The Unscalable Part of Cross-Border Selling Is the Conversation, Not the Catalog
Every cross-border operator I know has solved the same problem twice. First they solved logistics: the parcel gets from Shenzhen to a Midwest doorstep in nine days, tracked, duties prepaid. Then they solved creative: a UGC hook gets tested across forty ad variants before lunch. What almost nobody has solved is the middle layer — the actual conversation between a brand and a buyer who is 7,000 miles and one language away from the person who built the product. That gap is where refund requests become chargebacks, where “how do I use this?” becomes a one-star review, and where a $600 wholesale inquiry dies because nobody followed up with a real walkthrough. So when a tool like MyLens AI shows up promising one-on-one AI presentations at scale, my ears prick up — not because I need another deck builder, but because the unscalable part of this business has always been the human on the other side of the screen.
What MyLens actually is, stripped of the launch-day gloss
The pitch from co-founder Ardalan Mirshani is that AI can write and code for us, but it can’t understand for us — and the best teachers, mentors, and salespeople succeed by sitting with one person and adapting. MyLens tries to bottle that. You feed it raw materials and it generates visual slides in seconds (launch page). You “teach” it the audience, the goal, the guardrails. You send each recipient a private invitation link and an access code. The AI then presents live, answers questions from your source material, and — the detail that matters — generates new slides and visual layers mid-session when a question demands it. Afterwards you get a per-session review, with recommended follow-ups and flagged confusion points.
This is the company’s fourth Product Hunt launch; earlier iterations in November 2023 and June 2024 were timeline and visual-answer generators. That lineage tells you something: the team has been iterating on “turn messy information into a visual” for two years, and this launch is the commercial wrapper around it. Reviews on the page are thin — two reviews, one calling it a “game-changer” for turning complex data into visuals, one praising the timelines — so treat the social proof as early-stage, not battle-tested.
The problem it’s actually solving — and why it’s a sleeper for this industry
Here’s the reframe I’d push on any seller reading this. MyLens is not competing with Canva or Gamma on deck aesthetics. It’s competing with the live human who would otherwise have to run the same onboarding call forty times a week. The maker explicitly names 1-1 customer success, support, and onboarding as core use cases, and confirms it can walk customers through complicated proposals and business reports.
Now map that onto cross-border operations. You have three chronically understaffed conversation surfaces:
Supplier and wholesale onboarding. A US boutique wants to stock your line. Today that’s a PDF line sheet, a Zoom call in a bad timezone, and a follow-up email that gets ignored. An AI session that walks the buyer through MOQs, margin math, and your returns policy — answering “what’s the landed cost at 200 units?” from your actual data — is a genuinely different sales motion.
High-consideration product education. Anything with setup complexity — a 3D printer, a smart garden, a piece of commercial kitchen gear — generates a support ticket per order. A private walkthrough link in the post-purchase email could cut that.
Localization at the last mile. This is the one I’d bet on hardest. A Spanish-language or German-language AI presenter, trained on your English source material but instructed to answer in the buyer’s language, is a localization layer that doesn’t require hiring native CS staff in five markets.
Why Amazon sellers should care more than Shopify ones
Shopify DTC brands already own their customer relationship and can email, SMS, and retarget freely. Amazon FBA sellers can’t. Their buyer is behind a wall — Amazon Seller Central gives you a messaging window, not a relationship. That constraint is exactly why a private invitation link is interesting: it’s a compliant-ish way to pull a buyer into a branded, one-on-one experience outside the marketplace’s UI, where you can actually teach them how to use the product and preempt the negative review. The risk, obviously, is that Amazon’s communication policies are strict about soliciting off-platform contact, so any use here needs a compliance read before a single link goes out. I’d frame it as post-purchase education, not review farming, and keep every message inside the approved templates.
How it stacks up against what operators actually use today
The honest comparison set isn’t other AI presentation tools. It’s the stack a seller already owns:
- Klaviyo flows for post-purchase education. Cheaper, but one-to-many — no adaptation, no Q&A.
- A Zoom + Loom combo for wholesale and onboarding calls. Human, warm, and completely unscalable across timezones.
- Intercom or Zendesk AI chatbots for support deflection. They answer questions but don’t present — there’s no narrative, no visual walkthrough, no “let me show you.”
- Gamma and Tome for deck generation. They solve the artifact, not the delivery.
MyLens’ differentiation is the combination of live adaptation and per-recipient sessions with review analytics. The mid-session slide generation is the feature that separates it from a recorded Loom — a good presenter does exactly that, and most pre-built decks can’t. The post-session review layer (“who understood, who struggled, what questions came up”) is the part I think operators will underrate and later love, because it turns a support cost center into a product-feedback pipeline. If forty buyers all ask the same question about your sizing chart, that’s not a CS problem — that’s a listing-copy problem, and now you have the data to fix it.
Where the math breaks
Two hard questions. First, cost per session. The launch page doesn’t disclose pricing — not disclosed anywhere in the source — so I can’t tell you whether this is a $29/month tool or an enterprise contract. For a seller running 500 post-purchase walkthroughs a month, the unit economics only work if the AI session costs meaningfully less than the support ticket it replaces. Test that before you scale it.
Second, conversion versus deflection. The maker says the AI can answer beyond the original slides using provided sources and can generate new visuals in real time, with creator-set guardrails controlling how far it roams. That’s the right architecture. But “understands where someone is stuck” is currently, by the maker’s own admission, “a mix of signals” that they’re “working on making much smarter over time.” Translation: the struggle-detection is early. Don’t build a business process on the assumption that it reliably diagnoses confusion.
Where I think it falls short for cross-border specifically
I’ll be blunt about the gaps, because that’s the job.
Language is unproven. Nothing in the launch material confirms multilingual presentation or translation quality. For a cross-border operator, that’s the whole ballgame. If the AI presents only in the language you taught it, the localization use case collapses and you’re back to hiring native speakers.
Trust and brand control. Letting an AI represent your brand live, one-on-one, with buyers, is a reputational bet. The guardrails exist, but a hallucinated claim about shipping times or warranty terms in a private sales session is a real liability. In regulated categories — supplements, electronics with safety claims, anything with compliance language — I’d want a hard review gate.
The “presentation” framing may undersell it. The launch leans on decks and slides, which makes it read like a sales-enablement tool. The more valuable use for sellers is asynchronous, link-based product education — closer to an interactive manual than a pitch. If the product team keeps optimizing for the pitch, they’ll miss the ops buyer.
Early social proof. Two reviews, both from the timeline era, both positive but thin. This is a fourth-launch company still finding product-market fit. That’s fine — it also means you should pilot, not commit annually.
What I’d watch / test next
This week, if you want to pressure-test the thesis without a big commitment:
- Pick one high-ticket SKU with a real setup or usage learning curve — not your bestseller, your most confusing product.
- Build a five-minute walkthrough from your existing manual, FAQ, and unboxing video, and set explicit guardrails: no pricing promises, no shipping-time claims, no warranty language outside the approved text.
- Send it to ten recent buyers as post-purchase education, with a plain-language opt-in and a clear “talk to a human” escape hatch. Track ticket volume and review sentiment against a control group of ten.
- Run the same test on one wholesale inquiry and watch whether the session shortens the close.
- Ask the vendor directly about multilingual output and per-session pricing — both are undisclosed on the launch page, and both decide whether this is a toy or a line item.
My working hypothesis: MyLens is a preview of a category — AI-run, one-to-one customer conversations — that will matter enormously to cross-border sellers within two years, and is currently a useful pilot with real gaps in language and trust. Watch the localization roadmap. That’s the signal.
