How to Make an AI UGC Ad for a SaaS or Mobile App

If you sell a SaaS product or a mobile app, there's nothing to photograph — no box, no bottle, no packaging. On Flovaly the cleanest way to make an AI UGC ad for software is Talking Actors' script-only mode: a saved character delivers your pitch to camera, no product image required anywhere in that flow. You can also feed a screenshot into the guided UGC flow's product-photo field — it'll accept it — but that flow was built to have an actor hold a physical object, so the result is a real trade-off worth understanding before you upload anything.

What actually happens if you upload a screenshot

Checked directly against the generation code: the guided UGC route's productImageKeyfield has no content validation — it accepts any uploaded image, screenshot included. But the worker that composites the shot doesn't just place the image somewhere on screen. It sends your character's reference photo and your product image to an AI image-editing model with an instruction that the actor is “holding and presenting the exact product shown in the second reference image” and to keep it “unaltered — same shape, colors, label and branding.” That prompt is written for a physical product photographed against a background — a bottle, a box, a gadget. A flat app screenshot doesn't have a shape to hold, so you're asking the model to improvise something an actor can grip out of a rectangle of UI.

The two honest options, not a workaround

This isn't a case where a clever trick gets you a native screen-recording ad — Flovaly has no B-roll overlay feature and no way to composite live app footage behind an actor at all, so anything claiming otherwise would be overclaiming. There are two paths that actually work with what's built:

Script-only, no image at all.Talking Actors' prompt mode takes a character and a script, full stop — verified against the same API schema used in Flovaly's general service-pitch guide, there's no product-image field in that request. A presenter states the problem your app solves, why it solves it, and where to get it, the same way a founder would pitch on camera. This is the more honest fit for software: nothing on screen pretends to be your product, because nothing needs to.

A phone mockup, if you want a product shot.If you'd rather have something visual in frame, upload a photo or clean mockup of a phone with your app's screen showing, rather than a bare screenshot. A phone is a real object with a shape an actor can plausibly hold — much closer to what that compose prompt is built for than a flat rectangle of interface. It won't look like native screen-recording B-roll, but it composites far more naturally than a raw screenshot does.

Writing the pitch itself

Software sells on the problem it removes, not the interface it shows — a stock avatar holding a phone doesn't explain what your app actually does. Lead with the specific frustration your app solves in the first line, name the app once, state the concrete outcome (time saved, money saved, problem gone), and close with exactly where to get it — download link, free trial, sign-up page. Our guide to UGC ad scripts that convert covers this hook-first structure in more depth; it applies here whether or not there's a product image at all.

Picking a character

A premade stock avatar is enough to test a script fast. Once a pitch is working, a custom actor built from a text prompt or an uploaded photo gives your app a consistent presenter across every variant, rather than a shared face other products might also be using. Either way, Talking Actors reads scripts from 8 curated stock voices — American or British accents only, no cloning — worth checking against your target audience before you commit to a script; see how to pick a voice for AI Talking Actors for the full list.

What it costs

Script-only Talking Actors and the guided UGC flow are priced differently:

Talking Actors is 5 base credits plus 2 for the voiceover, 1.5× at 1080p. The guided UGC flow has no separate voiceover charge since speaking is embedded in the script directly. On the $29/month Starter plan (150 credits, ≈$0.1933/credit), a 5-second 720p clip runs about $1.35 either way. See the full AI video ad cost breakdown for every other generation type.

The short version: for pure software with nothing to hold, skip the product-photo field entirely and let a Talking Actor pitch it straight to camera — it's the mode actually built for that. Reach for the guided UGC flow only with a phone mockup in hand, not a bare screenshot. Try either with the $1 trial. If what you're actually selling is a physical product rather than software, our guide to making a product demo ad on Flovaly covers the guided UGC flow's other, product-photo-first path in more depth.

FAQ

Can I upload an app screenshot as the product photo on Flovaly?

Technically yes — the field accepts any image with no content check. But the compose step behind it is built to have the actor hold a physical object, so a flat screenshot doesn't composite the way a bottle or box shot would. A phone mockup showing your app's screen is a closer fit if you want something visual in frame.

Does Flovaly overlay real app footage or screen recordings behind an actor?

No. There's no B-roll compositing feature in Flovaly at all — every clip is either a Talking Actor delivering a script to camera, or the guided UGC flow's actor-holding-product composite. Neither overlays a screen recording as background footage.

What's the best way to advertise a SaaS product with no photo?

Talking Actors' prompt mode — a saved character and a script, no image field in that request at all. It's the same mode covered in Flovaly's general service-business pitch guide, applied here to software specifically.

How much does a SaaS pitch video cost on Flovaly?

Talking Actors runs 7 credits for a 5-second 720p clip (5 base plus 2 for the voiceover), 12 for 10 seconds, or 17 for 15 seconds — roughly $1.35 to $3.29 on the $29 Starter plan. 1080p applies a 1.5× multiplier.

Can I clone my founder's or team member's voice for the pitch?

No. Talking Actors reads scripts from 8 curated stock voices, American or British accents only — there's no voice cloning on any Flovaly plan, so a script-only pitch always uses a stock voice, not a specific person's real one.

FAQ