AI & Cannabis with Martez Knox: The Full-Funnel Approach to Dispensary Growth

<iframe width="100%" height="400" src="https://www.youtube.com/embed/ITBU-GJvmBs" frameborder="0" allowfullscreen title="AI & Cannabis with Martez Knox — COO Wisdom Pods"></iframe>

Watch on YouTube →

The COO Wisdom Pods interview with Martez Knox gets into the commercial mechanics of BakedBot in a way that's unusually direct: how does AI actually move numbers for a dispensary, and what does the math look like?

The answer Martez gives is the one that tends to make dispensary operators sit up straight.

The $400,000 Opportunity

Martez opens with a benchmark that frames everything:

"The average dispensary is doing like two million a year. Our tool delivers a 20–25% increase in per-transaction checkout. We're talking about an extra $400,000 here."

Twenty to twenty-five percent. On a per-transaction basis. That's not a loyalty point accumulation — that's a customer who would have bought one product walking out with two, because the recommendation was right.

This is what personalization at the transaction level actually does. When a customer knows what to add to their cart — because the recommendation is based on their actual purchase history and preference profile — the average order value goes up without any promotion or discount required.

Why BakedBot Attacked Cannabis

The host asks the obvious question: why cannabis? Why not retail broadly, or another vertical?

Martez's answer reflects the founding thesis:

"We wanted to start where we could actually disrupt. The way people buy cannabis — we believe that experience is completely broken. People go to a dispensary and speak to a bud tender they've never met before, and the bud tender is expected to know them based on whatever brief conversation they had."

The cannabis industry in 2023 had a knowledge gap that no other retail category had. Over 700 active strains. Complex terpene and cannabinoid profiles that interact differently with different individuals. A customer base that ranges from first-time medical patients to daily recreational users. And a budtender workforce with no systematic way to match any of it.

BakedBot's AI solves that problem by replacing the guesswork with data.

The Terpene and Cannabinoid Model

In this interview, Martez goes deeper on the science behind Smokey, BakedBot's AI budtender:

"When you look at a strain or any cannabis product, you have terpenes, you have cannabinoids — tons of ingredients inside this magnificent plant. What we've done is look at how do the cannabinoids affect certain psychographic things — particularly like: how do you feel after you smoke this? How do you feel after you smoke a Wedding Cake?"

The model doesn't just look at strain reviews. It looks at the actual molecular composition of products and maps those compositions to reported effects, customer preference data, and purchase history. Over time, the model gets more accurate for each individual customer — the more they buy, the better the recommendations get.

This is the flywheel that makes BakedBot's data advantage compound over time. An MSO with a million transactions has a better model than a two-location operator with ten thousand transactions — unless that smaller operator is using BakedBot, which pools learning across the platform.

The Full Funnel in Practice

Martez describes BakedBot's three-layer product architecture:

Attract — Programmatic advertising. Dispensaries can't run ads on Google or Facebook. BakedBot's programmatic layer puts compliant cannabis ads in front of qualified audiences in markets where advertising is permitted.

Nurture — Marketing dashboard with email and social campaigns. Once a customer is in the system, they get compliant sequences that keep them engaged between visits.

Convert — AI budtender. When the customer arrives — online or in-store — Smokey closes the gap between their intent and the right product.

Each layer feeds the next. Better advertising brings in better-matched customers. Better nurturing keeps them engaged longer. Better conversion means higher per-transaction value. The math compounds.

---

Full interview at COO Wisdom Pods.