An MVP — minimum viable product — is the smallest version of a product that real users can use for its intended purpose. In Australia in 2026, a genuinely scoped MVP from a small engineering team typically lands somewhere between $9,000 and $15,000 AUD, with narrower single-workflow builds starting around $2,000 to $6,000. Quotes ranging from $5,000 to $250,000 for what founders describe as "the same app" are common, and the variance is almost entirely explained by scope definition rather than by rates.
Why do MVP quotes vary so wildly?
Because "MVP" has stopped meaning anything specific, and most quotes are answering different questions.
When a founder describes their product in a paragraph, they are describing a vision. When an agency quotes it, they are pricing an interpretation of that paragraph. One agency reads "marketplace" as a listing page with manual matching; another reads it as two-sided onboarding, payments, escrow, ratings, dispute handling and a mobile app. Both are honest readings. The quotes differ by an order of magnitude.
Three other factors explain most of the remaining spread:
Who does the work. A senior engineer who has built the thing before is expensive per hour and cheap per outcome. This is not a platitude — the dominant cost driver in software is rework, and rework comes from decisions made by people who have not seen the failure mode before.
Whether the scope is fixed. Time-and-materials quotes are structurally lower at the point of sale and structurally higher at the end. A fixed-scope quote prices the risk of being wrong; an hourly quote transfers it to you.
What is counted as "the MVP". Some quotes include design, deployment, a staging environment, and handover documentation. Some include none of it, and those things arrive later as change requests.
What actually drives the cost up?
In rough order of impact.
Integrations with systems you do not control. The single most underestimated cost. Connecting to Xero, a bank feed, an insurer portal or a legacy internal system means writing against someone else's API, their error cases, their rate limits and their outages. A product with three external integrations is not slightly more complex than a product with none; it is a different class of project.
Payments and anything regulated. The moment money moves, or personal information of a sensitive kind is stored, the compliance surface changes and so does the testing burden.
User-generated content and moderation. Easy to build, expensive to make safe.
Real-time anything. Live updates, collaborative editing, presence indicators. Each pushes you into architecture that costs more to build and considerably more to operate.
Mobile as well as web. Frequently doubles the surface area for a fraction of the early learning. Most MVPs do not need it; a responsive web app answers the same questions.
Design ambition. A functional interface built on a solid component system is cheap. A bespoke, brand-differentiated interface is a separate project running alongside the engineering.
What does not drive cost as much as founders expect: the number of screens. Ten simple screens over one clean data model is a smaller job than three screens over four integrations.
What should you actually build first?
The smallest thing that answers your riskiest question.
This is the discipline that separates a $12,000 MVP from a $120,000 one, and it is almost never a technical decision. Every product has one assumption that, if wrong, makes everything else irrelevant. Usually it is "will anyone pay for this" or "will people actually change how they work to use this." The MVP's job is to test that assumption with real users, as cheaply as possible.
Concretely, that usually means: one user type rather than three. One workflow, end to end, rather than five partial ones. Manual back-office processes instead of automated ones — if a human doing the matching by hand proves people want the matching, you have learned the same thing for a tenth of the cost. No admin dashboard; use the database. No settings; pick sensible defaults. No onboarding flow; call your first ten users.
The instinct to build more is understandable and almost always wrong. Features you build before you have users are features built on assumptions, and the ones that survive contact with real usage are rarely the ones founders expected.
How does the R&D Tax Incentive change the real number?
Materially, but later, and not for the whole build.
Australian companies with aggregated turnover under $20 million can claim a 43.5% refundable tax offset on eligible R&D expenditure. Refundable matters here: a pre-revenue company with no tax payable can still receive it as cash.
Three caveats that determine whether this is real money to you:
Only the experimental portion typically qualifies — the parts where the technical outcome genuinely could not be known in advance, not the whole build. Building a login screen is not R&D.
Registration is due within 10 months of the end of the income year in which the activities occurred. Miss it and a valid claim is simply gone.
Eligibility is decided by AusIndustry and the ATO, on advice from your registered R&D tax advisor. No development agency can promise you an offset, and any that does is telling you something about how they operate.
The practical effect is that a build costing $12,000 today may effectively cost meaningfully less once a successful claim lands in the following financial year. Plan and fund the full amount; treat the offset as runway that arrives later. We cover this in more depth in funding your MVP with the R&D Tax Incentive.
What should you refuse to compromise on?
Three things, because each one is far more expensive to fix later than to get right now.
Owning your code and IP. Full source, in your repositories, from day one. This is not only about independence — it is also relevant to the R&D incentive, which contemplates the claiming company owning the IP, directing the work and bearing the risk.
Where your data lives. Australian-resident infrastructure or your own servers. Once you have users, migrating jurisdictions is painful, and if you sell into regulated sectors you will be asked. There is also a live compliance dimension: under s 16C of the Privacy Act, an Australian entity remains accountable for personal information it discloses to an overseas recipient.
Building onshore if you intend to claim. Overseas activities are generally ineligible without an advance finding from AusIndustry, obtained before the work happens.
So what is the honest number?
For a scoped MVP — one user type, one core workflow, minimal external integrations, web-first — expect $9,000 to $15,000 AUD from a small senior team, with a working product in front of users in weeks rather than months. Narrower builds automating a single workflow sit in the $2,000 to $6,000 range.
If your quote is dramatically below that, check what has been excluded. If it is dramatically above, check whether you are being quoted an MVP or a full product — and whether the scope can be cut in half without losing the question you are trying to answer. Usually it can.
That is the conversation our MVP development engagements start with: what is the riskiest assumption, and what is the smallest real thing that tests it.
Ready to find out what yours would cost?
If you have a product idea and want a straight answer on scope, timeline and price, it is worth a conversation. Book a discovery call and we will scope it properly, tell you which parts to cut, and quote a fixed price — your IP, your repositories, built in Australia.
Sources
- ATO, Rates of R&D tax incentive offset — 43.5% refundable offset
- ATO, Refundable and non-refundable offsets — aggregated turnover under $20 million
- business.gov.au, Overview of the R&D Tax Incentive — 10-month registration deadline, overseas and advance findings
- OAIC, APP 8: Cross-border disclosure of personal information
Price ranges reflect Zatersio's own fixed-scope pricing for Australian SMB and startup builds, not an industry survey. This article is general information, not tax advice — eligibility for the R&D Tax Incentive is determined by AusIndustry and the ATO.
