How a Three-Location Dispensary Grew Repeat Customer Revenue by 28%

When a multi-location cannabis retailer first approached Chapters Data, their headline numbers looked healthy. Transaction volume was up. New customer acquisition was strong. But the owner had a nagging suspicion: too many first-time visitors never came back. His gut said the problem was product mix. The data said something different.

This is the story of how that dispensary used customer segmentation — starting with nothing more than 18 months of POS transaction history — to change how they communicated with customers, what they stocked, and ultimately how they grew.

The Problem Wasn't the Product

The dispensary ran three locations across two markets. Monthly transactions averaged around 4,200 across all three stores. On the surface, things were fine. But when we pulled their cohort retention data, the picture sharpened immediately.

Only 38% of first-time buyers made a second purchase within 90 days. Industry benchmarks for cannabis retail put healthy 90-day second-purchase rates between 45–55%. They were losing the retention game by a wide margin — and because they weren't measuring it, they didn't know.

The instinct to blame the product assortment wasn't entirely wrong. But the bigger issue was more nuanced: different types of customers needed different types of engagement, and the dispensary was treating all of them the same. Every customer got the same SMS blasts, the same promotions, the same budtender script.

When your messaging is generic, your results will be too.

Building the Segmentation Framework

The first step was straightforward: take 18 months of anonymized, loyalty-linked transaction data and sort customers into behavioral groups. No custom software. No machine learning. Just three variables:

Recency: How recently had they purchased?
Frequency: How often were they buying?
Value: How much were they spending per visit and in total?

This is the classic RFM framework, and it works precisely because it's simple and actionable. A small business owner can understand it without a data science background, and the output maps directly to campaign decisions.

Within 30 days of loading the data, the dispensary had a clear segmentation picture — and it immediately changed how they thought about their customer base.

The Three Segments That Changed Everything

Engaged Regulars — 22% of customers, 61% of revenue

These customers had purchased 3 or more times in the past 6 months, with an average basket of $72. They had strong preferences for flower and concentrates. They knew what they wanted and bought it consistently. They didn't need to be sold — they needed to be retained and occasionally surprised.

Explorers — 31% of customers, 28% of revenue

This group bought regularly but inconsistently: 1–2 times per quarter, average basket of $58. Their category distribution was more diverse — edibles, topicals, and vapes alongside flower. They were curious. They were educable. And they were the highest-upside segment in the entire customer base.

At-Risk and Lapsed — 47% of customers, 11% of revenue

These were the customers the owner had been sensing. Most had purchased once or twice, then gone quiet. A significant subset had visited multiple times but hadn't returned in 90+ days. This group was consuming acquisition budget with little compounding return.

Building Campaigns That Matched Each Segment

With segments defined, the next step was replacing broadcast promotions with communications that matched actual customer behavior.

Protecting the Regulars

The Engaged Regulars didn't need discounting. They were already spending consistently. What they responded to was recognition and access.

The dispensary built an early-notification system: new product arrivals were announced to this group 24 hours before general promotions went out. They also introduced a lightweight tier in their existing loyalty program — no points overhaul required — that unlocked early access to limited drops and quarterly member events.

The message to these customers shifted from "here's a deal" to "you're one of our best customers, so you hear about this first."

Result: the Engaged Regulars group's 6-month purchase frequency increased from an average of 4.1 to 5.3 visits. Marginal revenue per customer rose roughly $140 over the measurement period, and churn in this segment dropped from 14% to 9%.

Accelerating the Explorers

The Explorer segment held the biggest opportunity. These customers were already engaged — just not deeply. The dispensary's own budtenders reported that Explorers often asked for category guidance but didn't always get it amid the pace of a busy floor.

The solution was a product education email sequence triggered at the 30-day mark for any customer who had made 2 purchases but hadn't returned in 3 weeks. Each email focused on a single product category with a practical "how to choose" framing — not a promotional blast.

Three emails. No discounts in the first two. A modest 10% offer in the third, specifically on a category the customer hadn't tried yet based on their purchase history.

The results were significant. Within 90 days of launching this sequence:

Explorer 90-day retention jumped from 41% to 58%
Average second-purchase basket for this group rose by $14 as customers began exploring adjacent categories
Open rates on the triggered sequences ran 31–38%, compared to 12–14% on previous broadcast promotions

The key difference wasn't the offer — it was the relevance. A message that says "you've tried flower, here's what to know about vapes" lands differently than "10% off this weekend."

Reactivating the Lapsed — and Knowing When to Stop

The At-Risk and Lapsed segment required a more considered approach.

For customers 60–90 days post-purchase, a simple reactivation offer with a flat dollar discount off their next visit captured a portion of them. Win rate on reactivation was modest — around 18% — but on a base of several hundred lapsed customers, that added up.

For truly lapsed customers (120+ days dormant), the math was different. Reactivation rates dropped low enough that the dispensary made a deliberate, data-backed decision: stop spending budget on this group. Shift that spend toward reinforcing the Explorer segment, where it compounded over time.

This single decision freed up approximately $800 per month in promotional spend that had been going toward customers statistically unlikely to return — and redirected it toward customers who were already showing engagement signals.

The Results: 12 Months Later

Twelve months after implementing the segmentation model and its three associated campaigns, the dispensary ran a comparative analysis against the prior-year period.

The results across the board:

Overall repeat purchase rate (90-day): 38% → 51%
Revenue from existing customers: up 28%
Average customer lifetime value (18-month window): up $94
Churn in the top-20% segment: down 38% relative to prior period
Net Promoter Score (measured via post-visit SMS survey): up 11 points

New customer acquisition stayed roughly flat throughout this period — this was not a volume play. It was a retention play. And because retaining an existing customer costs substantially less than acquiring a new one, the margin improvement exceeded what the topline revenue figure alone suggests.

What Made It Work

A few factors separated this project from segmentation efforts that stall.

They used data they already had. No new platforms. No expensive integrations. The POS data existed. The loyalty data existed. What was missing was the analytical framework to turn it into actionable customer segments — and the willingness to act on what it revealed.

They stopped treating email like a megaphone. The shift from broadcast promotions to behavior-triggered, segment-specific messaging was the single change most directly responsible for Explorer retention gains. Generic blasts informed nobody. Triggered sequences informed the right people at the right time.

They accepted that not all customers are worth the same investment. Letting go of the lapsed segment — deliberately, based on data — felt counterintuitive. Owners often resist it emotionally. But continuing to spend promotional budget chasing customers who are statistically unlikely to return is opportunity cost against the segments where the math actually works.

Someone with authority reviewed the data weekly. One pattern we see consistently across clients who succeed with analytics: a decision-maker is looking at the numbers on a regular cadence. At this dispensary, the GM pulled up segment dashboards every Monday morning. That rhythm kept campaigns from going stale and surfaced early signals when behavior shifted.

The Bottom Line

Customer segmentation isn't a new idea. But for cannabis retailers still managing by transaction averages and intuition, it represents a significant and available competitive advantage — one that doesn't require new technology or a data team to capture.

This dispensary didn't build anything from scratch. They took 18 months of data they already owned, sorted their customers into three behavioral groups, and built communications that matched what each group actually needed.

The result was a 28% lift in repeat customer revenue — driven not by acquiring more customers, but by getting more value from the ones they'd already earned.

Retention is undervalued: A meaningful improvement in repeat purchase rate compounds faster than most acquisition investments
RFM segmentation is a starting point, not a ceiling: Once segments are defined, campaigns can grow increasingly sophisticated over time
Behavioral triggers outperform broadcast: Messages triggered by purchase history consistently outperform generic promotions in cannabis retail

At Chapters Data, this is the kind of analysis we bring to cannabis retailers and small businesses every day — turning existing transaction data into a clear picture of which customers need what, and when. If you're sitting on months of purchase history without using it to understand customer behavior, that's revenue being left on the table.