The quiet threat hiding in your product photography backlog
Cross-border sellers have spent a decade optimizing the wrong end of the funnel. We obsess over CPC, ACoS, and conversion rate, then hand our creative pipeline a folder of flat-lay JPEGs and hope for the best. Meanwhile, the platforms that actually move units — TikTok Shop, Temu, SHEIN — are rewarding interactivity, not static catalogs. So when I saw Liinks launch diiverge.co on Product Hunt, my first thought wasn’t “cute toy.” It was: this is the creative primitive Amazon sellers will be forced to copy within eighteen months.
What diiverge actually does — and why the mechanism matters more than the demo
The pitch from maker Charlie Clark is deliberately small: “Start with a picture. See where it goes.” You click a glowing dot in a scene, choose what happens, and the next moment unfolds as a short film. You can go back, try another branch, or open a map to see how the story forks. Public worlds are free to explore; the studio lets you turn your own photo, painting, or screenshot into an adventure using paid scene packs, then share the link.
That’s the whole product. No dashboard, no analytics suite, no API. Which is exactly why it’s worth studying.
The underlying mechanic — a branching narrative graph where every node is a generated short clip, and every node persists for the next visitor — is a content structure. Content structures are what cross-border operators actually buy. You don’t buy “an AI video tool.” You buy a way to turn one SKU photo into forty pieces of platform-native creative without hiring forty editors.
Why Amazon sellers should care more than Shopify ones
Shopify merchants already live in a world of flexible PDPs, embedded video, and Klaviyo flows that can trigger on almost any behavior. Adding a branching story to a DTC landing page is a weekend project for a decent Liquid developer.
Amazon sellers cannot do that. Your A+ Content modules are templated, your Brand Story slots are fixed, and your video uploads are capped. The only lever you have is the main image stack and the short video on the listing — and every competitor has the same levers.
That asymmetry is the opportunity. If diiverge-style branching content becomes a standard ad format on Meta or TikTok — and the engagement numbers on interactive video suggest it will — the sellers who can produce it cheaply will outbid the ones who can’t. Amazon sellers have the most to gain precisely because they have the least native flexibility inside the marketplace itself.
How it differs from the tools you’re already paying for
Let’s be blunt about the incumbent stack.
If you want AI video today, you’re probably looking at Runway, Pika, or Sora. All three are generation engines. You prompt, you get a clip, you move on. There’s no persistent structure, no graph, no “every path stays available for the next visitor.”
If you want interactive content, you’re looking at Verse or Storyline — authoring tools built for L&D teams, not for a seller trying to ship a TikTok hook by Friday.
If you want UGC-style video at scale, you’re looking at Creatify or Arcads. Great for talking-head ads. Useless for anything that requires a viewer to make a choice.
diiverge sits in a gap none of these occupy: a branching graph where each node is generated media, and the graph itself is the shareable asset. The closest analogue isn’t a video tool at all — it’s Twine, the open-source branching-narrative engine that indie game developers have used for fifteen years. Twine has no AI generation. diiverge has no authoring depth. Somewhere between them is the product a cross-border seller actually needs.
Where the math breaks
Here’s the part the Product Hunt thread glosses over. Paid scene packs mean every branch costs money. A simple three-choice, three-level story has 27 terminal nodes. At even modest per-scene pricing, you’re looking at real spend before you know whether the format converts.
Compare that to a Helium 10 subscription or a Jungle Scout seat — fixed monthly cost, unlimited queries. The mental model is completely different. Creative tools that meter generation punish experimentation, and experimentation is the entire job.
I’d want to see a flat-rate tier before I’d recommend this to a seller running Amazon PPC at any real scale.
What cross-border sellers can borrow from this — regardless of whether they adopt it
Three transferable ideas, in order of how fast you can ship them.
One: treat your product photos as level-one nodes, not final assets. The maker’s framing — “start with a picture” — is the right mental model for a creative pipeline. Your hero image isn’t the end of the workflow. It’s the seed. Every lifestyle shot, every UGC clip, every comparison graphic should be traceable back to a single source image and a single branch decision. If your current DAM can’t answer “which asset came from which source,” you have a structural problem, not a volume problem.
Two: persistence beats novelty. The comment from Rabnoor Singh about showing “how many branches nobody has opened yet” is the single best product idea in the thread. An unexplored count is a stronger click reason than a play button. Translate that to e-commerce: “1,247 shoppers chose the blue variant” outperforms “Shop now” almost every time. Social proof as a navigation mechanic, not a badge.
Three: let the audience build the map. Sagar Deore asked whether people mostly make their own worlds or explore others’. That’s the same question every DTC brand should ask about UGC. The answer, historically, is that 1% create and 99% consume — which means your job is to make the 99%’s exploration feel like progress. Wishlists, saved carts, and “recently viewed” are all primitive versions of this. None of them are as sticky as a branching path you can see.
The memory question nobody answered
Gal Dayan asked the sharpest technical question in the thread: does the system remember earlier choices within a playthrough, or does each scene generate independently? The maker didn’t answer in the scrape.
This matters enormously for commercial use. A branching ad that remembers the viewer’s first choice can personalize the offer. A branching ad that doesn’t is just a fancy slideshow. Until that’s clarified, treat diiverge as a prototyping tool, not a production system.
Where my judgment says it falls short
I’ll be direct, because the sellers reading this don’t have time for hedging.
No commerce integration. There’s no mention of Shopify, no Amazon Attribution, no TikTok pixel, no way to tie a branch choice to a purchase event. For a consumer toy, fine. For a seller, that’s the whole point. A branching story that can’t tell you which path drove the add-to-cart is a cost center.
No analytics surface. The maker describes exploring, going back, opening the map. Nothing about completion rates, drop-off nodes, or branch popularity. Every serious creative tool in 2025 ships with at least basic funnel visibility. Triple Whale built an entire company on the premise that sellers will pay for attribution. diiverge currently offers none.
Paid scene packs create the wrong incentive. Covered above, but worth repeating: metered generation is the enemy of iteration. The winners in this category — Canva, CapCut — won by making the marginal cost of a draft effectively zero.
No team or brand controls. No mention of shared workspaces, brand kits, or approval flows. Cross-border teams run across time zones. A tool that assumes a single creator is a tool that gets abandoned the moment you hire a second person.
None of these are unfixable. All of them are the difference between a Product Hunt darling and a line item in a seller’s stack.
What I’d watch / test next
This week, before you spend a dollar on diiverge or anything like it, do three things.
First, audit your current creative pipeline. Pick your top five SKUs and trace every asset back to its source image. If you can’t, you’ve just found your bottleneck — and it isn’t generation speed.
Second, run a manual branching test. Take one product photo, write three possible next scenes by hand, and ship them as a three-frame Meta carousel with a different CTA on each frame. You’ll learn more about interactive intent from a $50 test than from any Product Hunt demo.
Third, watch whether diiverge ships a flat-rate tier, a Shopify app, or an analytics dashboard in the next ninety days. Any one of those signals a pivot toward commercial use. All three together means it’s time to take the category seriously.
The format is coming to cross-border commerce whether we adopt this particular tool or not. The only question is whether you learn the grammar now, while it’s cheap, or later, when every competitor already speaks it.
