Do you really need a video agency for a SaaS promo?

I've spent fifteen years making these videos. Motion design, promo films, product launches, for teams at Microsoft, xAI and Cloudflare, and for a long tail of startups you've never heard of. Most of that time I've been inside the process rather than running it, which is a useful seat: you see what the work costs and what the client thinks they're paying for. So when I say hiring a video agency is sometimes a waste of your money, I'm arguing against my own invoice.

I'm going to defend agencies too, because the takes I keep seeing are lazy. Usually some version of: agencies are overpriced middlemen, just use a tool. That's cheaper to say than the honest answer, which is that it depends entirely on what you're buying.

What you genuinely get from a good agency

Someone who says no. This is the most valuable thing on the list and nobody prices it. Say you've got eleven features you love. A creative director's job is to tell you the video covers three, and that the joke on slide four doesn't land. Most founders I've worked with are too close to the product to make that cut, and internal teams have a hard time saying no to the person signing their paycheck. Someone in the room with taste and no career risk is rare.

Art direction that gets invented. Somebody sits down and builds a visual language that belongs to your product and nothing else: a way of moving, a particular relationship between type and space that stays recognisably yours across a whole campaign. That takes exploration, dead ends, and work that gets binned. You're paying for the binned work too, and you should be.

Live action. The moment you need humans, a location, a camera, sound, you need people. There's no software substitute for someone who knows how to light a face or get a natural performance out of a founder who has never been on camera.

Something you keep. Good studios don't just hand you an MP4. You get title treatments, transitions, a colour and type spec, project files your next editor can actually open. If you're making forty videos over three years, that groundwork pays back. If you're making one, it doesn't.

What else is on the invoice

When you get a quote in the thousands, a real chunk of it isn't creative work, and it took me an embarrassingly long time to understand that properly.

There's project management: the producer who schedules, chases, invoices and turns your Slack messages into a brief. There are revision cycles, priced in because they're inevitable. In my experience, three rounds tends to produce a request for a fourth. There's calendar risk: a studio holds time for you, and if you go quiet for two weeks, as I understand the maths that time is dead but already paid for. There's overhead: rent, licences, the accountant, salaries of people who will never touch your project. And there's margin, which is fine, that's a business.

None of that is a scam. It's what it costs to run a studio that will still be there when you call next year. But you should know that when you buy a 40-second promo from an agency, you're buying a coordination service with a video attached.

The timeline follows from that. Kickoff, script, feedback, storyboard, feedback, animation, feedback, sound, delivery. Every handoff is a place where your project sits behind someone else's. In my experience the actual hands-on hours are a small fraction of the calendar time. The rest is waiting.

The test I'd apply

One question: is this video the thing, or is it about the thing?

If the video is the thing, your homepage hero or the asset you'll run paid spend against for the next two years, hire an agency. Get a good one, pay properly, and don't design it by committee.

If it's about the thing, a feature announcement, a changelog explainer, something for the pricing page, you're somewhere else entirely. That video needs to be clear, and it needs to be out this week. It does not need weeks of calendar and an invented visual language.

I got this wrong for years. I argued for the full treatment on everything, partly because the full treatment is more fun to make. Then I'd watch the film get quietly retired when the product moved on, and watch the client hesitate to replace it because of what it had cost. That's the trap: agency-quality work on a disposable asset isn't a bonus, it's a liability. You end up with a beautiful video of an interface that no longer exists.

The other useful signal is whether you already know what you want to say. If you can write the sentence ("this turns X into Y for Z people") you probably don't need a strategist. If you can't, no software is going to find it for you, and a good agency will.

What the cheaper end actually looks like

Roughly: a freelancer, a template tool, or a generative tool.

Freelancers are good, and I'd take a good freelancer over a mediocre agency any day. What you lose is the producer layer, which is exactly the point. You're the producer now. You chase and you decide. If you have fifteen spare minutes a day, be honest about whether you can do that job.

Template tools I mostly dislike, and I'll say why plainly. Templates are built to fit any product, which means they fit yours loosely: you bend your message into a slot designed for someone else's. The pacing belongs to the template, so your best line lands on a slide that's already leaving. And the good templates are recognisable. You've seen that swoosh. So has everyone you're about to send the video to.

Generative tools are where I ended up building, so treat the next bit as disclosure rather than a review. PromoHyper is mine: you paste a product URL, it reads your brand, writes and voices a script, scores the music and designs each frame, with no template slot to bend your copy into. From $29, ready in minutes, and it exports 16:9, 9:16 and 1:1, which matters more than it sounds like it should. Try this: export one frame of your dashboard shot, crop it to vertical, and look at it on your phone at arm's length. The small labels you'd never think twice about on a monitor stop being words and start being texture. Cropping a landscape promo to vertical rarely survives contact with a phone.

I built it because I kept watching founders pay several thousand dollars and wait weeks for a video the next redesign would make obsolete. Bad trade in both directions. It's rarely the job a studio wants either.

I've compared agencies, freelancers and AI head-to-head if you want the full picture.

If you do hire a video agency, hire them well

Bring a written point of view: one sentence about what this video has to accomplish and who has to believe it. A mood board of videos you liked isn't that. In my experience the quality of what comes back is mostly decided before anyone opens a project file.

Consolidate your feedback. One document, one round. Four people giving notes in three different places is how a good project turns mediocre, because the team starts averaging opinions instead of deciding. I've done it myself, and the result is beige.

Ask for the project files, and ask before you sign. Some studios hand them over and some don't, and the ones that don't have, I'd assume, made a business decision about you coming back. That's not evil, but you should know which deal you're in.

There's no shame in either choice. I've made videos that genuinely needed a room full of people: a director, a crew, a composer. I've also watched founders spend money they obviously didn't have on a film that mattered less than a good screenshot. Neither price point is the mistake by itself. The mistake is not knowing which kind of video you're making before the money starts leaving.