Most project managers can price the part of an IoT deployment they can hold in their hands. A temperature sensor is a known quantity. A LoRaWAN gateway has a line on a vendor quote. You add up the bill of materials, pad it for spares, and you have a hardware number that survives a budget review. That number feels like the project. But the hardware is the cheapest thing you will buy, and it is usually the only line PMs get right. The money that decides whether a deployment ships on time and stays alive after launch lives in connectivity contracts, platform fees, integration hours, and the operations bill that starts the day you go live and never stops. This guide breaks a mid-market budget into the categories that actually consume it, shows where the spend really lands versus where teams expect it, and is honest about the cases where TagoIO is the wrong cost choice.
Hardware is the line everyone gets right and overweights
Sensors and devices run anywhere from $10 to $500 each depending on what you are measuring, and gateways or edge units land between $200 and $5,000 per box. For a mid-market pilot of a few hundred to a few thousand nodes, you can put a defensible hardware number on paper in an afternoon. Industry breakdowns put hardware at roughly 30 to 40 percent of the initial spend, and that share keeps shrinking as sensor prices fall.
The trap is that hardware feels like the whole project because it is visible. A box arrives, you can mount it, you can point at it in a meeting. So budgets get anchored to it. A team that spends three weeks negotiating sensor unit price and twenty minutes on the connectivity contract has the priorities backward. Buy good hardware, buy spares, then stop optimizing it and move your attention to the lines that grow.
Connectivity is small per device and large at scale
Connectivity is where the per-unit number looks harmless and the total surprises you. Per device, the range is wide: low-power NB-IoT or LTE-M can come in under a few dollars a year, cellular data plans run roughly $0.20 to $5 per device per month, and higher-data 5G nodes push past $10 a month. LoRaWAN on unlicensed spectrum can carry no recurring per-device fee at all if you own the gateways, which is exactly why it wins for dense, low-bandwidth fleets.
The category matters because it is recurring and it scales with your fleet. A thousand cellular devices at $3 a month is $36,000 a year that did not exist in your hardware quote, and it repeats every year you operate. Pick the radio technology against your data rate and battery needs, not against the cheapest module, because the radio choice sets a recurring bill you carry for the life of the deployment.
Platform pricing: per device or tiered, and both can bite
Managed IoT platforms generally price between $1 and $5 per device per month, or move you onto an enterprise agreement once volumes climb. Raw cloud primitives look almost free per unit. AWS IoT Core, for example, charges fractions of a cent per device per year for the connection and about a dollar per million messages. The catch with the cheap-per-unit cloud option is that you are buying plumbing, not a product, and you will build dashboards, user management, alerting, and access control yourself.
For a mid-market fleet, platform cost is real but rarely the line that breaks the budget. The decision that drives total cost is not the sticker price per device. It is how much you build on top of the platform versus how much comes included. That is the integration line, and it is the one teams underestimate most.
Integration and development is usually the biggest line
This is where the money actually goes. Custom IoT application work runs roughly $75 to $200 per hour, full software builds start around $50,000 and climb quickly, and connecting a new deployment to the systems you already run, the ERP, the maintenance scheduler, the billing system, is the hidden line that eats unbudgeted weeks. The visible work of getting a device to send a reading is small. The work of turning that reading into something a person acts on, routed to the right system, with the right permissions, in a portal a customer trusts, is most of the project.
If you take one number from this guide, take this one: integration and development is frequently the single largest category in a mid-market deployment, and it is the line most often left off the first budget. Plan for it explicitly, scope it against real systems you have to touch, and pad it, because legacy integration always finds work nobody quoted.
Ongoing operations is the bill that never ends
Launch is not the finish line, it is the moment the recurring bill turns on. Annual maintenance for the software side of an IoT deployment typically runs 15 to 25 percent of the initial development cost, every year. On a $200,000 build that is $30,000 to $50,000 a year for security patches, platform updates, firmware pushes over the air, connectivity management, and the data storage that grows as your fleet talks. IoT carries a physical dimension that pure software does not: devices in the field fail, batteries die, and hardware needs eventual refresh.
Budget operations as a line from day one. A deployment costed as a one-time capital project and handed off with no operating budget is a deployment that goes dark within two years.
When TagoIO is the wrong cost fit
Honesty matters more than a sale here. If you have real engineering capacity and you want to trade cash cost for labor, a self-hosted open-source platform like ThingsBoard Community Edition can be the right call. You run it on your own servers, you pay no per-device platform fee, and you accept that your team owns uptime, upgrades, and scaling. That is a genuine cost win when staff time is cheaper for you than recurring SaaS spend and you have the people to carry it.
Likewise, if your team already lives deep inside AWS and has the cloud engineers to assemble the pieces, raw AWS IoT Core plus the surrounding services can beat a managed platform on per-unit price. You are buying primitives and building the product yourself, which suits a team that wants total control and has the hours to spend.
And to be straight about our own line: building a polished, white-labeled customer portal on TagoIO with TagoRUN can run more development hours than wiring up a commodity dashboard, because you are shipping a branded product, not a generic admin view. If all you need is an internal chart of a few sensors, a managed platform built for white-label delivery is more than the job requires, and you will feel that in the hours.
Where TagoIO fits
TagoIO earns its cost when integration and time-to-launch are the lines that hurt, which for most mid-market teams they are. Serverless Analysis scripts let you process and route data without standing up your own backend, which pulls hours out of the integration line that usually dominates the budget. With 500+ device integrations, the connect-the-radio work that eats early weeks gets shorter. TagoRUN gives you white-label portals so a customer-facing product is a configuration job rather than a from-scratch build, and TagoCore covers open-source edge when you want processing closer to the device. For teams that need a defensible compliance posture, TagoIO is ISO 27001 certified and GDPR-aligned, and the multi-tenant model fits deployments serving multiple customers from one stack. The trade is clear: you pay platform fees so your engineers spend their hours on what makes the deployment yours, not on rebuilding plumbing.
Next steps
Price the platform line against your real device count at tago.io/pricing. See how teams in your sector scoped similar deployments at tago.io/use-cases. And if integration is your biggest line, a delivery partner can compress those hours, start at tago.io/partners. Budget all five categories, not just the one you can hold, and the deployment that ships will be the one that is still running two years from now.
