The product system behind my $100M public company

We built four products that fed each other, and the system they created together was worth more than any of them individually.

By Matei Olaru · 2025-11-14 · 12 min read

Most people who've heard of Lift & Co. know it as a media company in the cannabis industry. It was a multi-sided platform with B2B monetization. Consumers used us for reviews and product discovery. Businesses paid us for data, targeted advertising, tradeshow placement, and content distribution to retail employees. We built the consumer product, monetized it through B2B channels, and the two sides made each other more valuable.

I want to tell the product story because I think it's more interesting than the business story. We built four products that fed each other, and the system they created together was worth more than any of them individually.

Starting from zero

Lift started as a blog. I took over as CEO when it was a bare-bones content site with no real revenue or product. We bootstrapped by building an events business. It eventually grew large enough to nearly sell for north of $25M. We raised ~$15M along the way. But the consumer platform is where the interesting product work happened.

Product 1: The consumer marketplace

The cannabis industry had just been legalized in Canada and nobody had reliable product information. Consumers were walking into stores blind. We built a review and recommendation platform, like Yelp for a category where Yelp didn't exist yet.

Users reviewed products, rated experiences, and built preference profiles. We scaled to over a million annual users.

A key product decision was for reviews to generate structured data. Every review captured specific attributes (effects, duration, use case, product format) in a way that could be queried, segmented, and sold. The consumer-facing product was primarily a data collection engine. We developed a version of a recommendation engine to guide consumer choices and the datasets collected powered our most ambitious B2B business.

Product 2: The rewards loop

The marketplace needed fuel. We needed people to keep reviewing, and the reviews needed to be detailed enough to generate useful data. So we built a rewards program: users who reviewed earned points, redeemable with licensed producers for products. Licensed producers were banned from advertising products so instead they advertised us because reviews were in effect advertisement for them.

This created a flywheel: More reviews meant more data. More data meant a better consumer experience (better recommendations). A better experience meant more users. More users meant more reviews.

If we had to pay for data acquisition through marketing spend, the model wouldn't work. The rewards program turned our users into our data pipeline. We hit 70% market penetration with the program among active cannabis consumers.

Product 3: The data and advertising business

With over a million users generating structured behavioral data, we had something nobody else in the industry had: a real picture of what consumers wanted and why.

We partnered with Nielsen to stitch our first-party preference and review data against their DMP data. We also started to collect purchase receipts from users to collect sales data. This combination let us build targeting and segmentation products for brands that were completely flying blind in a new legal market. They couldn't advertise on most platforms (cannabis restrictions on Google, Facebook, etc.), so we built an insights product using our data, data which was quite good at predicting consumer purchases.

The business model: brands paid us for market intelligence (what's selling, who's buying it, why, what they'll buy next), and they paid us for targeted distribution (reaching the right consumers through our platform). We monetized the data the marketplace generated, which the rewards loop accelerated.

Product 4: The education platform

This one came from a completely different angle and ended up being one of our strongest revenue engines.

The Ontario government needed every cannabis retail employee to complete a training certification before they could work in a store. We built it and convinced the regulator to give us an exclusive government mandate and became the sole certification provider for the province. CannSell captured roughly 52% market share for cannabis retail training.

The product positioning was bigger than certification revenue. We now had a direct relationship with every budtender in Ontario, the people standing behind the counter influencing purchase decisions. So we built a content distribution layer on top of the LMS. Brands paid us to distribute product education and incentive programs directly to the retail workforce.

That created a second monetization loop. The first loop was: consumers generate data, data gets sold to brands. The second loop was: brands pay to educate retail employees through our platform, employees recommend products, consumers buy them, consumers leave reviews on our marketplace, which generates more data.

Two loops, four products, all feeding each other.

The system

Here's what the platform looked like when all the pieces were connected:

The marketplace generated data. The rewards loop kept the marketplace growing. The data layer monetized what the marketplace collected. The education platform gave us a distribution channel to the point of sale, creating a second data input (what brands are pushing, what's being recommended) and a second revenue stream. Each product made the others more valuable.

I designed this vision in the first few months of taking over as CEO and executed against it, leading key aspects of product. The business grew revenue 20x in 36 months. We went public on the TSXV at a $100M valuation. Growth 500 ranked us in the top 15% of Canada's fastest-growing companies. We scaled to 60+ employees and struck multiple seven-figure deals with government and corporate partners.

What this taught me

I've built other things since Lift, but the lesson from this system keeps showing up in everything I work on. The best products aren't standalone tools. They're nodes in a system where each piece generates value that flows to the other pieces.

When I look at how most companies are building AI products right now, I see the opposite: isolated features bolted onto existing workflows, each one justified by its own ROI, none of them connected to each other. The companies that will win are building AI product systems where the data from one application improves every other application. That's the same architecture we built at Lift, just with better technology.

I was CEO of Lift, but I was also the chief product officer in practice. I came up with the product ideas, designed the loops, helped write the specs, and worked directly with PMs and engineers to ship them. The CEO title was the thing that let me clear organizational obstacles. The product work was the thing that mattered most.