The email arrived at 2:47 PM on a Tuesday. Subject line: Dimplex oil radiator switch feels loose. Not exactly the kind of message you want to open right after lunch.
I'm the quality and brand compliance manager at Dimplex. I review every heater, fan, and dehumidifier that goes out the door—roughly 200+ unique units per year. I've been doing this for over four years, and I've rejected my share of first deliveries. But this one felt different.
A customer who'd ordered 500 units of our Dimplex oil radiator wrote to say the control switch on one sample had too much play. "It's not broken," they said. "But it doesn't feel right compared to the old units we've been using for years."
My first instinct was to dismiss it. A little wiggle in a rotary switch? That's within tolerance. The heater functions perfectly. But something about their tone made me pause. So I walked over to the testing bench, pulled a unit from the latest batch, and turned the switch.
Sure enough—it wobbled. Not dangerously, but noticeably. Compared to our previous batch, the detent felt softer, the shaft had maybe a millimeter of lateral play. I checked our spec sheet: we had a torque requirement for the switch, but nothing that measured side-to-side motion.
We called the supplier. They'd quietly substituted a slightly different switch component to reduce cost. It met all the functional specs we'd listed, but nobody had thought about feel.
Here's where the conventional wisdom let me down. Everything I'd read about quality assurance said that as long as the product passes functional testing, cosmetics and tactile feedback are secondary. In practice, for our B2B customers who install dozens of these units, a loose switch reads as "cheap"—even when it isn't. It undermines confidence in the whole unit.
The customer also mentioned they were comparing our radiators with a propane heater and a Dewalt fan for their workshop expansion. "We're not looking for the cheapest option," they wrote. "We just need stuff that holds up." Those two products—both from completely different categories—were setting their quality expectations.
And then came the kicker. They asked, almost as a side note: "Also, any idea how to unlock Honeywell thermostat? The one we bought for the office has a keypad lock we can't bypass."
Now, Honeywell isn't our brand. But the customer didn't care. They saw us as heating and cooling people, so they assumed we'd know. I'm not 100% sure about the exact Honeywell model, but I knew the general issue: many programmable thermostats have a keypad lock feature, and the unlock sequence usually involves holding the menu button and entering a 4-digit code. I walked them through it over the phone—they'd accidentally enabled the lock while trying to set a schedule. It wasn't a defect. Just a newer feature that older users didn't expect.
That moment was a turning point for me. Here's the thing: what was best practice in 2020 may not apply in 2025. The fundamentals of heating technology haven't changed—you still need a resistance element and a thermostat. But the execution has transformed. Switches are no longer just on/off toggles; they're part of a user interface. Thermostats now have locking mechanisms and smart schedules. Customers expect consistency across every touchpoint, even when comparing our heater to a random fan.
So we did something about it. I wrote a new spec for the switch: a side-load test, a minimum detent force, and a visual alignment check. We sent it to the supplier and told them they could submit the cheaper part again, but only if it passed the new requirements. They redid the tooling. We also added a simple "user experience" checklist to our final inspection—things like tactile feedback, labeling legibility, and whether a first-time user could figure out the controls without reading a manual.
That cost us some money and a few weeks of delay. But it was worth it. The customer approved the revised units, and their follow-up orders have grown. More importantly, we've seen fewer service calls about "confusing controls" in general.
To be fair, this approach worked for us because we're a mid-sized manufacturer with direct relationships to our B2B buyers. If you're dealing with high-volume retail or disposable products, the calculus might be different. Your mileage may vary if your customers prioritize bare-minimum cost over every other factor.
I've also stopped assuming that a component substitution is fine just because it passes functional specs. People assume that the lowest-cost supplier is the one who cuts corners. The reality is more subtle: suppliers often make changes to improve their own margins, and unless your spec is precise, they'll meet the letter without understanding the intent.
And about that USB-style framing? It's like the USPS envelope standards—they define exact dimensions for letters because if everyone just used "about right," the whole system would jam. Same logic applies to a heater switch. If you don't define what "right" feels like, you'll get "close enough"—and "close enough" rarely feels good to a paying customer.
We also took a look at our marketing materials after this. Per FTC guidelines, any claim we make about our products needs to be truthful and substantiated. But "truthful" doesn't mean "complete." We now audit our product pages to make sure images show the actual control panel and switch position—not a stylized version that looks different from the real unit. Because a customer who finds a different-looking switch in the box will think the product is defective, even if it works perfectly.
In the end, that one email taught me more than any audit report could. Industry evolution isn't just about new technology—it's about shifting expectations. Ten years ago, nobody would've compared a heater's switch to a fan's. Now they do, because both are just "products I bought online" showing up at the same time.
So here's my advice, and I'm saying it to myself too: don't assume the old standards still make sense. Test the product the way your customers will experience it, not just the way the test stand was designed. And if someone asks you about a Honeywell thermostat, just help them. It's not your product, but it's their trust.
That's the lesson. It cost us a rework and taught us a lesson. A lesson learned the hard way, but one we needed.