on, then the social kit cut from the same material. A world is the easiest kind of project to build a system around, because the rules are already fictionalised and everyone on the team can repeat them.

The same principle drives an NFT collection reveal, where the world has to be established before the individual pieces mean anything. A trailer is that argument at higher intensity, with a verb attached. Recent films and identity systems are in our selected work.