Most diagnostic connectivity products are software. Cloud platforms, integration engines, results middleware. They take input from device software clients (sometimes provided, sometimes not), normalize it, and route it to wherever the customer needs it. The model works in well-controlled environments — a hospital network with reliable infrastructure, a lab with clean network conditions, a region where someone else handles the connectivity hardware.
It doesn’t survive contact with country-scale diagnostic networks. Here’s why.
1. Software-only middleware assumes a network it doesn’t control
The first thing to break, almost always, is the assumption that the device can reach the cloud reliably. Hospitals have IT departments that handle this. National programs running POC devices across rural facilities don’t. The network is whatever’s available — a 3G signal that comes and goes, a Wi-Fi access point shared with the clinic’s unrelated systems, a router someone plugged in three years ago that nobody’s looked at since. Pure SaaS middleware doesn’t manage any of that. It expects it to work.
When it doesn’t work — and it doesn’t, regularly — the customer either builds their own connectivity layer or accepts that significant portions of their data are missing. Both outcomes are common. Neither is what the customer thought they were buying.
2. Crossing borders is harder than vendors admit
A country boundary is a substantive engineering event. Different mobile carriers, different roaming agreements, different telecom regulations, different data residency rules, different national health information systems with different protocols and different procurement frameworks. Pure software middleware that runs on the assumption “the data gets there” runs into all of these issues, all at once, the moment it’s deployed in the second country.
Programs operating across multiple countries end up with one of two responses. They run separate implementations per country — multiplying their operational burden — or they accept that the multi-country dashboard their vendor sold them only really works for the home market. Either way, the multi-country promise isn’t quite what it looked like in the demo.
3. Software can’t push firmware
In any sufficiently long-lived diagnostic device deployment, firmware updates are not optional. New protocols, security patches, calibration adjustments, fixes for issues that surfaced in the field. In a clinic-scale environment, an engineer drives to the device and updates it. In a country-scale environment, that engineer drives to seven hundred devices across hundreds of facilities, multiple times a year. The cost is real, both in money and in keeping the fleet current.
Software-only middleware doesn’t help with this. Pushing firmware to a device requires owning — or at least standing alongside — the connectivity hardware between the device and the cloud. Vendors who’ve avoided that part of the stack avoid it because it’s harder. Their customers pay for that avoidance, in field-service costs, and in fleets that drift out of currency over time.
4. Operations is where it actually breaks
Even when the network holds, even when borders haven’t multiplied, even when firmware is current, the day-to-day reality of running thousands of distributed devices wears down software-only deployments. SIMs need replacing, devices need swapping, configurations drift, sites change names, operators rotate, consumables change form factors. Someone has to manage all of this, and that someone is usually the customer — because the vendor’s product only handles the part of the work that compiles cleanly.
The right unit of integration
Connectivity, at country scale, isn’t a piece of software. It’s hardware plus SIM plus cellular operations plus cloud platform plus integration engine plus field-services discipline — together, as one stack, owned by one party that has incentives aligned with the customer’s deployment actually working. Anything less is the customer absorbing the parts the vendor decided not to ship.
The teams running the most successful national POC programs all end up with the same realization eventually: the right partner is the one who shows up with the gateway, the SIM, the cloud, the integrations, and a team that owns operational uptime.
What this means for buyers
If you’re evaluating diagnostic connectivity products, the questions that matter are not the ones procurement frameworks tend to ask first. They are:
- Does the vendor ship the hardware that connects the device, or do you?
- Whose SIM is in the gateway? Who’s negotiating with carriers?
- Who pushes firmware? Who pays when devices fall out of currency?
- What happens when the network in country three is structurally different from country one?
- Who runs operations across the fleet — them or you?
If the answer to any of those is “you,” what you’re buying is a piece of software, not a connectivity solution. There are situations where that’s fine. Country-scale diagnostic networks aren’t one of them.
That’s the gap we built ConneX to close.
Reading this and recognizing your program? Talk to us