7 Ways to Validate a Product Idea Without Building an MVP (2026)

"Just skip the MVP" has become its own cargo cult. Three years ago the indie-hacker church chanted "ship the MVP, the market will tell you". In 2025 it flipped to "never build, just run a fake door". We're now watching founders run a Reddit-ad smoke test for a network-effect marketplace and call a 1.8% CVR a "kill" — when the test was never going to validate that product in the first place.

Skipping the MVP isn't the answer. Building a different artifact that produces signal faster than an MVP can — and matching the artifact to the product's risk profile — is.

Below are seven methods we've seen work. None of them is "skip the MVP". Each one is a specific test for a specific shape of risk, with a real 2026 cost, a kill criterion, and a list of when it's the wrong call. The method has to fit the product. That's the whole framing the cargo-cult version misses.

This piece is the ranked-listicle sister of the cheapest way to validate a product idea, which compares 11 methods on raw signal-per-euro. If you haven't read validate or build an MVP first for the order-of-operations argument, start there. This article assumes you've already decided to validate before building and you want the menu.

How we score each method

Every method below carries a uniform scorecard:

• Cost — realistic out-of-pocket range in EUR, doing it yourself in 2026.
• Time to answer — total elapsed days from "decide to test" to "have a number you can act on".
• Signal quality — 1 to 5, how much you should trust a positive result.
• Effort — 1 to 5, focused-work hours.
• Kill criterion — the single number that says "stop, this isn't working".

We've personally run six of the seven on real ideas this year. Two passed. Four got killed by Sunday.

1. Landing page + paid-ad smoke test

Cost: €150–300 · Time: 10–14 days · Signal: 5 · Effort: 3 · Kill: under 2% CVR after 1,000 visitors, or CPL over €10

The canonical demand test. Build a one-page pitch, send €150 of paid traffic at it from Reddit, Meta, or Google Search, watch what strangers do.

Why it dominates the list: it's adversarial. Strangers don't owe you politeness, paid traffic doesn't grade on a curve, the dashboard is auditable. Foti Panagiotakopoulos at GrowthMentor shipped €418 of Google Ads over 14 days at a 16.89% landing CVR, €0.94 CPC. His own caveat is the honest part: "since there was no paywall in front, it did not prove that users were willing to pay for our service." The smoke test answered demand-for-the-promise; method 2 is what answers demand-for-the-price.

Tools: LemonPage (page + Reddit/Meta/Google ads + measurement in one workflow), Carrd Pro Lite at $9/year (page only — bring your own ads), Framer at $15/month (design-led, monthly cost), Webflow (heavy for this job, designer-grade). Pick based on whether you want one workflow or four; the page itself isn't the differentiator.

What you do with the data: package the five numbers (CTR, CVR, CPL, qualitative comments, click demographics) into one slide. That slide is now the spine of your fundraising deck if you're raising. Investors spend more time on a clean traction slide than on the next ten pages.

When to use it: any new product, any audience, any vertical, when you have no warm list and no existing distribution. When NOT to use it: deep-tech (the test can't validate technical risk), regulated industries (a CVR doesn't unlock a banking licence), or true network-effect products where value emerges only after multiple interactions.

2. Pre-sale via Stripe Payment Link or Gumroad

Cost: €5–20 setup · Time: 3–7 days once you have traffic · Signal: 5 · Effort: 3 · Kill: zero paying strangers in 7 days at the planned price

The "real money before code" test. Same one-page pitch as method 1, but the CTA changes from "Reserve your spot" to "Pay €29 now". A Stripe Payment Link charges 1.5% + €0.25 on EU cards, takes twelve minutes to wire. Gumroad takes 10% + $0.50 on direct sales but handles VAT as merchant of record. For a first cycle, Stripe wins.

Joel Gascoigne ran the canonical version of this for Buffer in 2010. Two-page site, then a third page in the middle showing pricing, button to test willingness to pay. 120 email signups across 7 weeks. 50 became users on launch day. First paying customer 4 days later. His own framing: "I didn't get ‘a billion signups’, in fact in a long 7 week period I only got 120 signups. But I spoke with a lot of those people during that time." Self-funded into product-market fit.

The harder modern version: Justin Welsh in 2024 spent $40.93 of paid promotion in 10 minutes and generated $7,500 of pre-sales for a product that didn't exist. Verified to his own newsletter; treat the exact ratio as illustrative, not industry benchmark.

When to use it: when you have at least a trickle of traffic, from method 1 or a small audience. When NOT to use it: when there's literally no audience yet (a pre-sale page with zero visitors is a ghost — pair with method 1 first), or when the product price is so high that pre-paying without seeing it is unreasonable (€10K B2B contracts, regulated services).

3. Demo video / Wizard-of-Oz

Cost: €0–50 · Time: 1–2 weeks · Signal: 4 · Effort: 3 · Kill: under 1% click-to-signup on the video, or under 500 views from the seeded channel

You film a 3–4 minute demo of a product that does not exist. The product behind the screen-recorded UI is you, manually, in a Google Sheet. The customer sees magic.

Drew Houston ran the textbook version on April 5, 2007. Posted a 4-minute Dropbox screencast to Hacker News — title: "My YC app: Dropbox — Throw away your USB drive". Waitlist went from ~5,000 to 75,000 in a single day. Sequoia led $1.2M seed shortly after. Drew, this past March on X: "My (perhaps questionable) strategy for getting into @ycombinator: 1) Post a Dropbox demo on Hacker News 2) Pray @paulg @jesslivingston see it. It actually worked."

The pure Wizard-of-Oz cousin: Aardvark's peer-to-peer Q&A routed every user question to a human respondent and pasted answers back, simulating an algorithm that didn't exist. Validated the matchmaking premise. Acquired by Google for $50M in 2010.

When to use it: when the product is visually demonstrable, when there's a seedable channel (HN, YouTube, a niche subreddit), or when the back-end is fakeable for the test window. When NOT to use it: when value emerges only across long usage (B2B retention products, habit apps), or when no channel will surface a video to your audience.

4. Concierge / done-by-hand service

Cost: €0 (your time) · Time: 2–6 weeks · Signal: 5 · Effort: 5 · Kill: cannot close 3 paying customers manually in 4 weeks

You deliver the outcome by hand for a fixed fee, to 3–5 named customers, before a single line of code. The product looks unautomated to them because it is.

Manuel Rosso did this for Food on the Table. He wanted to build automated meal-planning. Instead he went to one Austin grocery store, manually scanned weekly sales, hand-built shopping lists for individual families, delivered via email. For months. Validated demand and willingness-to-pay before writing the matching algorithm. Scaled to thousands of users; acquired by Food Network in 2014.

Early Wealthfront ran the same play. Pre-robo-advisor, founders personally gave investment recommendations to early clients to test demand and pricing. Only after the demand was clear did they automate. They now manage roughly $50B AUM across ~700,000 clients. The concierge phase was the validation phase. (Detailed pre-automation client counts come from secondary sources; the trajectory is verified.)

The kill criterion here is the sharpest line in the whole pre-MVP playbook: if you can't close 3 paying customers manually in 4 weeks, automation isn't the missing piece — demand is. Founders who skip concierge and go straight to code build the wrong automation 60–70% of the time, in our experience. Manual fulfillment surfaces the messy edge cases — the refund, the wrong size, the hand-holding — that no spec doc catches.

When to use it: services-flavoured products, AI workflows where the back-end is the expensive part, marketplaces, B2B with high-touch onboarding. When NOT to use it: pure self-serve consumer products at €5/month (concierge math doesn't scale to those margins), or when "manual" version is technically impossible (latency-sensitive infra, real-time games).

5. Fake door inside an existing product or community

Cost: €0–50 · Time: 1 week · Signal: 4 · Effort: 2 · Kill: under 2% click-through on the door OR no inbound after the click

This is the non-paid cousin of method 1. Add a button to an existing product, an existing audience's feed, or an existing community where you already have credibility — "Try this feature" or "Buy this" — that leads to a "coming soon, leave your email" page. Track click-through against the existing baseline.

The signal is strong because the click happens inside real workflow, not in response to a paid ad. There's no novelty effect from the ad creative; the user wasn't hunting for the feature, they encountered it while doing something else.

Dan Kim ran a 3-pronged fake-door Instagram campaign for WWW, a "social stock market for culture" concept. Campaign C hit 16.5% CTR (vs the 1–2% benchmark). 121 visitors, 3 sign-ups → 2.47% conversion. His framing on what the data meant: "The high CTR proved the ‘desirability’ of the concept. The low CVR revealed the ‘inadequacy’ of my initial MVP." He iterated rather than killed — and that iterate-vs-kill judgement is the actual founder call. The test surfaces which assumption was wrong; it doesn't make the call for you.

The ethical cost of fake doors is real. Mitigate by making the "coming soon" page generous: early access, a discount, a personal reply from a real person. Don't dark-pattern users into rage.

When to use it: when you already operate a product (testing a new feature or pricing tier), or you have an audience (newsletter, Twitter, niche subreddit) where you can drop a credible "we're working on this" link. When NOT to use it: when you have neither a product nor an audience — you're better off paying for traffic via method 1.

6. Cold ICP outreach (B2B)

Cost: €0–20 · Time: 2–4 weeks · Signal: 4 (B2B) / 2 (B2C) · Effort: 5 · Kill: under 5% reply rate AND no second-meetings booked in 50 emails

Email or LinkedIn-message 10–20 named prospects in a tight ICP. Ask for 20 minutes to discuss the problem, not to pitch the product. Cold reply rates in 2024–25 averaged 3–5.1% B2B, top-quartile 15–25%. Of the replies, half show up. That gives 2–4 calls per 50 emails — usable for B2B problem validation when contract values justify the slow signal.

The validation number isn't the demo. It's the booked rate. People don't put time on calendars for problems they don't have. Reply rates roughly triple when the email asks about a specific problem ("is this the hardest part of your week?") instead of pitching a solution. Mom Test discipline: don't pitch, ask. Founders who lead with the product get filtered as salespeople; those who lead with the problem get treated as researchers.

The 3-7-7 follow-up cadence captures 93% of replies by Day 10. Two follow-ups, spaced as named, are the floor.

When to use it: B2B SaaS at €500+/month, services, enterprise tools, anything where a single sale matters. When NOT to use it: B2C at low ACV — strangers don't schedule demos for €15 apps and the signal is muddied by polite curiosity. Methods 1 and 2 dominate there.

7. Refundable-deposit waitlist

Cost: €5–20 setup + processing · Time: 2 weeks · Signal: 4 · Effort: 2 · Kill: under 0.5% deposit-conversion on a list of 500+ qualified leads

The upgraded waitlist. Free email signups are cheap; refundable €5 deposits are signal. The CTA on the landing page becomes "Reserve your spot for €5 (refundable)" instead of "Join the waitlist". The drop in conversion is the price of an honest answer.

Industry-reported lift: refundable-deposit waitlists convert 3–5x higher to paying customer than free waitlists, with 60–80% open rates vs 15–25%. We haven't been able to point to a single named primary case study with full deposit-specific numbers — Robinhood's million-person referral waitlist was free, not deposit-gated, so it isn't quite this method. Treat the multipliers as directional, not gospel.

What it actually does well: it kills vanity-waitlist theatre. Hustle Fund's Brian Nichols is on record: "A list of 500 people with a 40% conversion rate is infinitely more valuable than 10,000 names with a 2% conversion rate." A deposit-gated list of 100 is more useful in a pitch room than a free list of 8,000.

When to use it: as an upgrade to method 1's CTA, especially for B2C with novelty appeal, or for any product where you want to surface a small high-conviction list instead of a big-but-noisy one. When NOT to use it: when the price tag of the eventual product is so high a €5 deposit is meaningless signal (enterprise software), or when refunds in your jurisdiction add legal complexity you can't absorb at this stage.

The 7 methods, side by side

Match the method to the risk profile

This is the part the cargo-cult version of "skip the MVP" misses. The right question isn't "which method is cheapest?"; it's "which method actually tests the riskiest assumption in my product?"

A few patterns that hold up:

• Demand risk dominant (most consumer SaaS, indie tools) → method 1 stacked with method 2. €200, 14 days, two of the highest-signal methods together.
• Willingness-to-pay risk dominant (you know people want it, unsure they'll pay) → method 2 standalone, on a warm list.
• Operational risk dominant (the back-end is the hard part — AI workflows, marketplaces, custom delivery) → method 4 (concierge) or method 3 (Wizard of Oz). Build the manual version, learn the messy operations, automate later.
• Channel risk dominant (you're not sure paid ads work for this audience) → method 5 (fake door) inside an existing community first, before spending on traffic.
• B2B with €500+ ACV → method 6 first for problem confirmation, then method 1 with LinkedIn ads. Outreach gives depth; paid traffic gives breadth.
• Network-effect or marketplace → none of the seven validates this cleanly. Pick the harder side, validate with method 4 (concierge supply), then ship a thin MVP. We say so explicitly in validate or build an MVP first.
• Deep-tech / regulated → none of the seven applies. Validation evidence is the technical milestone (a benchmark, a working prototype, a published result), not a CVR.

Pieter Levels has validated and killed over 70 product ideas, keeping only the 5 that now generate over $3M/year combined. His own framing on the trap of generic validation: "Like woah dude, I didn't even validate the nomad market, it's month two… [an advisor] is like ‘No, you have to go into all these different verticals.’ Then after three years I'm selling furniture or something, it's just too broad. Make it for yourself." The sharpest possible kill rate: an indie-hacker who tested 100 ideas in 30 days, dropped 97, and now runs three at over $180K combined revenue. "If they couldn't get 5 people to talk about the problem, the idea died."

A high kill rate is a feature, not a failure. The seven methods exist so you kill the bad ones cheaply and free the calendar for the survivors.

The trap nobody flags: "saved months of work" is post-hoc

Founders love writing "the test bombed and I saved 6 months." Read closely: most are conflating relief with insight. They wanted to walk away; the test gave them permission. Dan Kim's WWW story is the proof in the other direction — 16.5% CTR with 2.47% CVR didn't kill the concept, it surfaced a trust gap in the page. Same data killed many similar ideas because their founders chose not to iterate.

That choice — iterate vs kill — is the actual judgement call. The test doesn't make it for you. A low-CVR result isn't "the idea is bad"; it's "this offer didn't land with this audience at this price." You still have to decide whether to retry the audience, the price, the angle, or kill outright. Pre-commit the kill criterion before you start, and pre-commit one iteration before final kill. Both rules together stop the two failure modes (false-kill and rationalised-yes).

Recap and the kill funnel

Seven methods, one principle: pick the artifact that produces the cheapest honest signal for your risk profile, run it with a kill criterion written down before you start, and treat a kill as a win.

The 4-tool kill funnel article packages this as a stack: search-demand check, then method 1, then method 2, then converter conversations. Sequence + thresholds. That piece is the operational version of this list.

For the wider field — what AI is changing about validation timing, the 14-day vs 48-hour debate, the order-of-operations argument — see the pillar: how to validate a startup idea in 2026.

Validate your idea on LemonPage — the page, the ads, and the measurement, in one workflow. We built it because we kept losing four hours of plumbing per test, and the whole math of pre-MVP validation only holds when running tests stays cheap.

Stop confusing the MVP with the validation tool. Pick the test that fits the risk. Kill cheaply, build only what survives.

FAQ

Can you really validate a product idea without building anything?

Yes, for most consumer and indie B2B products. Methods 1, 2, 5, and 7 require zero code — a one-page pitch, a payment link, a fake-door button. Methods 3 and 4 require manual operations, not code. Only method 6 (cold outreach) sometimes benefits from a clickable mock. Across the seven, "no code" is the default, not the exception.

How much does a no-code validation test cost in 2026?

€100–€300 covers most pre-MVP B2C and indie B2B tests. €5–€20 if you have warm traffic and run a pre-sale only. €0 for concierge and Wizard-of-Oz, where the cost is your hours instead of euros. Above €500 you're scaling, not validating — pull back to a smaller test and protect the budget.

Which method gives the strongest signal for the lowest cost?

Method 2 (pre-sale via Stripe Payment Link) on a warm list of even 100 people. €5–€20 setup, real money in 3–7 days, signal quality 5. The catch: you need traffic. For cold ideas with no audience, method 1 (landing + paid ads) at €150–€300 is the floor. The two stacked together give the highest signal-per-euro of anything on this list.

Is a fake-door test ethical?

Yes, when the "coming soon" page is generous and visible — early access, a discount, a real reply from a person, a clear "this is in the works" message. It crosses into dark-pattern territory only when the user thinks they bought a product that does not exist with no acknowledgement. The Dan Kim WWW campaign and Buffer's two-page test both ran this method honestly; the model is well-established.

When should you skip these seven methods and just build the MVP?

Almost never, but two cases qualify. If the MVP genuinely takes a weekend (a one-page form-to-email tool, a small Chrome extension), testing it directly is cheaper than testing a fake door. And if you're building for yourself and the build is its own reward, the calculus is personal, not commercial. Outside those two, the seven methods dominate.

What's the kill criterion across all seven methods?

Different per method, but the spine is the same: a pre-committed number, written down before the test starts, with no permission to rationalize past it. Under 2% CVR after 1,000 visitors for method 1. Zero paying strangers in 7 days for method 2. Under 2% click-through on a fake door for method 5. The kill criterion is what separates a validation test from confirmation theatre — without one, every result reads as a yes.