Agency, freelancer or AI: who should make your SaaS promo video?

I've freelanced for years, worked alongside agencies on other people's projects, and now I build a tool that does the job with AI. So I've watched each one hurt from close up.

Most people pick by price. That's the easy variable and the least useful one. A SaaS promo video costs you calendar time and attention as well as money, and those two are what actually decide whether the thing ever gets finished.

Agencies: you're paying for someone who will say no

The real product of a good agency is judgment, not footage. Somebody senior sits and listens to you describe your product for as long as it takes, then tells you the thing you're most excited about isn't the thing your customer cares about. Almost nobody sells that separately, and it's worth a lot.

You also get a machine. Creative direction, a writer, designers, sound, a producer keeping it all moving. When the video really has to be good, a launch, a funding announcement, a homepage hero that'll sit there for two years, the machine earns its keep.

The costs are money and layers. Agency work usually starts at several thousand and climbs from there, depending on scope and where they're based. The layers are the part people underestimate. Your feedback goes to a producer, who translates it for a creative director, who briefs the editor. Days later you see a cut where your note was interpreted rather than executed. You say "not quite" and the loop runs again.

Calendar time is the real bill. Kickoff, script, script revisions, storyboard, approval, production, first cut, two rounds of notes. Every one of those steps has a gap in it where somebody is waiting on somebody else. On the agency projects I've been near, six weeks was the normal, well-run timeline. Not the disaster case. The good case. Nobody was being lazy; that's just how many handoffs there are.

Agencies suit you if the video matters more than the schedule and you genuinely want to be argued with. They suit you badly if you're still changing your positioning every other week. You'll pay for a video that describes a product you no longer sell.

Freelancers: cheaper, direct, and now you're the producer

Working with a good freelancer is my favourite way to make a video. You talk to the person doing the work, and your note goes straight into their timeline with nobody in between to soften it.

Rates vary a lot, but for a proper promo piece you're usually looking at somewhere in the high hundreds to low thousands, depending on length, animation load and how established they are. Faster than an agency too. Often a week or two.

Here's the trade you might not have priced in: the producer job doesn't disappear, it moves to you. Somebody has to write the brief, collect the screen recordings, say the music has the wrong energy, chase the revision, and notice the pricing on screen is out of date. At an agency that's a producer's job. With a freelancer it's yours, and it'll eat your week in the awkward way, twenty minutes at a time, always on the day you had something else planned.

Quality varies more here than anywhere else I've looked, so it's worth knowing how to judge somebody.

Ignore the showreel. Showreels are highlight compilations cut to a drum track, edited so fast that no shot has to survive a second look. Everyone's looks good. Ask instead for two or three complete pieces in the format you need, full length, start to finish. If you want a 45-second product video, a beautiful 20-second sneaker spot tells you almost nothing. Screen recordings, UI motion and explaining an abstract feature clearly are a specific skill, and plenty of brilliant editors are bad at it.

Then watch those full pieces with the sound off. If the writing still holds up silent, you've found someone who can write. After that there's really only one question: could you say what the product does after a single watch? That one fails constantly, including in work that looks gorgeous. I've done it myself. I once spent most of a project on an opening sequence I was very pleased with and left the part that actually explained the product to a rushed final third. It looked great. Nobody who watched it could tell you what the software did.

The brief matters more than the budget. Nearly every project I've seen go sideways started with a brief nobody wanted to write. Before you approach anyone, put on one page: who the video is for, the one thing they should walk away understanding, where it'll live, how long it runs, what has to appear on screen, and a couple of videos you like with a sentence on why. If you can't write that page, you're not ready to hire anyone. And by the third revision round you're usually not fixing the edit. You're fixing a script decision everybody nodded through in week one.

AI: fast and cheap, but your SaaS promo video still needs an opinion

This is what I build now, so read this section with the appropriate squint.

No scheduling, no gap between brief and quote, no waiting for a slot in somebody's week. PromoHyper takes your product URL, reads your brand, writes and voices a script, scores the music and designs every frame. From $29, usually ready in minutes, in 16:9, 9:16 or 1:1. If you want more than one, packs run $69, then $147 for three and $290 for ten.

The price is the real change. At this cost a video stops being a project. It's something you try on a Tuesday, look at, and redo differently. Almost nobody commissions an agency piece for one feature launch or a single ad variant, because the cost per attempt makes experimenting stupid. Here you make five and find out which one you can stand to look at.

Two caveats.

You still need an opinion. The tool can write a script, but it can't know your real differentiator is the onboarding rather than the integrations. The way I use it: treat the first output as a proposal and argue with it. Accept the first draft and you get something competent and forgettable.

And check whether it actually read your brand or just poured your text into a layout. Easy test: look at the result and ask whether it could have been made for a competitor by swapping the logo. If it could, it's a template. Everyone in this category says they don't use templates, so judge the output rather than the landing page.

There's one thing AI won't do. It won't tell you your positioning is wrong. A good creative director will, at length, and that conversation is sometimes worth more than the video.

So which one

Pick an agency if this video will be seen more than anything else you own, you need someone to push back on your thinking, and the schedule doesn't frighten you. Also pick an agency if internal politics mean the video needs an outside authority to survive committee review. That's a real reason and nobody says it out loud.

Pick a freelancer if you can write the one-page brief and will actually do the producer work. When it works it's the best quality per dollar I know of, as long as you hire on full-length pieces in your format rather than a showreel.

Pick AI if you need it this week, or if the honest alternative is making nothing at all.

The most common way to waste money here is paying agency prices for a video nobody ever pushed back on. If you're buying judgment, make them use it on you, including in the meeting where it gets uncomfortable. That argument is the thing you're paying for, and it's the part you can't get anywhere else.