I've Seen This Movie Before

Every AI market map I see right now looks like the 2015 martech landscape with new logos.

Hundreds of tiny squares. Categories splitting into sub-categories. Everyone promising the same outcome with slightly different words. You squint at it and wonder how anyone picks.

I know that map well, because I had a logo on it.

Our little square

My co-founders and I started Forewards (the legal name was Toocoo Media). We built referral marketing software for ecommerce stores. The pitch was simple and backed by real data: people are four times more likely to buy when a friend refers them.

For a while, it worked. In 2013 we cracked the Branham300, the annual ranking of Canada's top tech companies, at #249. More importantly, we were #2 on its Top 10 Growth list, with 2000% growth. By 2015 we had our own little square in the "Loyalty / Referral / Gamification" box of the martech landscape, wedged between a few dozen other companies that looked a lot like us.

And that's where things went sideways.

Falling in love with magic

When you're one logo in a crowded box, the pressure to stand out is enormous. So we did what a lot of smart teams do. We got clever.

We built an "Easy Referral Program" wizard. Step one asked the store owner to pick their customer type: do your shoppers love high quality things, unique finds, a great deal, or doing good? Then we'd tailor the campaign to match. Targeting, personalization, a guided setup. It demoed beautifully.

It was also a swing and a miss.

The word "Easy" was in the name, and it was the one thing the product wasn't. We had asked a busy store owner to stop and think like a marketing strategist before they could get any value. We were proud of how smart it was. They just wanted more sales by Friday.

Looking back, the lesson is one sentence: we optimized for innovation, not the customer.

We fell in love with the magic. The customer never asked for magic. They asked for a job to be done.

Growth stalled. We couldn't raise our next round. And we folded Forewards.

From 2000% growth to closing the doors. Being clever in a crowded box didn't save us. Being useful might have.

Same movie, new logos

This is why the AI market map makes me a little nervous.

The technology is far more powerful than anything we had in 2015. That part is new. The behaviour around it is not. Teams are racing to show the cleverest demo, the most autonomous agent, the most "AI-powered" feature, because that's what stands out in a crowded box.

But customers don't buy magic. They buy the result. And the best AI is usually the AI you barely notice, quietly removing a step from a workflow someone already hated.

If your product needs a wizard to explain why it's smart, that's worth paying attention to.

Use your words

A few years later I got a second chance to learn a related lesson, from the opposite direction.

At Honeybee, I had the single best idea of my career. I came from AdTech, and AdTech was really just a two-sided marketplace: buyers, sellers, and a platform in the middle making the match. Group insurance looked the same to me. Why should it be any different?

So the model became: an employer sets a budget, like a wallet, for each employee. Employees pick and buy the coverage that fits their actual lives. A cross-industry idea that felt obvious once you saw it.

Then I went out and pitched it. "Plan configuration. Flexible enrolment. Consumer experience. Mobile first."

I sounded like a nerd.

The idea was right. The words were wrong. Nobody lies awake thinking about plan configuration. They think about whether their kid's braces are covered.

That's the flip side of the Forewards lesson. Even when you've found real customer value, you can still bury it under your own vocabulary. Every one of those phrases was accurate. None of them sounded like a human being with a problem.

AI teams are doing this right now too. "Agentic orchestration" and "multi-step reasoning" are true statements about the product. They are not reasons anyone buys it.

Take big swings at good pitches

I closed the talk I gave on this with three lines from baseball, and they still hold up.

Take big swings at good pitches. Honeybee was a big swing, and it was worth it because the pitch was good: a real problem, a market ready for change. Big swings at bad pitches are just strikeouts with extra effort.

Sometimes you just need to make contact. Forewards didn't need a clever wizard. It needed to get a store owner live and seeing referrals as fast as possible. A single beats a strikeout every time.

Use your words. Describe what you do the way your customer would describe their problem. If they wouldn't say it out loud, don't put it on your homepage.

The AI market map will shrink, the same way the martech map did. The squares that survive won't be the ones with the most magic. They'll be the ones a customer could explain to a friend in one sentence.

I'd know. I had a square.