Willingness-to-pay signals are the cheapest form of market research a cross-border seller can run
Cross-border operators are sitting on a validation problem that has nothing to do with logistics. Most of us can source a product, spin up a Shopify storefront, list on Amazon, or push a TikTok Shop SKU live in a weekend. What we cannot do cheaply is find out whether anyone will actually hand over money before we commit inventory, ad spend, and a container deposit. A soft launch on Product Hunt called Would you pay is a small, imperfect experiment in exactly that direction — a Tinder-style swipe deck where the only question is “would you pay for this?” — and the mechanics behind it are more relevant to a private-label seller than the indie-hacker framing suggests. The interesting part is not the app. It is the discipline of asking a harder question than “do you like this?”
What Would you pay actually does, and why the framing matters
The product is built by Mustafa Ergisi, a maker who has shipped a string of side projects — AI2SQL, AI2image, CrewClaw — and got tired of the same pattern: months of building, a launch, then silence. His own words on the launch page: “Nice comments told me nothing about whether anyone would actually pay.” That sentence is the entire thesis, and it should be pinned above the desk of every DTC founder who has ever mistaken a viral TikTok for demand.
The mechanic is deliberately blunt. A maker adds a startup with nothing but a URL; the system reads the site and generates a card. Swipers see one card at a time and swipe right for “yes, I’d pay,” left for “no.” After ten swipes, the maker sees the percentage of people who said they would pay, plus who those people are — developers, founders, or marketers — and how many clicked through to the site. There is no signup and no email required from the swiper, which is the detail that makes the sample even remotely honest.
The results Ergisi published are refreshingly unflattering. He put his own products in first. The best one hit 38%, and some landed around 20%. In the first day, 6 makers added startups and 27 people swiped 172 times. Those are tiny numbers, and he does not pretend otherwise — but he calls the 20% outcomes “painful, but far more useful than guessing,” which is the correct emotional register for anyone who has ever watched a product die quietly after a warm reception.
Why Amazon sellers should care more than Shopify ones
Shopify operators can test demand with a $200 ad budget and a pre-order button. Amazon FBA brand owners cannot. By the time you have a listing live, you have already paid for a supplier deposit, a freight forwarder, a barcode, and probably a photoshoot. Amazon Seller Central gives you almost no cheap way to ask “would you pay $34.99 for this?” before the inventory exists. That asymmetry is why a swipe-based willingness-to-pay signal — however crude — is worth more to a private-label seller than to a dropshipper.
The same logic applies with extra force on Temu and SHEIN, where price compression means the only question that matters is whether your margin survives at the price the market will bear. A yes/no swipe tells you demand exists. It does not tell you whether demand exists at your cost structure. That gap is the whole game.
How it differs from the tools you already use
The honest comparison set here is not other Product Hunt launches. It is the validation stack a cross-border operator already pays for, and it is worth being precise about what each one actually measures.
Helium 10 and Jungle Scout measure existing demand — search volume, review velocity, BSR trends. They are backward-looking by design. They tell you a category is already being bought, not whether your specific angle will be bought. Google Trends is directional noise. PickFu gets you real consumer opinions on packaging and creative, but it is a paid panel answering “which of these do you prefer,” not “would you pay.” Klaviyo and your email list can run a real pre-order test, but only if you already have an audience — which most first-product sellers do not.
What Would you pay adds is a stated-intent layer that is faster and cheaper than all of the above, and weaker than all of them too. Stated intent is famously inflated: people say yes to things they never buy. Ergisi’s own numbers make this visible — a 38% “would pay” rate on an indie tool is almost certainly not a 38% conversion rate. The value is not the absolute number. It is the relative number: product A at 38% versus product B at 20% is a real signal about which one deserves your next $5,000, even if both numbers are inflated by the same factor.
Where the math breaks
The launch thread contains its own best critique, and it comes from the commenters rather than the maker. Ben Slinger noted that without focus on types of ideas, the deck is “a lot to ask of people, and is very tech focused,” and that “it’s very hard to know from a one-liner whether something is a good idea or not.” That is the core validity problem. A one-line card plus a screenshot is not enough context to make a purchasing decision, so swipers fall back on vibes. Vibes are a fine filter for indie SaaS and a terrible one for a $40 kitchen gadget with a real cost of goods.
Maxwell raised the fix that Ergisi himself agrees with: “a price range on every swipe would sharpen this.” The maker’s reply is the most operationally interesting line in the whole thread — “A yes/no tells you if there’s demand, a price range tells you if there’s a business” — and he confirmed he is planning a “how much would you pay?” step right after a right swipe, weighing fixed ranges ($5 / $10 / $25+) against a slider. For a cross-border seller, that price step is not a nice-to-have. It is the only part of the output that maps to a P&L.
What cross-border sellers can borrow from this
Strip away the swipe UI and there are four transferable practices here, all of which you can run this week without paying for anything new.
Ask the payment question, not the preference question. Every survey you send your list, every poll you run in a Facebook group, every DM you slide into a micro-influencer’s inbox should end with a version of “would you pay $X for this?” Likes, saves, and “this is so cute” comments are the exact noise Ergisi built the product to escape. If your creative testing on Meta or TikTok is optimizing for CTR and not for add-to-cart intent, you are collecting the same useless signal.
Segment your respondents by role, not just by count. The feature Ergisi is proudest of is showing makers who said yes — developer, founder, or marketer — rather than just how many. The e-commerce translation is obvious: a “yes” from a 24-year-old urban renter and a “yes” from a 41-year-old suburban homeowner with two kids are not the same data point. If your pre-order landing page or quiz does not capture a role-like qualifier, you are throwing away the most useful half of the answer.
Force a binary before you force a build. The reason the swipe format works is that it removes the middle option. “Maybe” is where product roadmaps and inventory POs go to die. When you run your next product validation — a pre-order page, a Kickstarter-style reservation, a paid waitlist deposit — make the ask binary and make it cost something, even if it is just an email plus a stated price point. The friction is the feature.
Treat the click-through as the second signal. The app tracks how many swipers clicked through to the maker’s site, which is a behavioral tell layered on top of a stated one. In your own funnel, the equivalent is the difference between someone who says “I’d buy that” in a survey and someone who actually enters a card at checkout. Weight the second one ten times the first.
A sidebar for TikTok Shop and Etsy operators
If you sell on TikTok Shop or Etsy, you already have a cheaper version of this tool: post the product, watch the saves and the add-to-carts, and kill anything that does not move in 72 hours. The swipe deck is most useful to you not as a validation tool but as a positioning tool — writing the one-liner that makes someone swipe right is the same skill as writing the hook that stops a scroll. Ergisi’s tip to Chantal Ross — “describe your paid tier in the card’s one-liner, not just the app” — is a direct instruction for your listing title and your TikTok hook. Lead with the offer, not the feature.
Where my judgment says it falls short
Three things bother me, and I would rather say them plainly than dress them up.
First, the sample is too small and too self-selected to be decision-grade. Twenty-seven people swiping 172 times is a smoke test of the smoke test. The swipers are Product Hunt users, which is a heavily skewed population — technical, early-adopter, allergic to paying for consumer goods. If you are a cross-border seller of physical products, this audience is close to the worst possible proxy for your buyer.
Second, the deck is unfiltered by category, which Julian Ting pressed on directly. Ergisi’s answer is candid: right now everyone sees the same deck in random order, with new startups prioritized until they hit 10 swipes, and interest-based matching is “on my list as the deck grows.” That is a reasonable MVP tradeoff, but it means a beauty brand and a B2B dev tool are competing for the same swipe attention. Relevance matching is not a nice-to-have here; it is the difference between a signal and a coin flip.
Third, and most importantly for my readers: a “would pay” percentage is not a business. It does not tell you your landed cost, your return rate, your ad cost per acquisition, or whether the margin survives a Temu price war. The tool’s own maker admits the price question is still unbuilt. Until that ships, this is a directional input at best — useful for ranking three product ideas against each other, useless for deciding whether to place a 500-unit order.
What I’d watch / test next
Three concrete moves for this week. One: add a “would you pay $X?” question to whatever list or community you already own, and segment the answers by buyer role rather than counting raw yeses — if you do not have a list, build a 50-person one from your existing customers before you build anything else. Two: run the same one-liner test Ergisi runs, but on your actual buyers — put two versions of your listing title or TikTok hook in front of a small paid audience and measure add-to-cart, not CTR. Three: watch whether Ergisi ships the price-range step he promised after a right swipe, because a stated price point plus a stated intent is the first version of this tool that a cross-border seller could actually put into a P&L. Until then, treat every “would pay” number — including the 38% — as a ranking signal, not a forecast.
