This study demonstrated that customer service ranked as the most important attribute for the sample as a whole; however, there was considerable heterogeneity in terms of consumer preferences for cannabis retailer characteristics, as shown by the latent class model. Several preference patterns exist, and data from our study can help to better understand sub-groups of the cannabis consumer population to inform policies that impact how cannabis is sold in Canada.

While customer service was the most important attribute for the entire sample, this result was driven by two sub-groups (groups 1 and 4) that placed a very high priority on this characteristic. Group membership analysis showed that individuals in these groups were more likely to be those who purchased and consumed less frequently and those who started consuming after legalization. Interestingly, members of these sub-groups also placed particular importance on store regulation and the availability of extensive product information. The importance of customer service and product knowledge was supported by our previous work that consumers looked to cannabis store staff for information on the potential effects of a particular product, such as whether it would be relaxing, energizing, or could reduce pain, as well product lineage [9]. Other literature has shown that customers are comparing the customer service they get from regulated stores from what they are used to in the legacy market [21].

Consumer preference for customer service presents a challenge for licensed cannabis vendors, as retail regulations limit the extent to which cannabis store staff can share knowledge with consumers. The Cannabis Act [1] states that a person authorized to sell cannabis can only promote a product at the point of sale in terms of its price or availability, prohibiting any testimonials or personal endorsements of products. This means that the customer service offered by staff in regulated stores, if adhering to policy, is extremely limited. In some provinces, access to customer service and product information varies by store type. In Newfoundland and Labrador, the private cannabis retail model includes four tiers of cannabis stores [22]. Tiers 1 and 2 include standalone stores that prevent the entrance of minors and can only share information as outlined in the Cannabis Act [1]. Tier 3 and 4 stores allow the sale of other products, and minors are permitted on the premises. Staff in these stores can share government approved information through weblink and pamphlets, but it cannot be verbally shared between retailer and customer. It has also been noted that staff within publicly run stores are not necessarily hired based on their cannabis knowledge or ability to support customers in product selection [9]. These factors may deter consumers who place a high value on customer service from shopping for cannabis in regulated stores, particularly if they have access to knowledgeable sources elsewhere.

Product variety was moderately important for four out of five sub-groups in the current study, as indicated by consumers’ willingness to make trade-offs with other attributes in favor of a store with more selection. One attribute that was frequently traded off in favor of greater product variety was store regulation. This finding is in line with data from the 2021 Canadian Cannabis Survey [2] which found that of respondents that had used cannabis in the last 12 months, 55% reported having intended to purchase a product from a regulated source only to find that it was unavailable. This is less of an issue for dried flower cannabis or cannabis oils, but with the strict limitation of 10 mg of THC per package of cannabis edibles [1], many customers were not able to get their desired products from licensed stores. Additionally, some provinces have banned sales of all cannabis vapes due to the emergence of EVALI [23], making them only accessible to individuals in these provinces who have medical authorization. Limited variety and restrictions on product type and potency at regulated retailers may cause consumers to continue purchasing cannabis from unregulated sources.

Despite the perseverance of the unregulated cannabis market, purchasing a regulated product was preferred by most consumers. However, only 15% of the consumer sample placed Health Canada regulation status as the most important purchasing factor, as other groups were willing to trade off regulation in favor of proximity, customer service and sales. One issue with regulated cannabis stores is limited access, both in terms of proximity and hours of operation. With this in mind, privately run and hybrid retail models tend to have more cannabis stores per capita, with longer hours of operation than government-run retail models [24]. The increased accessibility to regulated cannabis products within provinces with private and hybrid retail models may be better positioned to strengthen legal sales.

There are mixed opinions on the value of increased public access to retail cannabis stores. On one hand, reducing access to cannabis retail stores may be an effective way to reduce consumption and related harms [24]. Meanwhile, increased access to regulated cannabis retail shops is an essential step towards decreasing unregulated sales. This perspective was supported by the literature, as Wadsworth et al. [25] found a positive relationship between the proximity of a regulated cannabis store and the likelihood of a respondent’s most recent cannabis purchase having been legal. For consumers who place a high value on proximity and convenience, a lack of legal storefronts may encourage them to obtain cannabis illegally. Provinces took a variety of approaches to spread out access to retail cannabis stores. For example, in Newfoundland and Labrador, retailer cannabis licenses are distributed based on postal code, to ensure widespread access while also not over saturating in some areas [26]. Other provinces, such as Manitoba and Ontario, have municipalities determine if a retail cannabis store will be allowed in their jurisdictions; however, limitations on store density do not seem to exist [27, 28]. In provinces that operate under a solely public retail model, the retail locations are determined by the province. There is an opportunity to do further research on these differences in store density and the impact it has on consumer decisions and behaviors.

Availability of sales and discounts was ranked as more important than most other attributes for about 39% of our sample (groups 2 and 5) and as the most important for 24%. Analysis of group membership data revealed that these groups also were more likely to be frequent purchasers and consumers, making up a large proportion of the cannabis market. Legal cannabis prices remained higher than those of the unlicensed market due to the expenses associated with extensive product testing and approval [9]. In a 2020 study, Mahamad et al. reported that legal dried flower cannabis was on average 19% more expensive than unregulated dried flower cannabis, noting that the price disparity increased with larger quantity purchased [29]. The Cannabis Act outlines maximum quantities that a person can purchase or possess at any time as well as strict regulations preventing any marketing or promotion activities such as discounts or bulk purchase offers [1]. These regulations prevent licensed stores from selling cannabis in quantities or at the price points that can be accessed through unlicensed sources. While this price disparity may deter some consumers, others may be willing to pay more for legal cannabis. Amlung and MacKillop found that consumers perceived legal cannabis as superior and preferred it when priced the same as or even slightly higher than unregulated cannabis [30]. However, once the price of legal cannabis passed a certain threshold, the unregulated product was preferred. Unless regulated cannabis stores can find a way to compete with unregulated prices, consumers primarily concerned with product cost may continue to buy from unregulated sources [29, 30].

There are several limitations to this study. Those limitations that are inherent to discrete choice studies, including ordering effect, hypothetical bias, and framing effect [14], are discussed along with the methods used to mitigate against them in the methods supplement of a previously published paper [17]. One attribute that was important when considering retailer preferences was store ambiance, as was noted in our preliminary qualitative study [9] and other retail preference studies [10]. Comments we have heard related to public stores suggested that these venues lack character. Meanwhile, private stores provide a variety of different atmospheres, with some catering to “stoner” culture and others displaying very sleek and minimalistic decor. It was not possible to include elements of store ambiance in this DCE as words could not adequately describe this attribute, and images would reveal other store characteristics that could influence choices. In future research, it would be interesting to look at the impact of ambiance on consumer behavior and see if consumers with certain characteristics were drawn to different store types. Another limitation of this research involved the demographic makeup of the participants. In the current study, 29.9% of participants reported an annual income of $100,000 or more. In comparison, the Canadian Census [31] reported that only 10.8% of 2021 respondents had an annual income of $100,000 or more. With this considered, the participant group in the present study was substantially more affluent than the general population of Canada, which may interfere with the ability to generalize findings to all Canadian consumers. Future research should explore whether less financially stable consumers would be willing to make the same price trade-offs in favor of other purchasing factors. Finally, consumer decisions on where to make purchases are also influenced by the characteristics of the products within the store. Product attributes (e.g., potency, packaging) could not be brought into this DCE; however, they were explored within the survey [16, 17]. To truly understand consumer choices, you need to consider both the characteristics of the products and retailer.