What we learned shipping BOH Wiz to 10 restaurants

I started BOH Wiz because I got tired of throwing out a $40 quart of pico every Friday morning that nobody used because the masking tape said "PCO 4/12" and nobody could remember if that was the prep date or the use-by date. The first version of the app printed labels. The version we ship now does seven things differently, and most of them came from operators telling us we were wrong.

Here's what we learned.

1The labels are the small problem.

We thought we were building a label printer. We were actually building a notification system. The single feature that moved the food-waste number in every kitchen we deployed in was not the printer or the labels. It was the day-before notification: "12 items expire tomorrow." Closing manager builds them into the special. Opening manager doesn't have to throw anything out. At one Chicago independent, this single change moved prep waste from roughly 12% of weekly food cost to around 8% in the first quarter. The labels were the means. The notification was the result.

2Every operator we talked to had been pitched a $10/month subscription system. Most had said no.

The pattern was identical across kitchens: a labeling vendor pitches $10/month, the operator does the math, signs, then quietly stops paying within 18 months because the printer always seemed to be out of labels and nobody on the line really used the app. The vendor's revenue model was perfectly aligned with the operator not using the system — they get paid either way. We made a deliberate choice to charge once for the printer ($299.99) and zero for the software, because our revenue can't depend on anyone forgetting to cancel.

If you want the math, we wrote it up: $299 once vs $144+/year forever.

3The line cook will not learn a new app. The manager will.

We over-built features for the prep cook in version one. Bulk import, custom shelf-life rules, allergen tags, multi-station printing modes. The prep cook ignored all of it and only used one button: Print. The manager actually wanted the analytics, the audit trail, the multi-location dashboard. So we split the product. The cook's view is one tap. The manager gets the dashboard at bohwiz.com/manager. Don't make the line cook care about the database.

4Founder-direct support is unsustainable. We're keeping it anyway.

The phone number on bohwiz.com rings to me, in Chicago, on my actual phone. Not a queue, not a tier-1 rep, not a chatbot. This is genuinely impractical — if we get to 100 customers it'll fall apart — but for the first 10, it's been the single most-cited thing customers mention when we ask why they stayed. The lesson isn't "founder-direct support scales forever." The lesson is that the early-customer experience is the marketing, and there's no shortcut.

5Multi-location operators do not want per-location billing. They want one bill.

The two multi-unit operators in our first ten told us the same thing within a week of onboarding: "If you charge me per location, I'm going back to handwritten." Our pricing model was already one-time, but the dashboard wasn't multi-tenant. We rebuilt the manager view in the next quarter so a single account can run two kitchens or twenty, with per-store data but one consolidated bill. We could have charged for it. We don't, because we won't make money the operator can't predict.

6The hardware decision is the durability decision.

We picked the Zebra ZD411 because Zebra direct-thermal printers shipped since 2015 last 5+ years in restaurant environments and Zebra parts are easy to source. A few competitors lock you to their proprietary printer; their printer goes EOL, you re-buy the hardware. The ZD411 also runs our 2 × 1 in. dissolvable label refills, which we make to a spec we tested in real cooler humidity (smudge-free at day five, dissolves in cold water in under 30 seconds, FDA-compliant for indirect food contact). The printer is open-format so technically other direct-thermal labels work too — but our refills are why most customers stick around once the printer is paired. Hardware lock-in is a short-term revenue strategy and a long-term churn driver. Label quality is the opposite.

7The mistake we made: we shipped Android too early.

The hardest lesson. We rushed Android to parity with iOS in the first year because the bigger operators we were courting are Android-first on shared kitchen devices. The early Android builds had two BLE pairing edge cases that crashed mid-print, which is the worst possible time for an app to crash. We've since pulled Android off the Play Store while we run the rebuild past more partner kitchens. iOS is live and stable; Android comes back when the partner kitchens sign off, not when our roadmap slide says.

The shorter version of this lesson: ship hardware-adjacent software when the device drivers are boring, not before.

What we're working on next

The current focus is widening the named-brand customer list, getting Android back to public release, and answering the same question every operator on a discovery call asks: "Will I still own the data if I stop paying you?" The answer is yes. We export your full label history to CSV, on demand, no cost, no friction. Hostage data is also a churn driver.

If you run a kitchen and any of this resonates, the printer kit is $299.99 with a 30-day return policy. The phone number is 866.626.4949 and rings to me. The email is hello@bohwiz.com.

— Jose Salazar, Chicago, May 2026