App cost calculators: what a ballpark figure can and cannot tell you before you commission software

App cost calculators: what a ballpark figure can and cannot tell you before you commission software
Photo by Rami Al-zayat / Unsplash

An app cost calculator will not give you a final price. What it can do is force the decisions that usually get skipped, and those skipped decisions are where most software budgets start to go wrong.

Overspending on a mobile app rarely begins when developers start writing code. It begins earlier, in the weeks when nobody has written down what the app actually does, and every conversation with a prospective supplier starts from a different set of assumptions. Several development agencies publish free estimating tools to capture enquiries, and some are genuinely useful as a structured questionnaire. LITSLINK runs one on the litslink.com page, which asks a series of questions about platforms, features, design and monetisation before producing an indicative cost and an estimate of development hours.

Official business population estimates have put the UK private sector at roughly 5.5 million businesses in recent years, with around three quarters of them employing nobody other than the owner. That is a large number of people making a four or five figure software decision without a finance director, a technical lead or a procurement team to sanity-check it. A calculator is a poor substitute for any of those, but it is better than a figure someone mentioned at a barbecue.

Here is how an early ballpark changes the decisions that follow.

It forces a written feature list before anyone quotes

A one line brief produces wildly different quotes. "An app like Uber, but for dog walkers" can be imagined as a £15,000 build or a £150,000 one, and three agencies will quietly imagine three different products.

Ten minutes spent answering questions about logins, bookings, payments, messaging, admin tools and user roles produces something crude but concrete: a list. That list can be pasted into every enquiry, which means the quotes that come back are finally pricing the same thing.

The questions tend to be the useful part, because they are written by people who have scoped many builds and they cover the items first-timers forget. Those usually include an admin panel, push notifications, how personal data is stored and handled, and how the business intends to make money from the app, which changes the build more than most owners expect.

Scope arguments are much cheaper on paper than halfway through development. A written list does not guarantee an accurate estimate, but it gives everyone the same starting point.

The features that quietly drive the price

Owners often treat features as roughly equal in cost. A login screen, a payment flow and a live map all sound like "another screen", so dropping any one of them feels like it should save a similar amount. Running the same estimate twice, once with a feature switched on and once with it off, shows that it does not.

Three categories account for most of the unpleasant surprises:

  • Payments, because taking money brings security requirements, refund handling, receipts, reconciliation and the rules card networks and app stores impose.
  • Real-time features such as chat, live location and instant notifications, which need servers that stay connected and have to be tested under load.
  • Integrations with outside systems, including accounting packages, booking platforms or older stock databases, where the timeline now depends on software you do not control.

A calculator will not explain why these cost more. It will show that they do, which is usually enough to start a better conversation with a supplier. Moving the most expensive nice-to-have into version two, then rerunning the estimate, is a quick way to see what a first release really needs to contain.

Platform choice usually moves the number most

The question of iOS, Android, both or neither tends to shift an estimate more than any single feature, because two native apps mean two codebases to write, test and maintain.

Cross-platform frameworks such as Flutter and React Native let one team ship to both app stores from a shared codebase, which is why many smaller projects now use them. There are trade-offs. Heavy use of the camera, Bluetooth or device sensors can pull a project back towards native development. For booking, ordering and loyalty apps, which is what a large share of small UK businesses are commissioning, cross-platform is often the default rather than the exception.

A website may do the job for less

A mobile-friendly website still covers most of what an occasional-use app would do, with no app store review process, no store commission and usually a lower build cost. Seeing the platform line on an estimate is often what prompts an owner to ask whether a native app is needed at all.

That question deserves proper thought before any money is committed, and the same budgeting discipline applies to web projects as well as mobile ones, where scope creep and vague briefs cause the same problems.

It gives you a yardstick for comparing quotes

Without a reference point, most owners pick the cheapest quote because there is no other obvious way to judge the three in front of them.

With a ballpark figure and a rough hour count, a quote at a third of that invites a polite question about what has been left out. One at three times the figure invites a question about what it includes that the others do not.

The obvious objection is that these tools are lead magnets. That is half right. Agencies build them to generate enquiries, not out of charity, and the output reflects one supplier's pricing model rather than the market. The practical response is to run two or three calculators with identical answers and look for where they agree. The area of agreement is the useful signal. The precise number is not, and it is not a quote, a valuation or regulated financial advice.

Why IT budgets overrun, and what that means for small builds

Large IT projects have a long record of costing more than planned. Research published by McKinsey in 2012 with the University of Oxford examined more than 5,400 IT projects and found that large ones ran 45% over budget on average and 7% over time, while delivering 56% less value than predicted. Software projects were identified as carrying the highest risk of cost and schedule overruns.

Those were corporate programmes with budgets above $15m, or very roughly £11m to £12m at recent exchange rates, so the figures do not transfer directly to a café ordering app. The underlying causes, principally vague scope and optimistic early numbers, are not confined to large projects, though the study did not measure small ones.

A ballpark will not prevent an overrun. It will flag the quote that sits far outside the range before anything is signed.

App store fees come out of your revenue, not your build budget

Store commission is a running cost that scales with success, and it is frequently missing from first-year budgets.

Apple charges a reduced 15% commission on paid apps and in-app purchases for developers in its small business programme, which covers those earning up to $1m in proceeds a year, or very approximately £750,000 depending on the rate on the day. Enrolment is not automatic, and the eligibility rules and annual thresholds matter if revenue grows or if a developer is part of a wider group of companies. There is also an application process and an annual re-qualification step that developers have to keep track of rather than assume.

Google's Play fees have been restructured more than once, partly in response to regulatory and legal pressure, and the detail now varies by market and by transaction type. Coverage of the changes explains that a lower headline service fee can sit alongside a separate billing charge, so the effective rate a developer pays can be higher than the headline figure suggests.

Because these schedules have changed repeatedly, the published fee pages for each store are the only reliable source at the point you commit. Whatever the rate, the principle holds: store fees are a percentage of takings, so they grow as the business does, and they should appear in revenue forecasts rather than the build budget.

The bills that start on launch day

Apple and Google release new operating system versions each year. Apps that are not maintained start to fail in small, irritating ways, such as a button that no longer lines up or a login that breaks after a security change.

Hosting, error monitoring, backups, certificate renewals, developer account fees and routine bug fixing all recur. Year two of an app is rarely a zero cost year, and an estimate that covers only development is answering half the question.

A reasonable planning habit is to treat the build as the deposit and the running costs as the mortgage, then budget for both before signing.

How the spend lands in your accounts and your tax bill

How app development costs are treated in the accounts affects reported profit, and in some cases the timing of tax relief.

Under UK GAAP, research costs must be written off as they are incurred, while development costs may be capitalised as an intangible asset where specific conditions are met. An accounting policy summary from ACCA sets out what the recognition criteria under FRS 102 require, including technical feasibility and the intention and ability to complete and use or sell the asset. Capitalising is a policy choice applied consistently, not a free decision taken project by project.

There are commercial consequences either way. Advisers have noted the practical considerations for technology companies weighing capitalisation, including the effect on reported profit, distributable reserves, covenant tests and how investors read the numbers. Capitalising flatters profit in year one and depresses it later through amortisation.

Two further points are worth raising with an accountant before the first invoice:

  • VAT. A VAT-registered business can usually reclaim input VAT on development services used for taxable business purposes, but the cash still leaves the account before the reclaim arrives. Businesses on the flat rate scheme are generally unable to reclaim VAT on services in the same way.
  • Research and development tax relief. Routine app building, using existing frameworks and known techniques, typically does not meet HMRC's test of seeking an advance in science or technology. Some genuinely novel technical work can qualify, but the assumption that any software project does is a common and expensive error.

Read the figure as a range, not a promise

Each of the benefits above comes down to the same move: getting a rough figure early, then doing something specific with it.

What the calculator does Mistake it prevents What to do with the number
Forces a feature list Briefing agencies on a vague idea Paste your answers into every enquiry so all quotes price the same product
Shows cost drivers Treating payments, live features and integrations as cheap extras Move the most expensive nice-to-have into version two and rerun the estimate
Makes platform explicit Paying for two native apps you did not need Compare native, cross-platform and web estimates side by side
Gives a quote baseline Choosing the cheapest quote with no way to judge it Question any quote far below or far above your range
Surfaces running costs Budgeting for launch and nothing after Add an annual line for store fees, hosting and updates before committing

If a tool produces a single figure, treat it as the middle of a range rather than a ceiling, and hold contingency on top. Most calculators price in US dollars, so convert at the rate on the day and allow for currency movement if the supplier is billing abroad.

Saving a copy of your answers is worth the thirty seconds it takes. When a proposal arrives weeks later, you can check it against what was actually asked for and spot features that have crept in or quietly disappeared.

What a calculator cannot tell you

  • Whether anyone wants the app, which only real customers can answer.
  • How well a particular team communicates, tests its work and meets deadlines.
  • What a specific integration will cost once someone has read the other system's documentation.
  • How much the price rises when the brief changes halfway through the build.
  • Whether an app is needed at all, or whether a good website would do.

The value of a ballpark is not precision. It is having a number in writing before the first invoice arrives, so the invoice has something to be measured against.

Sam

Sam

Founder of SavingTool.co.uk
United Kingdom