What 24 Months of Search Data Reveals About the Gauteng Cannabis Consumer

20 April 2026

In April 2024, I made a decision that seemed minor at the time: I connected my cannabis dispensary directory to Google BigQuery and started storing every search query, every impression, and every click in a permanent data warehouse.

I didn't know exactly what I was going to do with it. I just had a feeling that in a market as opaque and rapidly evolving as South African cannabis, the data would eventually be worth more than the directory.

Twenty-four months later, I have millions of impressions worth of search data. I have month-by-month click trends. I have the exact phrases Gauteng residents type into Google when they're looking for cannabis on a Friday afternoon. I have competitor search patterns, product-specific query data, and suburb-level demand signals that tell me, with reasonable confidence, which parts of Johannesburg are underserved right now and where search demand is accelerating fastest.

Nobody else in South Africa is sitting on this. Not a dispensary, not a research firm, not a government agency. It exists because I decided twenty-four months ago to store what most people throw away.

But aggregate data is only interesting if it pays rent. Here is what 24 months of consumer search behaviour reveals about the market, and exactly how much it's costing dispensaries that aren't paying attention.

What the Numbers Actually Mean

Let me make this concrete with a real example. The figures below are live data, pulled directly from my BigQuery warehouse as you read this.

Those impressions represent moments when a person in Johannesburg typed something into Google and the answer that came back pointed toward this business. They weren't passively scrolling Instagram. They were searching, the highest-intent behaviour a consumer can exhibit before making a purchase.

Because this business had no dedicated website, every one of those clicks landed on a third-party directory. They were, in the language of the data, working hard to figure out how to buy. The gap between clicks and sessions tells you they kept coming back, digging deeper, looking for a way through.

But what is that search traffic actually worth?

The Unit Economics of a Missed Search

Dispensary owners know exactly what sold yesterday. They rarely know the financial value of the customer who couldn't find them today. Let's translate the live click number into hard Rands.

Leaving high-intent search traffic on the table isn't just a marketing oversight. It is costing dispensaries hundreds of thousands of Rands in missed lifetime value, and that estimate uses a conservative 10% walk-in conversion on clicks that are already screened for purchase intent.

The Search Queries Tell You Everything

Aggregate numbers are interesting. Individual search queries are fascinating. Here are the actual phrases people typed into Google to find these listings, ranked by what they reveal about intent.

Branded queries — people who already know the name:

High intent, high likelihood of conversion. The business already owns these customers' consideration, they just need somewhere to land.

Discovery & competitor queries — the demand nobody expected:

Loading query analysis…

Hidden queries — the volume Google won't show you:

Loading hidden query analysis…

The Trend That Matters Most

Month-by-month, the click data across the locations tells a story of uneven but overall upward momentum. Here is the picture by suburb, live from the data warehouse:

Loading suburb analysis…

The Competitor Signal: Stealing Traffic

There is a final layer the branded and generic queries don't capture on their own: Competitor Conquesting.

Because the South African cannabis market is still digitally immature, most established brands rely on physical presence and Instagram awareness while neglecting local SEO. The data shows users actively searching for direct competitors, and clicking on Sativa listings instead, because a third-party directory is outranking larger brands in suburb-level results.

Business owners hate losing customers to competitors, but they love stealing them. Right now, a strong local SEO presence is a licence to legally siphon traffic from competitors' marketing spend. Most dispensaries don't know this is happening, because most dispensaries aren't collecting the data.

The Anatomy of a Missed Customer

Data without context is just numbers. Let's look at what those impressions and sessions actually represent in the real world.

It's a Friday afternoon in Craighall. A local resident has money to spend right now. They search "edibles gummies near me."

Because the local dispensary has no indexed website, they don't show up in organic results. My directory does. The customer clicks the link. They want to see a live menu, verify stock, and click a map pin to drive over.

Instead, they hit a third-party platform. They can't process an order. They can't see live stock. They have to context-switch to Instagram, DM an admin, or guess if the store is open.

This is the ultimate blind spot. The intent is there. But faced with that friction, many simply back out, search again, and go to a competitor whose map pin actually loads and whose live menu is visible. The sessions-vs-clicks gap in the data tells you those customers are trying hard—returning to the page, digging deeper, looking for a way through. The intent is real. The infrastructure to capture it isn't.

The 2027 Market & Your AI Data Moat

The South African recreational cannabis trade is moving toward full commercialisation. When the 2027 Omnibus Bill opens regulated retail, every serious operator will immediately start building digital infrastructure and competing for local search visibility.

But there is a second, arguably bigger shift happening simultaneously: the integration of Large Language Models (LLMs) into retail operations. Every franchisor will eventually want to use AI to forecast demand, automate localized marketing, and deeply analyze consumer behavior.

Here is the reality of AI: an LLM is only as smart as the context you feed it. If you plug an AI into a business with no historical data, you just get a generic chatbot. But if you plug an AI into years of your specific, suburb-level search trends and sales data? You gain a massive, highly specific competitive advantage.

This is why data collection needs to start now. The value of the infrastructure I build accrues and compounds over time. You cannot go back and collect 2024 search behaviour in 2028. It either exists or it doesn't. By starting today, as AI models get smarter, your proprietary dataset grows deeper, giving you the ultimate context window when the market fully opens.

Stop Guessing. Start Building Your Data Moat.

Right now, most South African cannabis operators are making decisions based on foot traffic, intuition, and Instagram engagement. But a mature digital footprint is the difference between capturing an impulse buyer and losing them to the shop down the road.

You cannot act on data you don't own, and you cannot train an AI on data you never saved. Here is what a modern, scalable dispensary data stack actually looks like, the exact infrastructure we build for operators at Full Extract:

• The Analytics Foundation: Proper implementation of GA4, Google Search Console, and Bing Webmaster Tools, correctly configured for cannabis retail with custom event tracking.
• BigQuery Data Warehouse: Connecting your search data to Google BigQuery, creating a permanent, queryable store of your search history that you own from day one.
• Custom Telegram Dashboards: Search impressions, query trends, and competitor signals delivered as a Telegram Bot you can check daily, rather than a dense PDF you ignore.
• AI-Ready Context & LLM Integration: Allowing you to ask plain-English questions about your sales and search data today, while quietly building the deep historical context your business will need to power custom AI models tomorrow.

We build the infrastructure. You own the insight. No lock-in, no black box. We configure it, explain it, and hand it over so you can start competing on evidence, not instinct.

The Data Is There. You Just Have to Collect It.

I build scalable data infrastructure exclusively for dispensaries, franchisors, and multi-location operators through Full Extract. Whether you are looking to fix immediate leaks in your sales funnel today, or you want to prepare the foundational data you'll need to run an AI-driven retail operation tomorrow, we provide the intelligence you need.

Stop bleeding traffic and start capturing history.
We will set up your analytics foundation, build your custom dashboards, and ensure that when you are ready to leverage LLMs in the future, you have the rich, proprietary dataset required to make it actually work.

Start Building Your Infrastructure →

All initial consultations are free. We'll show you exactly what the data says about your target market.