The short version
If our office had to buy electrical gear again tomorrow, I'd do three things:
- Run the TCO math before I look at a single price. A cheap UPS sitting in the storage room is not a deal — it's waste.
- Confirm whether the network card is included, and write the APC UPS default IP into the purchase request. That way IT can connect on day one instead of day three.
- Never touch a voltage tester without a three-point known-state check. Thirty seconds, worth every second.
Why I'm writing this down
I'm an office administrator for a 150-person B2B services company. I manage all office and facilities purchasing — roughly $180K annually across 8-9 vendors. I report to both operations and finance.
I took over purchasing in 2020. In 2023, we ran a small infrastructure refresh — replaced three aging UPS units, added a few new ones, and restocked the facilities team's tool kit. That refresh basically rewired how I think about "cheap" equipment. It also taught me why the old "IT picks it, admin orders it" split has a hole in it (more on that below).
The UPS purchase that made me re-read the spec sheet
The old units came from a distributor that no longer exists. We ordered three APC machines — two Back-UPS picks plus a Smart-UPS tower, model SMT1500I.
The SMT1500I spec sheet reads: 1500VA / 1000W, eight IEC C13 outlets, tower-to-rack conversion, LCD display, and a SmartSlot expansion bay. You can verify all of that on the product page at apc.com — the model number matches.
What the spec sheet didn't make obvious to me (and what I didn't think to ask) was this: SmartSlot is just a slot. The network management card is a separate purchase. I assumed "Smart" meant a network card was already inside.
That assumption surfaced the day the units were racked and IT asked, "What's the default IP?"
For what it's worth, the factory default IP on an APC Network Management Card is typically 192.168.1.100, subnet mask 255.255.255.0, with DHCP disabled out of the box. But that varies by card model and firmware version — the right move is to pull the manual for your specific SKU from schneider-electric.com. In our case we didn't have a card at all, so the real question became: do we add one?
Looking back, I should have asked about the network card before the purchase request went out. At the time, our old workflow was simple — pick the box, order the box — and "does the box have a management card" wasn't a question I knew to ask. Now it's a line item on every UPS request I write:
"SmartSlot network card included? If yes, which model? Default IP and initial credentials to be confirmed and documented by IT at receipt."
One line. It would have saved us three phone calls and about four hours of back-and-forth. The calls weren't expensive. The hours were — plus a few bemused looks from the IT lead who kept asking me, the admin, what his network card's IP was.
The old "IT picks it, admin orders it" split has a gap
For years our workflow was: IT selects the model and compatibility, admin places the order. That made sense back when buying a UPS really did mean buying a box of batteries with a plug. But from 2023 onward I've noticed the gap — IT only checks spec and compatibility. Nobody owns the battery replacement cycle, the install cost, or the rework risk. Those costs land on the purchasing side eventually, whether anyone planned for them or not.
So I started owning that part myself, and that meant learning enough of the vocabulary to ask IT smarter questions. Which is how I ended up staring at a VFD symbol on a facilities drawing a few weeks later.
Why this kind of purchase needs TCO thinking
The whole refresh — three UPS units plus the tool kit — came out to roughly $4,800. If I'd compared on sticker price alone, I would have missed at least three costs:
- Battery replacement cycle. UPS lead-acid batteries typically need replacing every 3-5 years. Spread that per-unit cost across the cycle and you get the real annual holding cost.
- Install and cutover. Weight, lifting, cable routing, transfer testing. None of that shows up on the invoice.
- Rework risk. Wrong model, wrong outlet count, wrong rack depth. When that happens, the fix usually costs more than the original unit.
My working definition of TCO is four things: unit price, hidden costs, time cost, rework risk. If you're buying a UPS for your office, put those four lines on paper before comparing quotes. It's more useful than shaving another $30 off the invoice. We've been burned more than once by a supplier whose unit price was the lowest in the room and whose total cost was the highest.
The other thing: a voltage tester, and a symbol I didn't recognize
The facilities team also got a Milwaukee voltage tester from that same order — one of the non-contact models. Our facilities lead walked me through his process on the spot, and I wrote it down because I think admin buyers should know this, even if we never touch a panel ourselves:
- Test a known-live outlet. Confirm the reading is normal — ours run in the 115-125V range on a 120V system.
- Test a known-dead outlet. Confirm the reading is essentially zero.
- Only now test the actual point you care about.
He calls it the "three knowns" check. I don't know if that's the official industry name, but the logic holds up regardless: calibrate the tool against a known live state and a known dead state before you trust it with an unknown. People who skip that step tend to have one memorable near-miss and then never skip it again.
During the same walkthrough, he pointed at a symbol on a facilities drawing and said, "That's the VFD." I stared at it for a good 30 seconds before asking what a VFD was. Variable frequency drive — used for controlling motor speed. The symbol is usually drawn as a circle containing a sine wave plus a thyristor or transistor notation; the exact rendering differs between ANSI and IEC drafting standards. I felt a little ridiculous, an admin buyer suddenly reading motor symbols. But if I'm going to sit in budget review meetings and justify why the facilities line item went up this quarter, understanding the drawing is the lowest-effort place to start.
Where this stops applying
All of the above comes from a 150-person, three-floor, office-based B2B services company with fairly stable load. If you're in a different situation, the calculus changes:
- Manufacturing. A UPS backing a production line instead of IT load is a completely different reliability problem.
- Seasonal operations. Demand spikes change your redundancy math — you can't just size to average load.
- Multi-site offices. The network card's default IP strategy needs central planning, not a one-off purchase request per location.
I can only speak to our setup. If your conditions look different, there are probably factors I haven't seen. But the two habits — putting TCO and the safety check into the purchase request itself — have held up across every variant we've tried so far. Those I'd recommend regardless of company size.

