The surface problem: the spreadsheet stops at price and torque
Every week, someone sends me a spreadsheet comparing drive-unit options. Column B is price. Column C is torque. That's it. On paper, they're buying a motor. In reality, they're buying an integrated e-bike drive system—and the system includes the battery, speed sensor, display, firmware, wiring harness, mounting hardware, and the engineering time to make it all work together. The spreadsheet almost never includes those.
I'm not a procurement manager. I'm the person they call when the spreadsheet fails. I coordinate rush orders for an e-bike drivetrain supplier. I've handled 200+ deadline-sensitive orders in six years, including same-day turnarounds for OEMs with stopped assembly lines. When I'm triaging a rush order, I rarely start with torque. I start with time.
Actually, let's start with something that sounds like a different topic: servo motors.
The deeper problem: component thinking in a system world
If you're buying a servo motor for a machine, the torque rating is important. But anyone who has specified a servo system knows the motor is the smallest part of the decision. You also need the drive, the controller, the feedback encoder, tuned gains, cable lengths, cooling, and an emergency-stop strategy. A servo motor quote without that context is a paperweight.
An e-bike drive unit is no different. The Shimano STEPS E5000 torque number—40 Nm, in case you're in the middle of comparing that spec—tells you what the motor is capable of. It doesn't tell you:
- whether the firmware matches the bike's wheel size and cadence targets
- whether the speed sensor is included and correctly installed
- whether the battery's discharge curve aligns with the drive unit's demand
- what changes when you swap from a 26-inch wheel to a 27.5-inch wheel
- whether the supplier has the Shimano diagnostics tools and training
That's the deep problem. People aren't comparing motors. They're comparing an integrated system against a component price. And until someone understands that, the price quote is fiction.
I learned this the hard way. A customer sent me a spec for their new city e-bike. They had selected a Shimano STEPS drive unit but went with a 'compatible' harness to save money. I assumed 'same specifications' meant same performance. Didn't verify. Turned out the compatible harness used a different pin-out, and the assist cut out after 30 seconds of riding. We had to rewire the whole prototype. The 'savings' on that harness was about $18 per bike. The engineering time to discover it? Many times that.
The real cost of component-level thinking
In March 2024, an OEM in the Pacific Northwest called at 4:37 PM. Their assembly line was supposed to ship a batch of cargo e-bikes in 36 hours. A supplier had sent them a drive unit that looked right, but the speed sensor wasn't in the box. The original order was short by seven sensors—they'd skipped them to reduce the unit cost. The supplier's standard delivery was five days. We found the sensors, paid $160 for overnight courier, and got them to the line in time. Save: $21 per bike. Extra cost: $160 in freight and about four hours of panic.
Not ideal. Workable. But it was a lucky ending compared to what I see in rush orders.
The bigger surprise wasn't the missed sensor though. It was the hidden value that came with the 'expensive' option—support, revisions, quality guarantees. An OEM called us last quarter because their test bench motor kept tripping overload. The question was, 'what size VFD for 5hp motor?' A quick answer would be '5hp.' But on a load bench, the VFD has to handle continuous torque at low speed and regenerative braking. The 5hp VFD they bought was undersized for the actual duty cycle. Replacement cost plus line delay was over $4,000. The extra $200 they would've spent on a correctly sized drive would have been cheap.
You'd think disc brake components are a different conversation. They're not. On another build, a client's brake caliper wouldn't fit the fork because the rotor was 180mm and the adapter was for 160mm. Or maybe it was the other way—it doesn't matter. The point is both components are on the same bike, and someone treated them as independent line items. That mismatch cost two days and a courier run from a local supplier. The part itself was $30. The total cost of the mismatch was about $700.
Why does this keep happening? Because the surface problem looks like a price problem. The deeper problem is that a bill of materials is not a design. If you compare components line by line, you'll always choose the cheaper option that is just different enough to not work.
The price of ignoring TCO
Total cost of ownership is not a buzzword. It's the only framework that explains why the $500 quote becomes $1,100 after shipping, setup, revisions, and missed timelines—while the $650 quote that includes support actually costs less.
In my work, I see three hidden costs repeat:
- Line downtime. An assembly line has a cost per hour. Every minute spent waiting for a missing sensor is money. The cheapest component that arrives late is more expensive than the premium one that arrives on time.
- Engineering and rework time. Every 'compatible' part that isn't exactly compatible costs an hour of someone's time. That person is a bike engineer, not a spreadsheet. Their time is not in the unit price.
- Emergency logistics. Speeding a parcel from one warehouse to another is expensive. It's also avoidable if the original order included the complete system.
There's also the risk cost. According to UL 2849, the safety standard for e-bike electrical systems, and EN 15194, the European standard for electrically power assisted cycles, the certification is about the whole system—not just the motor. A drive unit that's certified to those standards is a compliance anchor. But certification only protects you if the rest of the bike is specced correctly. (note to self: always ask for the full system drawing before promising a delivery date.)
Don't just take a supplier's word for reliability, either. Per FTC guidelines (ftc.gov), marketing claims need to be truthful and substantiated. If a quote promises 'zero failures' or 'perfect integration,' ask for the evidence. In my experience, that question alone filters out the weakest vendors.
What TCO actually looks like
Let's make this concrete. Suppose you're comparing two drive-unit quotes:
- Quote A: $550 per drive unit, no speed sensor, no technical support, 10-day lead time.
- Quote B: $650 per drive unit, includes speed sensor, firmware matching, and a local person who answers the phone.
The obvious calculation is $100 per unit. The real calculation includes the missing sensor ($21), the firmware matching fee ($40), the potential for a stopped line ($200/hour), and emergency freight ($150–400).
Quote A is not cheaper. It's just easier to see.
This is why I now calculate TCO before comparing any vendor quotes. I don't even look at the unit price first. I look at the set of assumptions behind the quote: what's included, what's excluded, who's responsible if the drive unit and battery don't communicate, and what happens when the line stops.
The same logic applies to every component in the e-bike system. Servo motors need drives and tuning. Disc brake components need adapters and alignment. A 5hp motor needs a VFD that's sized for the duty cycle, not just the nameplate. And a Shimano STEPS system needs the full ecosystem: Shimano Steps E5000/6000/7000/EP8 drive units, the proper sensors, batteries, and firmware.
The solution isn't complicated. It's just not in the first column of the spreadsheet. Build a TCO scorecard. Include lead times, support response times, and freight costs. Ask the supplier to identify the integration points. And if you're in a hurry, call someone who has done this before—because there's no substitute for knowing what the price quote isn't telling you.
That's the part that keeps me busy. And honestly, it's the part that pays for itself.