We pulled 31 days of register data out of our own dispensary last week and found that dispensary discount leakage ate 18.3% of gross sales — just under half of every dollar of gross profit the store made. Not one of those discounts was a mistake. We approved all of them, one at a time, over months. That is exactly the problem: discounting never arrives as a decision. It accumulates.
Collateral Base has helped more than 100 operators across 30-plus states, but the numbers below are not from a client file. They come from Pekin’s Local Dispensary & Supply, the store Thomas Howard owns and runs in Illinois, wired into an AI assistant through a read-only connection to its Dutchie POS. We publish our own figures because we would rather show you the leak in a real store than describe one in the abstract. Treat them as directional figures from our shop, not a representation of what yours will look like.
Here is what 62 live discounts, 1,240 customers, and one very uncomfortable report taught us.
1. You find out your biggest discount is staff, not a special
The single largest discount line in the store over those 31 days was Employee Discount: $4,150.28 across 278 line items and 99 transactions. It beat every customer-facing promotion we ran. Our best-performing eighths special — the one we build menus and emails around — came in at $1,587. Staff discounting was more than two and a half times larger.
Add Manager Discount ($1,395.41 across 59 transactions) and internal discounting reaches $5,545.69, about 19% of every discount dollar the store gave away. That came from 158 transactions out of 1,240 — roughly one transaction in eight.
This is not a theft story, and we want to be careful there. Employee discounts are a real, legitimate benefit and ours are policy. But the National Retail Federation’s National Retail Security Survey put average retail shrink at 1.6% of sales, or $112.1 billion, and internal loss is a meaningful slice of that. Our internal discounting alone was more than double that shrink benchmark as a share of sales — and unlike shrink, it is fully visible in the POS the entire time. Nobody was hiding it. Nobody was looking at it either.
The benefit of seeing it: you can size the benefit deliberately instead of discovering it. A cap, a category exclusion, or simply a monthly number in front of the managers changes behavior without taking the perk away.
2. You see which discounts stacked on top of each other
Every discount line in our data carries a flag for whether it stacked with another discount on the same sale. Those flags are where the margin actually goes.
- Of the 278 employee-discount lines, 116 were stacked on top of another promotion.
- Our 4,000-point loyalty redemption stacked on 113 of its 131 lines — 86% of the time.
- A mix-and-match pre-roll bundle stacked on 70 of 90 lines, across only nine transactions.
Read that last one twice. Nine customers, ninety discounted lines, nearly all of them double-dipped. Each individual discount looked responsible when we built it. The combination never got reviewed, because no report we had showed combinations.
What you get from watching stacking: you stop guessing which promotions are quietly subsidizing each other and start writing exclusion rules that are grounded in what actually rang up.
3. You learn what dispensary discount leakage costs against gross profit, not sales
Most operators measure discounting as a percentage of sales, because that is the number the POS hands them. Ours was 18.3% of gross. That sounds survivable.
Now run it against gross profit. The store did $58,766.62 in gross profit at a 45.7% margin in that window. Total discounts were $28,832.73. Discounting consumed 49% of gross profit. Half the money the store actually earned went back out across the discount ledger.
That reframe is the entire point. A 5% discount on a 45% margin item is not a 5% problem — it is an 11% bite out of what you keep. Once we started reporting discounts against gross profit instead of sales, arguments about whether a special was “worth it” got dramatically shorter. For the rest of the numbers we check every morning, see our breakdown of the dispensary KPIs we review daily.
4. You catch discounts that ring up but aren’t on any list
Fourteen of the 62 discounts that fired in our store came back flagged as unconfigured — meaning they hit real transactions but did not appear in the POS’s list of configured active specials.
This is the ground-truth rule we now run everything by: what the register did beats what the specials screen says. Configuration is a claim. Realized line discounts are evidence. Where the two disagree, the register wins, every time, and the gap itself is a standing agenda item.
Our loyalty redemptions, employee and manager discounts, and a legacy bundle all live in that unconfigured bucket. If we had audited only the specials screen — which is what most weekly promo reviews actually do — we would have missed $13,328 of discounting in a single month. That is 46% of every discount dollar the store gave away, invisible to the screen most operators audit.
The related version of this problem on the pricing side is covered in our piece on the pricing leak an AI menu read caught, and the inventory version in dispensary inventory management.
5. You stop paying to run specials almost nobody reaches
Building a special is not free. Somebody configures it, somebody merchandises it, somebody trains the floor on it, and somebody fields the questions when it does not apply to the item in the customer’s hand.
In our window, a handful of promotions produced real discount dollars from a tiny number of baskets — one moved $556 across just nine transactions. Meanwhile our entry-level loyalty rung touched 526 transactions. Both are “working.” Only one is worth the operational overhead of existing.
Sixty-two simultaneous discounts across 1,240 customers works out to roughly twenty customers per discount. No budtender can hold sixty-two offers in their head, which means the ones that get mentioned at the counter are whichever three the staff happens to like. You are not running sixty-two promotions. You are running three, and paying to maintain fifty-nine.
Want to look at your own register data this way? Talk to an owner who runs a dispensary on this. Bring a question about your own store and get a real answer from someone who has to live with the same report on Monday morning.
6. You give loyalty a budget instead of a hope
Loyalty points are a discount line. They belong in the same ledger as everything else, and when we finally put them there we did not like what we saw.
Earlier this year, points redemptions at our store were running north of 10% of net sales and were the single largest discount category we had — bigger than any special. So we rebased the program: the redemption rate was cut in half, the top tier threshold moved up, and the join and referral bonuses were doubled to soften the change for members. The earn rate did not move.
In the 31 days we just pulled, redemptions across all eight reward rungs totaled $6,140 — about 4.8% of net sales. The rebase landed almost exactly where we modeled it. Member behavior held; the giveaway halved.
The lesson is not “cut your loyalty program.” It is that a rewards ladder without a target percentage of net sales is an open tab. Ours now has a number attached to it, and we check it monthly. You can see how we describe the resulting program to customers on the store’s own rewards page — the consumer-facing side of the same decision.
7. You walk into the special-planning meeting already knowing what to cut
The compounding effect of the six items above is that promotion planning stops being a debate about taste.
Before, the meeting sounded like: that one felt slow, let’s try something different. Now it sounds like: this promotion moved $221 across 36 lines with six of them stacked, it overlaps two other active offers on the same category, and it is coming off the board. That conversation takes four minutes and nobody has to win an argument.
Pulling this by hand is the reason most stores don’t. It is a pivot table across tens of thousands of line items, joined to a specials list that disagrees with itself, rebuilt weekly. Our manager used to lose the better part of a day to it. Now the read-only POS connection answers it conversationally, we look at it together, and the hours come back.
Worth saying plainly: we deliberately did not automate the decision. The assistant reads the register and tells us what it sees. It cannot configure a discount, change a price, or end a special — every write stays in human hands, gated behind an owner who has to click. We built it that way on purpose. An AI that can quietly reprice your shelf is not a tool, it is a liability, and we would rather be slower than explain that one to a regulator.
How to audit dispensary discount leakage in your own store this week
- Pull realized discounts, not configured ones. Ask your POS for actual discount dollars by line over the last 30 days. Ignore the specials screen entirely for this pass.
- Sort by dollars and look at the top five. If a staff or manager discount is in there, that is your headline, not a footnote.
- Divide total discounts by gross profit, not sales. Whatever number you get is the honest one.
- Find your stacked lines. Any promotion stacking on more than half its lines needs an exclusion rule or an end date.
- Count your live discounts against your customer count. Under twenty customers per discount means you are maintaining offers nobody is reaching.
- Give loyalty a target percentage of net sales and check it monthly, the way you would any other line.
None of that requires software. Catching dispensary discount leakage requires deciding the question is worth asking every month instead of every time margin looks strange. The software just makes it take minutes instead of a lost afternoon. If you want the wider view of where retail cash quietly goes, we catalogued it in dispensary money leaks and in our guide to managing a cannabis dispensary.
Bring your own numbers
We are operators before we are consultants, and the fastest version of this conversation is the one where you bring a real question about your own store. No deck, no discovery call theater — just the report you are suspicious of and someone who reads the same one every week.
Talk to an owner who runs a dispensary on this.
Collateral Base provides cannabis business consulting, not legal services. Figures above are directional operating data from a single Illinois dispensary over a 31-day window and are not a representation of results at any other store. Discounting, advertising, and loyalty program rules vary by state — Illinois operators should review the requirements administered by the Illinois Department of Financial and Professional Regulation, and operators elsewhere should confirm their own state’s rules. For legal questions about promotions or compliance, talk to a licensed attorney — our sister brand Cannabis Industry Lawyer handles that side.
