The dispensary marketing plan: put retention first

Most dispensary marketing budgets are spent getting someone through the door once, and the money is made on the fourth visit. This plan is built the other way round. It starts from the customers you already have, puts the channels that carriers and regulators actually allow in the order you should turn them on, and ends with a 90 day calendar and the three numbers in your point of sale that tell you whether it worked.

It is written for a single store or a group of up to about six, where the person running marketing is also the person buying the software.

Where dispensary marketing budgets go wrong

The common mistake is buying reach in a category where reach is the most restricted and least measurable thing you can buy. Mainstream ad platforms decline cannabis creative whatever your state allows, so dispensary spend gets pushed into discount listings, geofenced display and paid directories, where the impression is counted by the vendor and the visit is counted in your point of sale, and nothing joins the two.

Three specific leaks:

Discounting to buy a visit you already had. A storewide percentage off reaches your regulars at the same time as a new customer, so you pay margin on purchases that were going to happen.

Paying twice for the same person. Budget goes to acquisition media that reaches your own existing customers, who you can already contact for far less through a channel you own.

Spending without a definition of a result. If the plan has no pre-set measurement window and no held-back group, every channel will look like it worked, and none of them can be compared.

The size of the prize is set by the ticket. According to Headset point of sale data, US cannabis sales reached $24.3 billion across the 16 state markets Headset tracks in the 12 months ending June 2026, spread across 473.8 million transactions at an average basket of $47.29 holding 2.74 items (Headset, cannabis industry statistics). At a basket under fifty dollars, an incremental visit is worth a few dollars of margin. That number is the budget ceiling for every channel in the plan.

Headset also puts Millennials at 42.8 percent of US cannabis spending in that period, ahead of Generation X at 23.5 percent and Generation Z at 21.2 percent. A plan built on one channel is a plan built for one of those cohorts.

The retention-first plan

Spend on repeat visits first, acquisition second, because a repeat visit costs a fraction of a first one and you already have permission to ask for it. Work in this order.

One, enroll everyone at the counter. A loyalty program you cannot join in fifteen seconds is not a loyalty program. A wallet pass, which is a loyalty card that lives in Apple Wallet or Google Wallet rather than in an app the customer has to download, is joined by scanning a code and tapping add. Sticky Cards issues that pass, with the earn and the redemption both happening inside your point of sale.

Enrollment is not the hard part, use is. The 2025 EY Loyalty Market Study found 92 percent of consumers already enrolled in at least one loyalty program, with almost half in more than five (EY Loyalty Market Study, 2025). Deloitte's retail research puts the average consumer at eight programs enrolled and five actively used (Deloitte, reshaping loyalty programs). A card on the lock screen is in the five. A card in a drawer is in the eight.

Two, earn and redeem in the point of sale. A reward the budtender applies as part of the sale gets redeemed. A code the customer has to find in an email at the counter does not. This is the single biggest difference between a loyalty program that moves repeat rate and one that fills a database.

Three, segment on behaviour you already hold. Your POS already knows who has not been in for 30 days, who buys only pre-rolls, who spends in the top fifth, and whose birthday is this week. Those are campaigns, not reports. Sticky Cards builds segments from live POS events, including a set of pre-built ones for exactly these cases.

Four, win back before they are gone. A customer 30 days absent is a message. A customer 90 days absent is a reacquisition cost. Set the trigger at the first window, not the second.

Five, then buy acquisition. Once the returning-customer machine is running, new customer spend is worth more, because the second and third visit now happen on their own.

Channels that work under carrier and platform rules

Pick channels by what will actually be delivered, not by what a media kit offers. Four that work for dispensaries, in the order of least friction:

Wallet push. A notification to the loyalty pass on the lock screen. It rides Apple and Google notification rails rather than mobile carriers, so cannabis copy is not subject to carrier content filtering.

Email. Allowed, and the workhorse for anything long form. Note that general purpose email platforms vary in whether they accept cannabis senders, and Klaviyo in particular does not allow cannabis SMS, which is why it sits alongside a cannabis platform rather than replacing one.

SMS and MMS, after 10DLC registration. 10DLC, or 10 digit long code, is the carrier registration scheme that application to person text traffic must go through on ordinary ten digit numbers. Cannabis is a restricted category within it, so the campaign needs specific handling and age gating rather than a standard registration. CTIA's rules are the reason: its Short Code Monitoring Program Handbook names content that endorses controlled substances as prohibited and requires appropriate age gating for controlled substances (CTIA, Short Code Monitoring Program Handbook v1.9, January 2024), and its Messaging Principles and Best Practices treat a promotional message as needing the consumer's written agreement before the first send (CTIA, May 2023). Sticky Cards handles the registration for the store, and sends SMS and MMS once it is in place.

In-store and the counter itself. Signage, receipts and the budtender's thirty seconds are the cheapest acquisition channel a dispensary owns, and the only one nobody can switch off.

What to treat carefully: geofenced display advertising, where the location data supply is now under direct federal enforcement. The Federal Trade Commission's December 3, 2024 order against Mobilewalla bans the sale of sensitive location data and, for the first time, prohibits collecting consumer data from real time bidding exchanges (FTC, December 3, 2024). Treat geofenced media as a test with a held-back control, not a standing line item.

State advertising rules sit over all of this. Audience age composition minimums, distance rules relative to schools and youth facilities, and required warning language are set state by state, so clear each market separately.

A 90 day calendar

Run it in three blocks of thirty days, one job per block, because a dispensary marketing plan fails from doing five things at once.

Days 1 to 30, build the base. Connect the loyalty platform to your point of sale. Decide the earn rate and two reward thresholds. Design the pass. Put QR codes at the counter, on receipts and on delivery bags. Train every budtender on the one sentence they say at checkout. Start the 10DLC registration now, because it is the longest lead item in the plan. Set your baseline: record repeat visit rate, average basket and days between visits for the last 90 days, before anything changes.

Days 31 to 60, turn on the three campaigns that pay. A welcome message on the first visit after enrollment. A 30 day absence win-back. A birthday reward. Hold back a control group of similar customers from each one, so you can tell what the campaign did. Review enrollment rate by location weekly and fix whichever QR placement is underperforming.

Days 61 to 90, segment and measure. Add two behavioural campaigns from your own data, usually a category-based offer and a top-spender perk. Compare the three month numbers against the baseline you recorded in block one. Only now consider acquisition media, and only with a defined test window.

Two things to keep out of the first 90 days: a tier structure, which needs a few months of data to set thresholds sensibly, and discount-led storewide promotions, which will make every other number harder to read.

Measuring it in the point of sale

Judge every campaign on three POS numbers against a held-back control group, over a window you set in advance.

Repeat visit rate. The share of customers in a period who come back within the next period. This is the number the whole plan exists to move.

Average basket. Revenue divided by transactions, for members against non-members in the same stores and the same weeks. Comparing members to a national benchmark tells you nothing. Comparing them to the people standing in the same queue tells you everything.

Days to next visit. The gap between visits, for members against non-members. This moves before revenue does, so it is the earliest signal that the plan is working.

How to run the comparison honestly: pick the window before the send, seven or fourteen days. Hold back a control group of similar customers. Compare the two groups on the three numbers above. Keep the same definition every month, because a metric you redefine is a metric you cannot trend.

Build the plan on your own numbers

Book a demo and we will run repeat visit rate, average basket and days to next visit on your own point of sale data, and show you what the plan is worth for your store.

Book a demo