The build estimate people make is usually the development cost. The number that matters is total cost of ownership, and the second and third years are where the two curves cross.
Change any input and the result updates. Nothing is emailed anywhere and nothing is stored.
Break-even is the device count at which licensing over the same period equals the total build cost. Below it, buying wins on cost alone; above it, the arithmetic starts favouring a build — if you can actually staff one.
Almost every in-house build estimate we have seen counts the development sprint and stops. Three costs arrive afterwards and they are what move the answer.
Protocol support is permanent work. A platform is not finished when it decodes your current trackers. New device models ship continuously, manufacturers revise firmware, and a customer will eventually standardise on hardware you have never seen. Every one of those is engineering work, forever, and it does not appear in a launch estimate. Our device guides cover six manufacturers and six protocol families — that breadth is the part that takes years, not the map screen.
Someone has to answer at 3am. A tracking platform that is down is a business that has stopped. Running it means monitoring, on-call rotation, incident response, backup verification and security patching, and those are roles rather than tasks. The calculator above lets you set retained engineers, but two is genuinely the floor if anyone expects a response outside office hours.
Scale changes the architecture. A design that works comfortably at two hundred devices frequently does not at five thousand — the write volume from the data calculator is exactly where this bites. The rebuild that follows is rarely budgeted, because at the time of the original estimate nobody expected to need it.
We license a platform, so read this with that in mind: there are real cases where building wins. If your business logic is the product rather than the tracking — an insurance telematics scoring model, a niche vertical nobody serves, hardware you manufacture yourself — then a generic platform constrains the thing you are actually selling. If you operate at a scale where per-device licensing has become the largest line in the budget, the arithmetic above will say so plainly. And if you already employ a platform engineering team with capacity, the marginal cost is lower than the calculator assumes.
License the platform for tracking, build only the layer that differentiates you on top of its API. You get device coverage, uptime and protocol maintenance as someone else's problem, while owning the part your customers are actually paying for. It is usually the cheapest route to a defensible product, and it is what we recommend to most of the companies who arrive convinced they need to build everything.
The longer-form version of this argument, without the arithmetic, is in white-label vs custom GPS platform.
If the numbers above point at building, we will tell you so rather than talk you out of it — and we build custom platforms too. Talk it through or message us on WhatsApp.
To a first production release covering a handful of device protocols and core tracking, plan in the region of a year with a small dedicated team. To match the breadth of an established platform — 1,000+ device models, dozens of modules, mobile apps, billing — considerably longer, because that breadth is accumulated over years rather than built in a sprint.
Three things: ongoing protocol support as new devices and firmware ship, the operational cost of someone being responsible for uptime outside office hours, and the architectural rework that arrives when volume outgrows the original design. None appear in a development estimate; all of them recur.
The calculator gives you the break-even device count for your own assumptions. Below it, licensing wins on cost alone. Above it the arithmetic favours building — but only if you can actually staff and operate a platform team, which is a separate question from whether the spreadsheet says yes.
Yes, and it is usually the right one: license the tracking platform and build only the differentiating layer on top of its API. You own the part customers pay for, and device coverage, uptime and protocol maintenance stay someone else's responsibility.
We do, so judge the inputs rather than our conclusion. Every assumption is editable and the formula is stated. Set the retained engineers to zero and infrastructure to nothing and building will win; we have left those inputs honest rather than rigged, because a calculator that only ever produces one answer is worth nothing to you and eventually to us.
Send us what you put in. We will tell you where the assumptions are optimistic, including when that argues against buying from us.
Discuss Your Numbers →