What AI can buildvs what actually ships
87 real AU projects. Benchmarks for founders in Australia and New Zealand.
Executive summary
In the last six months, a noticeably larger share of founders reaching out to me at 44Degrees already have something built. A vibe-coded prototype. A cheap offshore MVP. A “working” app from an AI tool that demoed well in a Google Meet. The build step got cheaper and easier. However, what did not get easier is knowing what to tell your AI agent to build or change: market intelligence, positioning, validation, UX, acquisition and retention strategy still all depend on judgement, taste, deep industry knowledge, and experience that unguided AI does not provide.
Across our project base at 44Degrees, the pattern is consistent: roughly 80-90% of pre-built apps that we review either require significant rework or a rebuild before we would call them launch-ready. Very few walk in ready to scale.
That's the overall shift we're seeing in 2026. What follows is the benchmark data that Google, ChatGPT, or Claude's AI won't give you:
- Cost: Our analysis of 87 real AU app projects (2022-2026) shows a validation-first startup MVP typically costs $25,000-$50,000 AUD for design and build, after a $12,500-$15,000 AUD validation gate. The full path commonly lands at $45,000-$85,000 AUD. The comparable build-first path runs $100,000-$150,000 AUD, and full build-first industry projects commonly reach $150,000-$300,000 AUD or much more as scope grows.
- Validation outcomes: For ideas founders bring us for a build/no-build decision, our rough split is ~15-20% GO (proceed to build), ~30-35% PIVOT (change the idea, model, or positioning), and 50%+ NO-GO (stop before major spend).
- Scope: at least 50-70% of originally specced features are commonly cut before MVP ships - often much higher.
- Time: Most MVPs ship in 6-10 weeks through the 44Degrees process when scope is disciplined upfront.
We reviewed Google search results for AU and NZ app cost queries in June and July 2026. They are dominated by agency guides with conflicting ranges - not a shared methodology or first-party sample. This report fills that gap with practitioner observations backed where possible by government data.
Methodology
First-party figures come from 44Degrees' assessment of its own project base (2009-2026). They are practitioner estimates and observations, not a formal survey or market-wide truth for all of AU/NZ.
| Dataset | What it covers | How to cite it |
|---|---|---|
| Cost benchmarks (n=87) | AU app projects analysed for the public cost guide (2022-2026) | 44Degrees sample of 87 AU app projects |
| Validation outcomes | Ideas founders brought for a build/no-build decision; 15+ years of patterns | Experienced, practitioner estimate |
| Pre-built rework (80-90%) | AI/offshore/DIY builds without expert product direction | Practitioner observation |
| Scope reduction (50-70%+) | Projects that proceed after validation | Common range before MVP ship |
| Time-to-launch (6-10 weeks) | End-to-end MVP through 44Degrees process | Typical band, not a guarantee |
Third-party figures use approved primary sources only (see Sources). Google search landscape notes were captured in June and July 2026. Google's suggested follow-up questions change over time, so we describe recurring themes in search results rather than quoting specific questions that may differ by the time you read this.
Market context - why AU/NZ rewards apps handsomely when they know what they're doing
First, the good news: AU and NZ are among the most mobile-first markets on earth, so a genuinely good app has the potential of reaching a large, engaged, high-value mobile audience.
Australia: ACMA reports 99.7% of adults had internet access in 2025, 97% via mobile, 92% daily mobile internet use (ACMA, Feb 2026).
New Zealand: WIP NZ 2023 (AUT, n≈3,100): 97% connect from home, 82% via mobile data, 75% use phones several times a day (WIP NZ 2023).
But that reward is contested. In 2025 alone, Apple's App Review team reviewed more than 9.1 million app submissions - up from 7.7 million a year earlier - rejected over 2 million for failing its guidelines, and blocked $2.2 billion in fraudulent transactions (Apple, 2025).
Apple and Google are also constantly reshaping how apps get discovered, from AI-assisted search to short-form video (Google I/O 2026).
A crowded, tightly policed store rewards the few who get product-market fit and positioning right, and punishes the many who don't - poor product-market fit is still the number-one reason startups fail (CB Insights, 2024).
Shipping an app is no longer the hard part. Whether users find it, convert, and keep using it still decides if the business works.
Australia is mobile-first - apps compete on phones, not desktops. Source: ACMA 2024-2025.
Getting listed is easy; standing out and staying trusted is the contest. Source: Apple, 2025.
The AI-build shift - faster to code, but strategy still decides what succeeds
Over the past six months, more and more founders arrive having already built something - usually without that strategic layer. Across these inbound projects, roughly 80-90% need significant rework or an entire rebuild before we would call them launch-ready. Not because AI cannot execute the fixes, but because no one with real product judgement or deep industry understanding knew how to guide what should have been built or changed in the first place.
Only when the pilot, or strategist, can clearly and intelligently guide the AI, can the AI ship your product quickly.
When we assess arrivals that skipped proper validation and expert product direction, failures cluster in four buckets:
- No fundamentals - unproven demand, no credible user acquisition plan
- Technical issues - wrong framework, bugs, store rejection risk
- No test marketing - acquisition cost never gauged with real spend
- Product and UX gaps AI cannot diagnose alone - wrong flows, weak positioning, defaults that lack taste. A strategist has to name the fix; then AI can often ship it quickly.
The validation gap - most ideas we assess don't proceed as originally pitched
CB Insights' 2024 update analysed 431 failed VC-backed companies: 43% cited poor product-market fit (CB Insights, 2024). Our own GO/PIVOT/NO-GO split below reflects similar results from ideas founders bring to us for a formal build/no-build decision via our pre-build App Idea Validation service.
| Outcome | Share of assessed ideas | What it means |
|---|---|---|
| GO | ~15-20% | Proceed to MVP design and development |
| PIVOT | ~30-35% | Solution, model, or positioning needs change |
| NO-GO | 50%+ | Stop as proposed - before major build spend |
Experienced practitioner estimate from 15+ years of practice.
Observations from our internal practitioners - not a formal survey. Range bands shown, not false precision.
What apps actually cost
The following benchmarks are from our own client sample (2022-2026), documented in our app development cost guide.
Validated MVP definition
- MVP design (Phase 3, after validation): $5,000-$20,000 AUD
- MVP development (Phase 4): $25,000-$50,000 AUD
- Full path with validation gate (Phases 1-2, ~$12,500-$15,000 AUD): commonly $45,000-$85,000 AUD total
| App profile | 44Degrees (validation-first) | Industry average (build first) |
|---|---|---|
| Startup app (MVP) | $25,000-$50,000 AUD | $100,000-$150,000 AUD |
| Growth-stage app (medium complexity) | $60,000-$120,000 AUD | $200,000-$300,000 AUD |
| Platform-scale app (complex) | $150,000-$300,000 AUD | $500,000-$1,000,000+ AUD |
Industry ranges reflect typical agency quotes and published guides for comparable scope in AU/NZ - not a single survey. See our cost guide for methodology.
Startup app (MVP)
Growth-stage app (medium complexity)
Platform-scale app (complex)
44Degrees n=87 sample vs typical build-first industry paths for comparable scope.
Path comparison
Build first
- $150,000-$300,000+ AUD upfront (rising with scope)
- 6-9 months before you know if the idea is worth building
- Weak idea = six-figure sunk cost
Validate first
- ~$12,500-$15,000 AUD validation gate
- 4-6 weeks to GO/PIVOT/NO-GO
- Weak idea = ~$5,000-$15,000 AUD learning cost
Validated builds ship ~30-40% of originally imagined scope. A ~$15,000 AUD validation can prevent a $150,000+ AUD (or more) expensive mistake.
65%+ lower total spend vs typical build-first paths - mainly from validating demand and willingness to pay before building, and cutting scope to only what testing supports.
Scope and time-to-launch reality
When a project passes our pre-build validation process and is ready to proceed, we commonly see at least 50-70% of originally specced features cut before MVP ships - often much higher. Furthermore, most MVPs ship in 6-10 weeks end-to-end when validation happens first and scope decisions are made in the validation window, not mid-build.
What founders ask Google (and what Google gets wrong)
Founders commonly search queries like “how much does app development cost australia” and “how to validate an app idea.” However, here's what Google returns today:
- Cost queries: Agency guides with wide, conflicting ranges - no shared methodology or first-party samples.
- Validation queries: AI tools, Reddit threads, and procedural how-tos - not expert analysis with clear GO/PIVOT/NO-GO outcomes or real project cost bands.
Founders who research only on Google also typically end up inheriting other people's marketing funnels or ideas, not the tools they need to innovate and cut-through in today's competitive market.
What this means for founders
- Budget ~$5k-$15k and 2-6 weeks for validation before any major build commit. Prove demand and willingness to pay and your ability to acquire users at a profit.
- Expect NO-GO or PIVOT more often than GO - and listen to the data.
- Expect to cut at least 50-70% of your v1 scope, often more; what you think you need is often not what the market is prepared to pay for. Don't over-build before first proving demand.
- If you already vibe-coded something, get expert product direction before you assume the problem is the code. Many fixes are fast with AI when someone experienced tells you what actually needs changing.
- Source expert talent or agencies who have been there and done it, and ask for their methodology, not just hourly rates.
Sources (third-party)
- ACMA telecommunications trends 2023-24
- ACMA mobile internet release Feb 2026
- World Internet Project NZ 2023
- Apple - App Store fraud & review data (2025)
- Google I/O 2026 - Play discovery updates
- CB Insights - Why Startups Fail (2024)
Press enquiries: enquire@44degrees.ai
44Degrees Limited (founded 2009) · Wanaka, New Zealand · 250+ app projects · AppMedia (appmedia.com.au) is 44Degrees's specialist App Marketing sister brand.