TL;DR: Most spirits launch trainings fail because they teach the brand story instead of teaching the bartender how to sell. The rep presents a 40-slide deck, hands out branded koozies, and nothing changes at service. This is the framework that actually moves bottles.
The rep shows up on a Monday afternoon. Half the staff is there because they were guilted into it. The other half is texting. Someone pours small tastes around minute 45, after everyone has already checked out. The rep leaves behind a binder nobody reads and a branded bottle opener that goes home in someone's pocket.
Two weeks later, a guest asks what's good in mezcal. The bartender says, "We have a few." That's it. That's the full recommendation. The launch spend was $30,000 and the velocity didn't move.
I've watched this happen enough times to stop being surprised by it. The problem isn't brand reps or bar staff. The problem is that the training was designed to educate, not to enable. There is a difference, and it matters more than anything else in this article.
Education transfers knowledge. Enabling gives someone a tool they can use in the moment they need it. Behind the bar on a crowded Friday, your staff does not need to know the full history of your brand's family estate. They need three sentences they can say with confidence when someone asks what to drink.
Build for the moment, not the binder.
This is the lens that changes everything: the training exists to enable a 15-second recommendation on a crowded Tuesday.
Not to build brand love. Not to impress the regional sales director. Not to satisfy a PowerPoint template someone built three product launches ago.
Fifteen seconds. That's the window a bartender has when a guest is holding a menu and asking what's good. If your training doesn't produce a usable answer to that question, the training didn't work.
This means everything in the training session gets filtered through one question: will this help my staff say something useful to a guest tonight? If it doesn't pass that test, it lives in the supplemental binder, not in the room. The session is not the full story. It's the three things that actually help someone sell.
The operators who get this right build short, high-retention sessions where the staff walks out with a handful of real tools. The ones who get it wrong confuse thoroughness with effectiveness. A staff member who can recite the full production process but can't answer "is this smoky?" in a way a guest understands is a training failure, not a success.
Give the staff two minutes of category education, not twenty. What's happening with this category right now, why guests are asking about it, and what positions it against the alternatives they already know.
For mezcal: explain the tequila relationship clearly (both from agave, different plants, different regions, different process). Give a one-sentence terroir point. Tell them guests who know bourbon are graduating into agave spirits and they want the story. That's enough.
This layer exists to remove the staff's anxiety about not knowing enough. Confidence doesn't require depth. It requires having something true and specific to say.
Pick one or two production details that are genuinely interesting and translate to a guest experience. Not the whole documentary.
Most brands have twenty fascinating production facts. Use two of them. The ones that connect directly to what the product tastes like or why it's different from what the guest just had. "Hand-harvested, estate-grown" tells nobody anything at service. "Fermented in open pinewood tanks, which is why it has that slightly funky, earthy note" is a sentence a bartender can use.
The filter is always: does this detail explain what's in the glass?
Taste the product before you explain anything else, or right after the two-minute category brief. Not at minute 45.
Give the staff language for what they're tasting. Not wine-review language. Bar language. "Smoke, roasted pepper, a little fruit at the back" is usable. "Complex and layered with notes of terroir-driven minerality" is not a thing any bartender should say to a guest.
Then connect the flavor to a guest preference. "If a guest likes Scotch, especially peated whisky, lead with the smoke. If they're coming off a margarita and want something more interesting, lead with the agave sweetness and the finish." That's a service tool. That's what sticks.
Write three actual lines the staff can say. Not talking points. Lines. Things someone can walk up to a guest and say tonight.
The three should cover: the opening recommendation, the comparison to something familiar, and the one-sentence closer. Give the staff the option to use any of them depending on where the conversation is. Some guests want to know more. Some just want a yes or no. The scripts cover both.
Role-play them before the session ends. It feels awkward. Do it anyway. Awkward in practice beats a fumbled recommendation in front of a real guest.
This one gets skipped constantly and it's probably the most important layer for actual velocity.
How is this product placed on the menu right now? Is there a cocktail featuring it? A flight? A pairing with a food item? What's the natural upsell path from the well call? If a guest orders a tequila cocktail, is there a line for "we actually have a mezcal from Oaxaca that would make this better"?
The brand didn't build your menu. You did. The training should close the loop between product knowledge and the actual selling mechanisms your bar uses. If the bottle is sitting at the back of the spirits list with no cocktail and no feature, that's a problem that needs fixing before the training, not after.
When I built the staff training for a Mezcal's on-premise launch, the production story was genuinely fascinating. Capon agaves harvested after decades of growth. Open pinewood fermentation. A 1.5-tonne tahona. The Mendez Leon family history in the Matatlán valley, which is significant in the mezcal world.
All of it was real, and most of it stayed in the binder.
What made it into the training: the capon agave yield. Industry standard is roughly one liter per ten kilos of agave. Rosaluna's capon process produces one liter per five. Twice the agave for the same volume. That is a concrete, memorable fact that explains why the product is different and why it costs what it costs. A bartender can say that in one sentence.
That's the test. Take the brand's full story, find the two facts that answer "why this product, why this price, why tonight," and put those in the training. The rest is there for the staff member who wants to go deeper. Some will. Most won't. Design for both.
10 to 20 minutes. Remember, someone is paying for that time. Probably the bar. Not two hours. Not a "quick thirty minutes" that turns into ninety anyway.
Spend the first few minutes on category context and production story. The tasting is where the learning actually happens. Everything before the taste is setup. Everything after is application.
After the tasting, introduce the service scripts. Run the role-play. It takes ten minutes and it's the highest-ROI ten minutes in the session. Pick pairs. Have one person play the guest, one play the bartender. Rotate. The staff will laugh at themselves and each other. That's fine. That's how it sticks.
End with either a spiff structure or a tracking mechanism, preferably both. "The bartender who sells the most Rosaluna bottles this month gets a $50 gift card" is not complex to administer and it signals that you're paying attention. Alternatively, do a brief check-in at the two-week mark where managers review pour counts and give the staff a quick refresher on whichever script they're using least.
Don't end the session without someone knowing their role in tracking results.
Remember primacy and recency. People remember the first and last things they hear.
Keep it to three metrics or fewer. Operators have real work to do, and a ten-metric dashboard will be ignored.
Velocity pre/post. Bottle movement in the two weeks before the training versus two weeks after. Straightforward, available from the POS, and the most honest signal you have.
BOH pour count. If your POS tracks by modifier or pour category, watch the featured product's pour count against comparable spirits. You're looking for a relative lift, not a number in isolation.
Staff confidence score. One question, surveyed informally at the two-week check-in. "On a scale of 1-5, how comfortable do you feel recommending this product to a guest?" If the average is below a 4, something in the training didn't land and you have a specific problem to fix.
Guest mentions are hard to track operationally unless you're already running a review-mining system. Nice to have. Not worth building a manual process around.
Too much history, not enough bar. The brand spent years building this product and they want the staff to appreciate the whole story. The staff has four minutes before the first guest walks in. Edit to what matters at service.
No tasting until the end. You cannot train palate and application in the same fifteen minutes. Taste first, teach around the taste. The sequence matters.
No scripts. "Tell them about the production" is not a script. It's homework with no rubric. Write the actual lines. Your staff will improve on them once they get comfortable, but they need a starting point.
No follow-up at two weeks. The forgetting curve is real. Whatever the staff absorbed starts degrading the week after the session. A fifteen-minute stand-up at the two-week mark, revisiting the scripts and pulling a quick velocity number, extends retention by more than anything else you can do.
How long should the training session be?
60 to 90 minutes. That's the window where attention holds and you can cover all five layers including a real tasting and a role-play. Under 60 minutes you're cutting something important. Over 90 minutes you're losing people.
Do we train servers too or just bartenders?
Train both, but differently. Bartenders need the full session including sensory language and scripts. Servers need a two-minute version: one sentence about the product, one comparison, one recommendation for guests who ask. Build the server brief as a separate five-minute segment at the end of the session, or a short handout they can read before service.
What if the brand rep wants to run it?
Let them run the category and production sections. They know that material and it's credible coming from them. Have your bar manager or lead bartender run the scripts and role-play. The rep knows the brand. Your manager knows your service floor. Both are needed, and splitting the session that way is the most effective format.
How do we make it stick past week 2?
Two things: a spiff structure tied to velocity, and a brief check-in around day 14. The spiff creates a reason to use the scripts. The check-in catches the staff before they fully forget and reinforces the habit before it's gone. Neither takes significant management time. Both make a measurable difference in retention.
How do we measure if it worked?
Bottle velocity before and after is the cleanest signal. Run a two-week baseline before the training, then track the same two-week window after. If velocity is up, something worked. If it's flat, pull the pour count data and talk to the staff, because the problem is either in the scripts or in the menu placement.
For operators: If you want a staff training built as part of a broader system that tracks performance and keeps your team selling consistently, that's what ASM Command does. Start at jlittrell.com.
For brand marketers: Running an on-premise launch and need the framework built before the rep hits the floor? Same link, or reach out at kmsops.com/connect.
More playbooks like this one live at The Ops Wire.
Jason Littrell spent 10 years behind the bar in NYC, including Death & Co, and served as USBG NYC president. He now runs his hospitality consulting firm entirely on AI. He hosts the Hospitality Strategy Lab podcast and writes The Ops Wire newsletter.
Jason