TL;DR: If your bar can't tell you what it comped last Thursday, you don't have a comp system. You have a suggestion box. Set up four reason codes in your POS, enforce them at the point of entry, run a five-minute daily report, and keep comps between 1 and 3 percent of revenue. That's the whole system.
Tommy Two-Checks bought back about ten drinks last Tuesday. His coworker Mariana bought back twelve. The difference is that Mariana rang every one of them in.
Tommy thought he didn't have to. "It's hospitality," he said. "It's the cost of doing business." Mariana looked up from the POS and told him exactly what it cost: $187. On a Tuesday.
That exchange is the entire problem with comp tracking in most bars, collapsed into forty seconds. One bartender treating buybacks as acts of faith, the other treating them as line items. Only one of them is running a business.
The bar that can't tell you what it comped last Thursday doesn't have a comp system. It has a general vibe and a variance report that says "shrinkage" without any explanation of what actually happened.
The fix is not complicated. It's not expensive. It doesn't require new staff, new technology, or a single policy meeting that runs longer than it should. It requires four reason codes, a POS setting, and five minutes of your morning.
Every bar comps. A round for the regulars. A drink for the birthday table. A buyback after someone orders their fourth round. That's not a problem. That's hospitality. The problem is what happens when none of it gets recorded.
The bartender pours the buyback, gives a nod, moves to the next ticket. Nobody rings it in. At the end of the month, the variance report shows $2,400 in unaccounted product. Was it theft? Waste? Forty-three buybacks across three bartenders on eight different shifts? Nobody knows. The report says "shrinkage" and everyone moves on.
"It's the cost of doing business" is what operators say when they've decided not to know. It sounds like hospitality philosophy. It's accounting avoidance.
Regular buybacks build loyalty. Loyalty drives return visits. Return visits drive revenue. That's a business model. But it only works if you know what you're spending. Untracked generosity doesn't build a loyalty program. It just leaks.
The operators who fix this aren't cutting buybacks. They're measuring them.
You don't need a dozen categories. You need four that capture the actual reality of what happens behind the bar on a busy night. These cover it.
A guest orders multiple rounds and on round three or four the bartender sends one on the house. Classic. Who authorizes it: the bartender, within house policy (one per visit, more for regulars at manager discretion). What it tells you: high buyback rates on your best bartenders are usually a good sign. High rates on everyone, every shift, is a policy conversation.
A manager or owner comps a drink or meal outright: after a service failure, for a VIP, or as goodwill beyond standard buyback range. Who authorizes it: managers only. Bartenders don't touch this code. What it tells you: if manager comp spikes on a particular shift, something went wrong that night.
A drink comes back because something was wrong: wrong spec, off product, guest ordered the wrong thing. Reactive, not proactive. Who authorizes it: the bartender on the ticket. What it tells you: too many sends-back on the same cocktail is a training issue. Too many on the same shift is a different conversation.
Product got knocked over or ruined before it reached the guest. Who authorizes it: the bartender on the spot. What it tells you: recurring spills are a setup or flow problem, not bad luck.
Four codes. Everything fits somewhere.
A well-run independent bar typically runs comps at 1 to 3 percent of revenue. That's the range where generosity is present, intentional, and affordable.
What does 1 to 3 percent look like in real numbers? On a bar doing $60,000 a month in revenue, that's $600 to $1,800 in tracked comps. That's not nothing. That's a real budget line. And if you're spending it intentionally on regulars and relationship-building, it's probably paying for itself in return visits and tabs.
What does 5 percent or higher tell you? It tells you one of three things: your bartenders are over-generous without clear guidelines, your manager comps are covering service failures that shouldn't be happening, or your comp rate is high because a chunk of comps aren't actually being rung in and you're only seeing part of the picture.
What does 0 percent tell you? This one surprises people. A 0 percent comp rate is not good news. It almost certainly means bartenders aren't ringing comps in at all. The buybacks are still happening. The product is still leaving the building. The data just isn't reflecting it. If your comp rate looks suspiciously clean, pull inventory and compare. The story will surface.
The target is 1 to 3 percent of revenue, tracked, with reason codes, reviewed daily. That number becomes a business signal instead of a mystery.
The reason most comp systems fail is that they're optional. The bartender remembers to ring in comps when things are slow. Forgets when it's slammed. And the data ends up incomplete at exactly the moments when the bar is doing the most volume.
The fix is structural. Configure your POS so that a reason code is required before a comp goes through. No code, no comp. The system doesn't let the bartender skip it. This is available in most major systems, whether you're on Toast, Square, TouchBistro, or anything else in that category. It's usually a settings change, not a feature you have to pay for separately.
Second piece: turn on bartender-level reporting for comps. You want to see each person's comp total, their comp rate as a percentage of their ring, and how that compares to the house average. The goal is not to punish the bartender who's above average. The goal is to have a real conversation about what's driving it. Sometimes it's a regular they're particularly close with. Sometimes it's a training issue. Sometimes it's the best thing happening on your floor.
Third piece: set up a daily comp report that takes five minutes to read. Total comps for the shift, broken down by reason code, broken down by bartender. Compared to the same shift last week. That's it. One screen. Five minutes.
The system only works if someone actually looks at the report. Build that into the opening manager checklist.
The GM opens the report before the pre-shift meeting. Here's what they're looking at.
Total comps for last night's service. Is that number inside the 1 to 3 percent range? Good. Is it notably higher? Find out why before the shift starts, not during it.
Breakdown by reason code. Where did the comps go? If 70 percent of last night's comps were manager comp, something went sideways with service. If spill is unusually high, check the setup. If buyback is running double what it was last Tuesday, find out who was on the floor and what happened.
Breakdown by bartender. Who rang in comps? Who didn't? Anyone significantly above the house average gets a conversation, not a write-up. The conversation sounds like: "I see your comp rate ran higher last night. Walk me through what happened." Often the answer is completely reasonable. Occasionally it's not.
Comparison to prior periods. Last shift, last week, same shift thirty days ago. Trends matter more than single data points.
The report isn't the end. It's the starting point for a two-minute conversation with the person who closes. That conversation is where the real information lives. The report just tells you where to point the question.
Bartenders resist comp tracking because they think the house is getting cheap. That's the wrong read.
Tracking comps doesn't reduce hospitality. It gives hospitality a budget.
A bar spending $1,800 a month on buybacks and tracking every one of them is making an informed business decision. They know which nights and which bartenders are driving the spend. They can see whether it's working. They can defend it because they have the data.
A bar spending the same $1,800 without tracking is just hoping.
The message to the team: we're not tracking comps to take buybacks away. We're tracking them so we know what our hospitality is worth. Every bartender who's good at this job knows buybacks are a tool. We're asking them to use it intentionally.
Tracked generosity is still generosity. It's just generosity that knows its own name.
Won't this slow down service?
Ringing in a buyback with a reason code takes about four seconds. Ring the item, tap comp, select the reason code, done. The bartender who says it will slow service is usually the one who hasn't tried it yet. The friction disappears after a week.
How do I get bartenders to actually ring in buybacks?
Two things work. First, make it mandatory at the system level. If the POS requires a reason code before the comp goes through, the bartender can't skip it. The structure does the enforcement. Second, be transparent: tell the team you're tracking comps to defend the spend and keep doing it, not to cut it. Bartenders who understand the reason behind a policy follow it. Bartenders who think it's arbitrary don't.
What's a healthy comp percentage for a busy bar?
One to 3 percent of revenue is standard for a well-run independent. High-volume bars with strong regulars run closer to 3 percent. Quieter concepts might sit around 1 percent. Anything above 5 percent needs a conversation. Anything at 0 percent probably means the data is incomplete.
Should I track comps by bartender or just by shift?
Both. Start with the shift-level view to see whether a particular night is running hot. Then go to bartender-level to understand who's driving it. Shift-level data tells you when. Bartender-level tells you who and sometimes why.
What if my POS doesn't support required reason codes?
Check your comp or discount settings before assuming it's not there. Most modern systems have it. If your POS truly can't do it, use a paper log behind the bar and reconcile against the POS at close. More friction, but it works. If comp tracking is one of several things your system can't handle, that's worth a broader conversation about whether your POS is actually serving your operation.
Step 1: Audit your last 30 days of comp data.
Can you even pull it? Go into your POS reporting right now and find your comp report for the last 30 days. If it exists and shows reason code breakdowns and bartender-level detail, you're ahead of most bars. If it shows a single dollar number with no context, you have work to do. If it doesn't exist at all, you know where to start.
Step 2: Set up the four reason codes in your POS this week.
Buyback, Manager Comp, Drink Sent Back, Spill. Four codes. Get them in the system before the next shift. Configure them so a reason code is required before the comp goes through. That one settings change does more work than a policy memo ever will.
Step 3: Get help building this into your operation.
If you want a system that runs this automatically, flags bartenders above the house average, and feeds the data somewhere you can actually use it, that's what ASM Command is built for. NYC operators can also reach out through Punch List NYC for hands-on help building this into your bar's workflow.
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