Blog: Applying Churn Models to Cannabis Using Predictive Analytics April 18, 2019

Blog: Applying Churn Models to Cannabis Using Predictive Analytics

administrator No comments

Predicting and preventing cannabis churn from Strain or Cannabis Product availability datasets represents a huge additional revenue source for every Cannabis organisation in the Weed Supply Chain.

What is Cannabis Churn?

Cannabis churn refers to when a Consumer – whether its a Medical Marijuana Patient, a Recreational User, or even long term cannabis Customers – ceases their relationship with an Organisation anywhere along the Supply Chain. Generally, an organisation typically defines Churn Metrics as a Consumer has “churned out” once a particular amount of time has elapsed since the consumer’s last interaction with particular Products, Services, Farmers or Retailers.

The full costs of Cannabis Consumer Churn includes both lost revenue and the marketing costs involved with replacing those consumers with new engaged customers.

Churn is a Key Performance Indicator that most every Chief Marketing Officer at large Organisations pays attention to daily. Often times during Quarterly Earnings Reports, Churn is mentioned as a leading indicator of a company’s performance.

Churn Metrics can also used as a focal point for programs and promotions such as direct mail campaigns, Churn Win-backs to get the Consumer back as an engaged party or even Multi Million dollar Advertisements to entice a party back to a company. Regularly appending psychographic and demographic attributes to Churn datasets provides Cannabis Executives an overall profile so that Returns on Investments for employee’s efforts can be calculated.

Reducing Churn is a key business objective of every Fortune 500 organisation. To date, I have not seen big Cannabis retailers attempting Churn Win-back programs such as Good, Better, Best pricing strategies that address the overall shortcomings of Strain availability data so Cannabis consumers can receive better treatment.

It is generally during a period in the “season” that a Cannabis Consumer will Churn out.

Is Strain Availability data a good Predictive Churn Metric?

As an example, consider Strain availability on a Cannabis Retail Store by Store basis. As a Retailer in the weed supply chain, most likely you want Cannabis Consumers to have a “like experience” at each location. If Jack Herer (JH) works as a good treatment for you then what happens if Retail Location A has that Strain while Retail Location B does not. As a Cannabis Consumer, I want to know what particular Retail Location has my Strain so that I am not left with a disappointing Cannabis shopping experience.

Why Should I Include Seasonality in the Cannabis Churn Dataset?

Seasonality is an interesting attribute to use in the dataset because it is generally during a period in the “season” that a Cannabis Consumer, generally speaking, will Churn out as a Paying Customer for the metrics as well as attributes included in the Regression models. If the ability to drill down on Geographical aspects of the data by Season is included, this would be very powerful analytics if weather conditions in a State vary on the maps.

Which Other Data Attributes Should I Consider in my Cannabis Churn Regression Analysis?

The possibilities are infinite and Churn Regression Models should be done on a regular basis, perhaps even daily. According to the Edward Jones marijuana report, any investment in Canada’s legal Cannabis market would be entirely speculative because: 1) the medical marijuana market is too small, 2) companies in the space have limited operating histories, 3) regulations continue to change, 4) the recreational market’s low barriers to entry will increase competition and pressure prices, 5) public marijuana companies have volatile stock prices, and 6) marijuana startups have more complex risks than traditional startups. Any of these variables, based on Edward Jones’s analysts, should be calculated from the top down.

Consider the following mental model: I am a business development Executive with Current and Future Cannabis Retail locations. I also interact with the grow facilities in the southern part of the State. Finally, my Cannabis Retail Shops have Cannabis Strains that other Dispensaries do not have.

In consideration of all the Cannabis variables above, tell me: which Paying Cannabis Customers will Churn out due to the availability of each Strain Type, overall Strain availability and the number of Retail Cannabis Customers populated in particular Markets. In theory, these too could be used as Regression Variables in the statistics engine.

Cannabis Churn is a complex issue that should be a part of every Cannabis Marketing team’s long term strategy. However, tough issues are fun issues and I am certain solutions can be engineered to assist with the tactical issues associated with a Cannabis Organisation.

Profile: Mark McDonnell, MS, CSM, CSPO

I’m passionate about building great Products that make people’s lives easier. I particularly enjoy Design Thinking Applications for individuals that have experienced difficulties in their life due to Depression, Anxiety or PTSD.

I have over 20 years of experience strategising innovative digital experiences for small startups to the world’s biggest media brands including XM/Sirius Satellite Radio, the Discovery Channel, ComScore Media Metrics and the RIAA.

I am a City kid, was born in SE Washington, DC (where I got my funny nickname Duppy) love Designing, and I’m excited to collaborate with you!

AgileMinder – We Believe that Happy People make Better Products and Services.

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: