Every agency deck says video. Video is engaging, video builds trust, video is the future — and video also happens to be the thing agencies charge more to produce, which may or may not be a coincidence.
Both work. They work at different jobs, and the honest answer is less flattering to whoever wants to sell you a production package.
What each one is actually good at
Images are fast, cheap and iterate well. You can make eight variations in an afternoon. That matters more than it sounds, because Facebook ads are won by testing volume more than by any single brilliant asset. Eight decent images that you can swap weekly will usually beat one expensive video you are now emotionally committed to running until it dies.
They also load instantly. On a phone, on a bad connection, in a scroll that's moving fast, an image has already communicated before a video has finished buffering.
Video is better at trust and at explaining. If the thing you sell needs believing — a medical treatment, a large job in someone's home, a price that makes people pause — a human face doing the explaining outperforms text on a photo. It is very hard to fake competence on camera for thirty seconds, and viewers know that.
Images sell things people already understand. Video sells things people need to be convinced about.
The rough rule
Lean image when the offer is simple and the decision is quick. Drain unblocked. AC serviced before summer. A price and a phone number. Nobody needs a film to understand a blocked drain.
Lean video when the ticket is high, the decision is considered, or trust is the barrier. Weight loss programs. Cosmetic work. A roof. Anything where the honest customer objection is "is this person going to rip me off", because that objection does not get answered by a stock photo.
Lean video, harder, when you are unknown. A new business with no reviews has a credibility gap that a face closes faster than any amount of copy.
The question that actually decides it
Not "which converts better". This one:
Creative is a consumable. It gets used up — a local audience is finite and they've all seen your best ad by week six. The format you can keep feeding is worth more than the format that tests marginally better and then stops because filming became a whole thing you keep putting off.
An owner who will happily talk to a phone camera for ninety seconds between jobs should be making video. An owner who freezes on camera and hates every second should not — they should be shooting photos of finished work, which they'll actually keep doing.
The best format is the sustainable one. This is boring advice and it is correct.
What actually makes either one work
Format matters less than these, in roughly this order:
Show the real thing. Your van, your crew, your finished work, your actual clinic. Stock photography is instantly recognizable and it quietly signals that there was nothing real to photograph.
Make it stop the scroll in the first second. Not with a logo. With the problem, the result, or a face. Video especially — most people decide within about a second, and an intro animation spends that entire budget on nothing.
Assume no sound. Most feed video is watched muted. If your message lives in the audio, most of your audience never received it. Caption everything.
Say the offer. A surprising number of ads are beautiful and never mention what is being sold or what to do next.
The approach that usually wins
Run both. Not as a philosophical position — because they answer different objections and you don't know which objection is blocking a given person.
In practice that means a handful of images doing the volume work and being swapped often, plus one or two video pieces carrying the trust load. Then watch cost per booked job by format, not cost per click, because video frequently produces fewer, better leads and looks worse on every metric except the one that pays you.
We've had accounts where video cost noticeably more per lead and still won, because the leads showed up to appointments and the cheap ones didn't.
One thing that is not worth your money
Highly produced brand video. The cinematic sixty-second piece with drone footage and a licensed track.
It looks wonderful, it costs thousands, and in a feed it performs about as well as a phone video of you explaining what you do — often worse, because it reads as an advertisement immediately, and the feed is very good at ignoring advertisements.
Save it for your website, where someone has already chosen to be interested. In the feed, spend the money on making twelve more ordinary things instead.