3 tests before hiring your first mobile dev

Hiring your first mobile developer in France in 2026 costs €500-700 per day senior freelance (roughly €100-140k/year FTE) (Malt — Freelance IT day rates), or €45-70k gross annual senior CDI (i.e., €60-95k loaded, equipment included) (Welcome to the Jungle — Flutter developer salary).

Without methodical upstream validation, a founder discovers if their idea deserves this budget 6 to 9 months after hiring — when the v1 ships and the first user metrics start to speak.

There are 3 tests documented by Amazon, Dropbox and Airbnb that allow this validation in less than a week each, for less than €500 per test, without writing a line of code. Each kills a different risk hypothesis. Three tests passed = you can hire with confidence. One test failed = you just saved €40-80k.

The #1 cause of startup failure (and what it teaches us)

According to CB Insights, which annually analyzes public post-mortems of closed startups, the top cause of failure is “no market need” — 35-42% of cases depending on the year (CB Insights — Top Reasons Startups Fail). This cause outweighs financing problems, team problems, and technical problems combined.

Translated to practice: most founders who fail didn’t get the CODE of their app wrong. They got wrong the very existence of a user need for what their app did. These are exactly the cases the 3 tests are designed to detect — before having spent 6 figures.

Test 1 — The press release (Amazon Working Backwards)

The method. Before specifying a single feature, write the press release for your app as if it existed and you were announcing its launch. One page max. Then write the FAQ (10-15 questions/answers a user, an enterprise customer, or a journalist would ask).

This method has been formalized at Amazon since 2004 under the name Working Backwards — “working backwards from the customer.” It’s documented in detail by Colin Bryar and Bill Carr, two ex-Amazon executives, in the book Working Backwards (Macmillan, 2021) (Bryar & Carr — Working Backwards (2021)).

The test. You know the test failed if:

• The press release is flat, generic, or describes a feature rather than a user benefit
• The FAQ contains more internal questions (“how do we architecture this?”) than user questions (“how much does it cost?”)
• You can’t write a title in 8 words or less that expresses the promise

Template to copy:

FOR IMMEDIATE RELEASE

[App name] — [8 words max expressing the promise]

Subtitle: [1 sentence describing the measurable benefit]

[City, Date] — [App name] launches today [1-sentence description].
Users [concrete action] and get [measurable outcome] in [realistic duration].

The problem [App name] solves:
"Every day, [user type] has to [painful action]. It costs them
[time / money / energy]."

What [App name] changes:
"[App name] enables [user type] to [easy new action] in [reduced time]."

User quote:
"[Short sentence describing an aha-moment]." — [First name, age, role]

Availability and pricing:
[App available on iOS/Android starting at $X/month]

Test cost: 30 minutes of writing. Zero €. Cost if you skip the test: 6 months of development to discover the release is flat.

Test 2 — Latent demand (Dropbox fake video, Buffer landing)

The method. Create a landing page describing your app with the main promise, a waitlist signup form, and a “Download the app” button that redirects to a “coming soon” message. Drive traffic to it (Facebook Ads, LinkedIn, target Slack/Discord communities). Measure the click-through rate on the “Download” button.

This method is known as fake door test in product discovery, or smoke test in growth hacking. It was made famous by two cases:

• Dropbox (2008) — Drew Houston created a 4-minute video showing the product as if it existed, with inside jokes for the Hacker News community. Overnight, the waitlist went from 5,000 to 75,000 people (TechCrunch — Dropbox's Minimum Viable Product). The product didn’t exist yet.
• Buffer (2010) — Joel Gascoigne created 2 successive landing pages. The first presented the functionality and a “Plans and Pricing” button. The second added tiers. He only wrote a single line of code when the 2nd landing was converting (Buffer — Idea to Paying Customers in 7 Weeks).

The test. You know the test failed if:

• Fewer than 10% of visitors who reach the landing click the CTA
• The waitlist contains only your friends and family
• Users who sign up don’t respond when you reach out

Test cost: no-code landing (Framer, Webflow, Carrd) — 3-4 hours. Ad budget to bring 500-1000 visitors — €200-400. Total: ~€400. Cost if you skip the test: 6 months of development to discover no one was interested.

Test 3 — Payment possible (Airbnb concierge, Zappos MVP)

The method. Find 5 target users. Deliver them manually the service your app is supposed to automate. Charge them at target price from user #1. There’s no app, no automation — just a human (you or a partner) executing the promise by hand for 2-4 weeks.

This method is the concierge MVP (sometimes called Wizard of Oz MVP). Three universally known cases:

• Airbnb (2008) — Brian Chesky and Joe Gebbia personally took professional photos of the first NY host listings. This manual action doubled the booking rate. The result convinced them to code the automation (Airbnb — Origin story).
• Zappos (1999) — Nick Swinmurn photographed shoes at local mall stores, put them on a simple site, and when an order came in, he bought the pair at the store and shipped it himself. No inventory, no logistics, no app (Forbes — How Tony Hsieh Changed Online Retail with a Simple MVP).
• Instacart (2012) — Apoorva Mehta and the 2 co-founders personally did the shopping for the first users. They collected the order via a hasty app and delivered same-day (First Round Review — Instacart origin story).

The test. You know the test failed if:

• You can’t find 5 users willing to pay the target price
• The 5 users test but don’t renew the following month
• Manual service takes 10× longer than you estimated (strong signal: the app won’t be profitable either)

Test cost: your time over 2-4 weeks (to measure seriously — this is the main cost). Optionally a simple landing site (~€200). Cost if you skip the test: 6 months of development to discover no one pays.

The decision matrix — when to hire

What these 3 tests don’t cover

The 3 tests answer three product questions:

• Test 1 — Is the idea formulable in an exciting way?
• Test 2 — Is there appetite from the target?
• Test 3 — Do users pay at target price?

They don’t answer these technical questions:

• Real-time latency (WebRTC, MQTT, live chat) — requires a technical prototype
• AI model accuracy — requires a test dataset and systematic evaluation
• Integration complexity with an external system (payments, ERP, PMS…) — requires a technical spike

If your idea depends on one of these hard technical constraints, plan a 2-to-4-week technical spike with a freelancer, in parallel with the 3 product tests. This dual validation (product + technical) is what makes a post-validation hire truly safe.

A method that replaces sterile debates

What I deliver on a mission isn’t just Flutter code. It’s a product validation system including test structuring, landing and concierge instrumentation, and reading the go/no-go signals. Code is the result of validation, not the reverse.

What to do this week if you’re a pre-hire founder

Three concrete actions, in order:

1. Today — block 30 minutes to write the press release for your app (template §Test 1 above)
2. This week — put a landing page live with the promise and a “Download the app” CTA that captures emails
3. This month — find 5 target users and deliver them the service by hand, charging the target price

If all 3 tests pass, hire with confidence. If a single one fails, you’ve just saved yourself €40-80k.

Main sources

• Bryar, C. & Carr, B. — Working Backwards: Insights, Stories, and Secrets from Inside Amazon, Macmillan, 2021
• CB Insights — Top Reasons Startups Fail
• Malt — Freelance IT day rates
• Welcome to the Jungle — Flutter developer salary
• TechCrunch — Dropbox’s Minimum Viable Product
• Buffer — Idea to Paying Customers in 7 Weeks
• Forbes — How Tony Hsieh Changed Online Retail with a Simple MVP
• First Round Review — How Instacart’s Founder Built a Billion Dollar Marketplace
• Marty Cagan — Inspired (SVPG)
• Day 1 — The five questions before developing a mobile feature — product method preceding these 3 tests