Article · 10 min read

Build, buy or white label? Dispatch software for fleets over 50 vehicles.

The three-way decision most operators get wrong in the same direction.

By the Limozoft team · Last reviewed 2026-10-19

Once a fleet passes fifty vehicles, someone in the room always says it: we should just build our own. It is a reasonable instinct and an expensive one. Here is the honest cost model for all three routes, with the crossover points spelled out.

Key takeaways

Build: $180k–$400k v1, 12–18 months, permanent engineering headcount
Buy: live in days, vendor branding, vendor roadmap
White label: your brand, your domain, someone else's maintenance bill
App store review, SDK deprecations and OS updates never stop
Payment compliance and PCI scope belong to whoever owns the code

Build: what the quote never includes

The $250,000 build quote covers a first release. It does not cover the things that actually consume the budget:

Two app store review pipelines, forever
iOS and Android OS updates breaking background location twice a year
Map SDK price changes and deprecations
PCI scope, SCA/3DS flows and chargeback tooling
24/7 on-call — because dispatch outages are revenue outages
Bus factor: the developer who wrote it leaves in year two

Buy: fastest to revenue, weakest on brand

Off-the-shelf gets you dispatching this week. The trade is that your passengers download an app with a vendor's name on it, book on a vendor's subdomain, and the vendor decides what ships next. For fleets under 25 vehicles that trade is almost always correct.

White label: the middle path most 50+ fleets should take

Your logo, your colours, your domain, your app store listings — running on a platform that already survived a thousand other fleets' edge cases. You keep the brand equity and the customer relationship; you skip the engineering payroll. See what that actually includes on our white-label dispatch page.

The break-even maths

At 50 vehicles, a white-label platform at $30 per vehicle per month costs $18,000 a year. A build costs that much in a single sprint month.

The build only wins if it produces revenue the licensed platform cannot — a proprietary matching algorithm, a regulated integration nobody sells, or a marketplace you resell to other operators. If you cannot name that revenue in one sentence, you are buying a slower copy.

The five-question decision test

Can you name the feature only you can build, and the revenue it creates?
Can you fund two engineers and a QA for five years, not one?
Do you need your brand on the passenger app? (If yes, rule out plain buy.)
Do you need to be live in weeks or is 18 months acceptable?
Who carries the pager at 2am on New Year's Eve?

Frequently asked questions

How much does it cost to build your own taxi app?+

A credible first version — passenger app, driver app, dispatch console, payments and notifications — lands between $180,000 and $400,000, then 18–25% of that per year to maintain. The build is the cheap part; the maintenance is not.

Is white label cheaper than building?+

Dramatically. White label puts your brand on proven apps for a setup fee plus a monthly per-driver rate, typically recovering the cost within a quarter versus a multi-year build.

When does building actually make sense?+

When your dispatch logic is a genuine competitive moat, you run 200+ vehicles, and you can fund a permanent engineering team. Below that, building buys you a slower copy of what you can license today.