Visuals for Experts and Service Businesses: How to Show What You Can't Photograph

Visuals for experts and service businesses: how to show a service, what to generate, and why real photos still matter.
Experts and service businesses face the opposite problem from product sellers: there's nothing to photograph. A consultation, a training session, ongoing support, a review call, none of that sits on a table waiting for a camera. Visuals for experts have to be built around meaning and outcome, not around an object. Here's what actually works.
Show the outcome, not the process
What actually sells a service is the change it brings to the client. Not "I offer consulting calls" but the state after: order instead of chaos, confidence instead of confusion, a system that runs instead of scattered attempts. That's something you can capture through an image, a metaphor, a scene. AI models handle this kind of abstract illustration well, because there's no specific object that needs to look accurate.
What to generate for an expert brand
A practical list to work from. Covers for posts, livestreams, breakdown videos. Illustrations for topics: metaphors and diagrams that explain an idea. Course and program design, with module covers, dividers, one consistent look. Banners for a landing page or ads. Anything that needs readable text in the frame works better with models strong at typography, like Ideogram and GPT Image 2.
Why you still need real photos of yourself
There's a catch, though. An expert sells themselves, and you can't fully hide behind generated illustrations: people buy from a person they can see. A generated avatar instead of a real face works against trust, especially in consulting niches. So a photo shoot, at least once, is still worth doing. AI covers everything else.
A consistent style matters more than any single nice image
Experts publish regularly, and mismatched styling dilutes your brand recognition over time. It's worth setting up a template once, with a color palette, a type of illustration, and a layout for text, then sticking to it. For more on locking in a style and why it still drifts, there's a piece on repeatability in AI generation.
Honestly, abstract metaphors can slide into generic stock imagery fast: handshakes, lightbulbs, images that say nothing. I try to pull the metaphor from something specific to the client rather than a generic vocabulary, though it doesn't always land. You can put together post and course visuals in Flami.
FAQ
How do I show a service that can't be photographed?
Through the outcome and the client's state afterward, not the process: order instead of chaos, confidence instead of confusion. AI handles these abstract images well, since there's no specific object that needs to look accurate.
What should an expert generate first?
Covers for posts and livestreams, metaphor illustrations for topics, course module design, and banners. For anything with readable text, use a model strong at typography so the lettering doesn't come out garbled.
Can a generated avatar replace a real photo?
Better not to. An expert sells themselves, and people buy from a person they can see. A generated face instead of a real one hurts trust, especially in consulting niches, so real photos are still worth having.
How do I avoid generic stock imagery?
Pull the metaphor from your client's specific situation, not from a generic vocabulary like handshakes or lightbulbs. Universal imagery doesn't tell your audience anything and makes the content feel faceless.
Sources
- Flami: seeds and repeatability in AI generation
- Flami: visuals for a store's social media channel
About the author
Daniel Reed
Head of content at Flami