[Series] IoT Startup: Reset Your Expectations / Timeline

You’ve got the pitch deck ready. Your Excel spreadsheet is beautiful—colour-coded Gantt charts, milestones aligned perfectly with funding rounds, everything tracking toward your glorious market launch. Six months from concept to shipping product. Maybe eight if you’re being really conservative.

Eighteen months later, you’ll ship your first production units.

And honestly? You’ll get off easy.

If you’re reading this because you’re thinking about starting an IoT company, or you’ve just kicked one off and the timeline is already slipping, let me save you some pain: whatever timeline you’ve estimated, triple it. Then add 20%. You’ll probably still be optimistic.

The Seductive Math

The pitch is intoxicating in 2026. Hardware is cheap—ESP32s cost a few dollars, and there are even cheaper options if you’re willing to deal with less documentation. Cloud infrastructure is commoditised—AWS IoT Core, Azure IoT Hub, they’re practically plug-and-play. AI coding assistants can generate firmware faster than you can describe what you want. Manufacturing? Alibaba will connect you with a factory in 48 hours.

So the math seems simple:

• Month 1-2: Prototype and proof of concept
• Month 3-4: Firmware development and cloud backend
• Month 5: Testing and compliance prep
• Month 6: Manufacturing setup and first production run

Ship it. Ring the bell. Count the money.

Where This Falls Apart (Spoiler: Everywhere)

Here’s what will actually happen to you, and what I’ve seen happen to nearly every IoT startup I’ve watched go through this:

Months 1-3: The Honeymoon

Your first prototype works! Kind of. It connects to WiFi maybe 60% of the time. The sensors read… something. The data gets to the cloud when the stars align. But it works, and that’s enough for you to feel like a genius.

You start talking to potential customers. They love it. They also mention it needs to work with their existing system (which you’ve never heard of), survive -10°C to 50°C (your prototype is tested at “room temperature in your garage”), and oh, can it run on battery for a year?

Timeline impact: +2 months just understanding actual requirements.

Months 4-7: Hardware Reality

Your second prototype addresses the new requirements. Now it doesn’t work at all. Turns out adding a bigger battery creates a thermal problem. The WiFi antenna you copied from a reference design doesn’t work when enclosed in your case. The sensors you’d specced are discontinued.

You find new sensors. They use a different protocol. Your firmware needs a rewrite. The new antenna needs the PCB redesigned. Each iteration is 3-4 weeks waiting for boards to arrive from fabrication.

Timeline impact: +4 months of hardware iteration, and you’re not even close to done.

Months 8-11: The Compliance Nightmare

You finally have hardware that works reliably. Time for compliance testing! How hard can it be?

Very hard, as it turns out. Your first RF test fails spectacularly—you’re spraying harmonics everywhere. Back to the drawing board on EMI mitigation. Add ferrite beads, revise the ground plane, add shielding. Wait for new boards. Test again. Fail again, but differently.

Four attempts later, you pass pre-compliance testing at a local lab. Now for the actual certification body. They find new issues. The test setup is different. The requirements are interpreted differently.

Timeline impact: +5 months. Yes, really.

Months 12-15: Firmware Is Never Done

While hardware is iterating, your firmware development is ongoing. Except “ongoing” means “constantly rewritten to accommodate hardware changes.” That OTA update mechanism you built? Doesn’t work reliably over cellular. The power management code? Causes I2C bus lockups under certain conditions you didn’t test for.

Your security audit comes back with 23 issues. Memory optimisation takes three weeks because you’re 4KB over budget. The field trial units start failing after two weeks because of a race condition that only manifests under production load.

Timeline impact: This doesn’t add time, it runs parallel, but it pushes your “actually ready” date out continuously.

Months 16-18: Manufacturing Is Not Clicking ‘Order’

You have certified hardware. You have stable firmware. Time to manufacture!

The factory needs your BOM in their format. Three components aren’t available through their suppliers—you need to find alternatives or ship components to them. The assembly process has issues with your board layout (components too close together for their pick-and-place). First run has a 30% yield because the factory tech didn’t torque the enclosure correctly and cracked the PCBs.

Firmware flashing at scale needs a production jig you didn’t plan for. Quality control processes need to be defined and implemented. Packaging needs to be designed, prototyped, and manufactured.

Timeline impact: +3 months from “ready to manufacture” to “actually shipping.”

What You’ll Learn

The six-month timeline isn’t wildly unrealistic for any one of these components. Six months of pure firmware development? Doable. Six months of pure hardware iteration? Possible. Six months to navigate compliance? Tight but feasible.

The killer is that they don’t run in sequence—they run in parallel, and they all interfere with each other. Hardware changes break firmware. Compliance failures require hardware redesigns. Manufacturing realities force design compromises that need firmware updates.

And this assumes nothing goes really wrong—no major design pivots, no supplier bankruptcies, no key engineer departures. You might avoid the worst of component shortages, but even in 2026, component availability can still bite you when you least expect it.

For Your Timeline

If you’re planning an IoT startup in 2026:

• Budget at least 12-18 months from concept to first production units, even for “simple” products
• Double it if you have any regulatory requirements beyond basic RF compliance
• Add 50% if you’re doing anything battery-powered
• Add another 50% if you need IP68 rating or industrial temperature ranges
• Assume at least 3-4 hardware revisions, not 1-2
• Budget 2-3 months for compliance testing, with margin for failures
• Manufacturing setup is never less than 6-8 weeks

The good news? Every month of delay is a month you’re not supporting a half-baked product in the field. The customers who wait for your v1.1 will have a much better experience than the ones who got the “shipped on time” v1.0.

Have you been through an IoT product development cycle? What did your timeline look like versus the plan? Let me know in the comments—misery loves company.