A year ago, the hard part of video for a B2B marketing team was making enough of it. That is changing fast. AI tools can now produce a narrated, animated, captioned video from a paragraph of text. Volume is no longer the bottleneck it was.
So a new problem has taken its place. Scroll through LinkedIn on any weekday and you will see it. The same smooth camera moves. The same stock-style b-roll. The same upbeat voice, the same text that slides in from the left, the same gradient behind it. Different companies, one look. If you covered the logo, you could not say whose video it was.
The new sameness
This matters more than it seems. In a recent post I argued that mid-market software companies lose to louder competitors, not better ones, and that steady presence is the fix. But presence only works if buyers know it was you. A video they enjoy and then credit to your competitor is worse than no video at all.
More output with no brand in it is just noise. It fills the feed and leaves no memory behind. For a company trying to earn a place on the buyer’s day-one shortlist, that is the wrong trade.
Why do AI videos look generic?
There are three reasons, and none of them is that the tools are bad.
- Defaults are shared. Every team using the same tool starts from the same templates, transitions, voices and music. Unless someone changes all of them, every time, the result converges.
- Models learn the average. A model trained on huge amounts of video is very good at producing what video usually looks like. “Usually” is the opposite of distinctive.
- Prompts do not remember. You can describe your brand in a prompt, but the next video starts from zero. The brand lives in whoever happens to write the prompt that day.
Some tools answer this by asking marketers to become directors: write better prompts, describe each camera move, adjust each scene. That can work, but it moves the craft onto the people with the least time for it. A VP Marketing should not have to direct every video to keep the brand intact.
Distinctive brand assets apply to motion too
Marketing science has a useful frame for this. In Building Distinctive Brand Assets, Jenni Romaniuk of the Ehrenberg-Bass Institute shows how brands grow by building assets that buyers link to them: colours, logos, shapes, characters, sounds, taglines. A good asset passes two tests. Buyers link it to your brand, and they do not link it to anyone else.
Most B2B brands already apply this to their logo and colours. Few apply it to how their videos move. Yet motion is one of the first things a viewer takes in. The pace of a cut, the way a headline arrives, how the product UI is framed, the voice that reads the line: all of these can become assets, or all of them can be borrowed from a template.
The test is simple. Mute a video, hide the logo, and show it to a customer. Do they know it is yours? If not, the video is spending your budget on someone else’s category.
What a motion profile should capture
The way out is not better prompts. It is a model of your brand that every video starts from, so the brand is the default rather than the exception. We call this a motion profile. In our view it should capture:
- The basics: logo, colours, fonts and how they are combined on screen.
- Pacing: how long a frame holds, how fast scenes change, where the pauses fall.
- Movement: how text and shapes enter and leave, which transitions you use and which you never use.
- Product framing: how your UI is shown, how close the crop is, and how the eye is guided across it.
- Voice: the narrator’s tone, the words you use and the words you avoid.
The source material for all of this usually exists already: your brand book, your website and the two or three videos everyone in the company agrees are the best you have made. The job is to learn from them once and apply it every time.
Formats your buyers learn to recognise
A motion profile keeps every video on brand. Formats make them recognisable at a glance. Think of the way a news programme always opens the same way, or how a good podcast has a fixed structure. Repetition builds memory.
For a software company, that means a few signature looks, one for each kind of video:
- a launch format for new features, the same shape every time
- a customer story format with a consistent arc
- an explainer format for terms and ideas your category argues about
After a few weeks, buyers start to know what a launch from you looks like before the name appears. That is when volume starts to build a brand instead of diluting it.
When generic AI video is fine
To be fair, generic video has its place:
- internal training and updates that never leave the company
- quick tests of a message before you invest in it
- rough drafts that a designer will rework anyway
The line is simple. If a buyer will see it, it should look like you. If only your team will, speed can win.
This is the approach we are taking at Saltfish: learn a brand once, then make every video in it. See what we are building, or book a call and bring the video your team is proudest of.
