Our first thirty
Saturdays.
Thirty Saturday rushes. Roughly ten months of Saturday-only nights at the Cantina. Around 28,000 plates served. Here is the honest accounting — the math, what worked, what broke, what surprised us, and what we’re looking at next.
All figures are operator estimates rounded to honest precision — the cambro side of the kitchen, not the CPA side. Variance noted in §01.
Thirty Saturdays. Around 28,000 plates.
The Cantina opened the Saturday-only late-night format roughly ten months ago. We ran it every Saturday since. Thirty Saturdays in, which is the only honest place to stop and look backwards — any earlier and you’re reading noise, any later and the story’s already a different story.
Here is where the 28,000 plates comes from. Our internal POS ticket count across the thirty Saturdays lands at 27,840 entrées plus a long tail of sides and add-ons. We don’t count a side of slaw as a plate. We count brisket, ribs, pulled pork, turkey, the burnt-end mac bowl, the Friday taco carry-over (we run a Saturday taco slot too, late), and the sandwich list. Round that to 28,000 and you’re honest. Round it to 30,000 and you’re selling.
The $1.2M gross is plate-count times the average ticket ($42 on a Saturday — Saturday tickets run higher than the weekday average because the room turns toward platter orders and the bar pulls harder), minus the four sellouts where we served fewer than the average. The actual POS gross across the thirty came in at $1.18M. We rounded to $1.2M the same way we rounded the plates. The variance is real and small.
Four sellouts in thirty Saturdays. That’s one every 7.5 weeks — call it once a month-ish during the busiest stretch and zero during the slower months. We treat sellouts the way the trailer year treated them — as good problems. Every sellout is a Saturday we put more brisket on the schedule the next month. There’s a version of this business where sellouts are a brand signal you chase. We don’t. Sellouts are how we calibrate the next month’s prep, not a marketing event.
The other number worth saying out loud: the Saturday-only format is the highest-margin night in the program. Saturday rent is the same as Tuesday rent; Saturday payroll runs about 2.3x a Tuesday; Saturday gross runs about 5.4x. The arithmetic is the reason we built the late-night Saturday format in the first place. It’s also the reason any thirty-Saturday math overstates how rich the rest of the week looks. Tuesday through Thursday is where the floor lives.
The brewery model and the mug club.
The brewery-truck model carried us through the thirty Saturdays. The Cantina is the loud number on the marquee, but three brewery trucks (no rent, roughly 12% gross to the brewery, four staff per truck against eighteen at the flagship on a Saturday) are what kept the slow Tuesday and Wednesday months from punching a hole in cash flow. The full business case lives at why we park at breweries. Short version: the brewery already has the foot traffic; we bring the program. Neither side compromises on the thing they care about.
The mug club — somewhere around 900 members today, paying $99/yr for the numbered ceramic mug and the four-location perk grid — provides the predictable floor under Tuesday through Thursday at the Cantina. Mug-club members are not the people who eat with us once a quarter on a date night. They’re the people who walk in on a Wednesday because the mug is on the rack with their number on it. That behavior runs the slow nights. The catalog of perks is at /mug-club.
The thirty-Saturday format itself worked. There was an early version of the conversation where we considered Friday-and- Saturday or even Wednesday-through-Saturday late-night. We kept it to Saturday because one big night is easier to staff than two mid-sized ones, and because keeping the format scarce is what makes it the night people drive across town for. The first ten Saturdays we second-guessed the decision. By month three we stopped.
Two operational things that worked better than we expected: the cambro discipline (an extra hour of rest beats a fast slice every time — the over-rest costs less than an under-rest), and the four-station floor layout (brisket slicer, sandwich plate, sides, bar handoff — keeps the ticket time under eight minutes during the rush even when the line at the door is forty deep). We’ll write the station math up properly when we have more than thirty data points.
Labor on the brisket. Catering pricing.
The 14-hour brisket cycle eats more labor than we projected. We modeled the brisket program at $7/lb in labor cost — the overnight pitmaster shift plus the slice-and-cambro at lunch, spread across the average yield per packer. Actuals across the thirty Saturdays came in at $9.20/lb. That’s a 31% miss. The honest reason is two things stacked: we under- estimated the slice-and-trim labor (a properly bear-clawed point cut eats time we didn’t book in the model), and we over-estimated yield per packer in winter (cold ambient, longer stalls, more trim — covered in part in the 12-hour brisket timeline).
We’re not changing the brisket. We are changing the model. The next thirty Saturdays’ pricing reflects an honest $9/lb labor floor on brisket, which is the number we actually pay, and the plate price has moved up $1 to absorb roughly two-thirds of it. The Saturday taco slot — built off brisket trim — absorbs the rest. We don’t make this public on the menu page because nobody wants to read a footnote about labor cost on a brisket plate. But we’ll say it here.
The catering side was under-priced for the first four months. Honestly. We were quoting off the pre-Cantina trailer-only pricing model — which assumed one cook, one truck, and a four-hour pop-up. The reality of a 200-person wedding rig at the Cantina is a separate prep day, a separate packer cook on a separate offset, and four staff hauling cambros at 4pm on a Friday. We were quoting that at a margin that worked for the trailer and lost money at the flagship. We fixed it in March. The calculator at /catering reflects the new floor — every quote since March has been at an honest margin, and the four months of under-priced catering before that we’re writing off as the cost of learning the catering side is its own business.
The Wednesday floor is the third broken thing. Wednesday is still soft. Mug club holds Tuesday and Thursday up; Wednesday hasn’t found its draw yet. The sourdough-night experiment is one bet on fixing it — sourdough loaves baked overnight in the brisket smoker, Wednesday-Thursday only, paired with smoked-tomato bisque. We’re piloting in July. Other shapes (a trivia residency, a slow-cook tasting night) are on the whiteboard. Wednesday will get solved. It isn’t solved yet.
Texans, Pueblo drivers, and podcast traffic.
The biggest surprise of the thirty Saturdays is who’s in the mug club. We assumed Colorado Springs locals, four-times-a-month regulars, the people who actually walk into the Cantina on a Tuesday. Most of the roster is exactly that. But around month three we noticed a cluster of out-of-state signups — about forty memberships shipping to Texas addresses, mostly Austin and Houston, almost all within a six-week window. We asked. The answer: a Texas brewery-industry forum had passed our name around because of the brewery-partner model. Texans love the idea of a chef who parks at breweries (Texas brewery scene has its own version of this argument; ours is apparently the cleanest shape they’d seen). Most of those forty members will never eat at the Cantina. They’re paying $99 because they want to read the model. We’re fine with that.
The second surprise: Pueblo customers driving up for Friday tacos. Pueblo is forty minutes south of us. There’s no shortage of food in Pueblo. But the Friday taco slot — small batch, brisket trim, salsa rojo, sold out by 1pm — has pulled a real Pueblo regular crowd. Three different tables in the last month told the front of house they drove up for the tacos. That changes how we think about the Pueblo market. (More on that in §05.)
The third surprise: podcast and YouTube traffic is now out-pacing local press for driving customers in the door. We’ve had a couple of local-paper writeups in the thirty-Saturday window — good ones, accurate, well-meant — and they brought a measurable bump. Each podcast feature brought more, and the YouTube short that the Saturday rush walkthrough generated (three minutes, no narration, just a packer brisket coming off the offset at 7am) drove more first-time customers through the door than any single press piece. We are still figuring out what to do with that. The honest read is: long-form audio and short-form video work for this category; print works for credibility, not for traffic.
Sourdough night, a Pueblo conversation, a cross-training rotation.
Three things on the next-thirty horizon, in honest order of certainty.
First: the sourdough-night experiment starts the first Wednesday in July. Eight-week pilot. Sourdough loaves baked overnight in the brisket smoker, Wed/Thu only, $9/loaf, paired with smoked-tomato bisque. The hypothesis is direct: the smoker is hot anyway, the bisque uses brisket-trim stock, and the loaf is shelf-stable enough that Thursday’s bake feeds Friday’s sandwich menu if Wednesday doesn’t move it. Worst case we eat the loss for eight weeks and learn the loaf doesn’t pull a Wednesday floor. Best case we have a new midweek draw. We’ll write the pilot up at week eight either way.
Second: we are exploring a Pueblo truck partnership. I want to be very clear about the hedge here — there is no contract signed, there are no announcement plans, and please do not call the brewery. We have had two conversations. Both went well. Pueblo’s a real market, our Friday tacos already pull from there, and the brewery-partner model is the easiest shape to extend without a new flagship lease. If this lands it’s a fourth truck and the first stage stop outside Colorado Springs. If it doesn’t, we’ll keep looking. Either way we are not naming the brewery yet.
Third: the pit team rotates through Front Range BBQ for cross-training. This is the one that needs a paragraph of honest framing. Front Range BBQ is a Colorado Springs operator that smokes a different style than we do — more hickory, sweeter rub, sauce-forward. We are not converging on their style and they are not converging on ours. We are sending our pit team to spend shifts on their line because a pitmaster who’s only ever cooked one program is a pitmaster who hasn’t pressure-tested their own instincts. Front Range agreed to host. We’ll host theirs back when they want to come. This is not a merger, an acquisition, a content collab, or any of the things the local-restaurant rumor mill will probably tell you it is in the next six weeks. It’s cross-training. That’s the whole shape.
Beyond those three, the longer arc is the same as the one we wrote up in from trailer to flagship: more brewery partners where the cadence fits, a possible second flagship up the Front Range when we find the building, catering becoming its own vertical inside the business. None of that is news in the next thirty Saturdays. It’s the horizon, not the calendar.
To the people who actually made the thirty Saturdays.
Thirty Saturdays is a number with my name attached to it in the press, and a number with seventeen other people’s names attached to it on the schedule. To the pit crew — the overnight pitmaster who started a Friday 8pm shift thirty times in a row, the slice line that bear-clawed pork through a forty-deep door line without losing the bark, the sides and bar staff who carried the front of house when the kitchen window was three tickets behind — none of the numbers in §01 happen without you. To our brewery partners, who keep letting a smoker park on their side lot and trust that we’ll feed their drinkers without burning down their parking lot — thank you for the handshakes that never became contracts. To the customers who waited an hour in line in February (it was cold; we know), who drove up from Pueblo for a Friday taco, who signed up for the mug club from a Texas address because they liked the model — you are why we cook. We are thirty Saturdays in. We are not done.
See you on Saturday thirty-one.
The model, or the mug club.
The brewery-partner business case lives at /brewery-partners. The membership grid that holds Tuesday through Thursday lives at /mug-club.