The AI Hype Cycle, What it Really Means

The Multi-Generational AI Strategy: Why Smart Marketing Leaders Plan Beyond the Current Hype Cycle

I talk with a lot of people about AI. Everyone is experimenting and at different stages of their AI strategy.

Some are all in.

Many feel late to the party.

And some think AI is overhyped.

But is it? Is AI overhyped? It's hard to say because AI isn't just one thing.

The Hype Cycle Reality Check

Scott Brinker's latest analysis at ChiefMarTec reveals something most people miss about AI adoption: "AI" as a category is useless for strategic planning. Different AI applications are moving through the hype cycle at completely different speeds, and what used to take a decade now happens in years or months.

A new SAS report tracking 10 different AI applications in marketing shows this split clearly. Some tools are taking off:

• Chatbots/customer interactions: +44.2% growth in one year
• Text generation: +32.4% growth
• Trends analysis: +56.5% growth

Others are actually losing steam:

• Video generation: -9.1%
• Customer-journey mapping: -4.3%

This isn't random. It's the hype cycle in action, and figuring out where each tool sits gives you a real advantage.

Mapping the Current Landscape

If you plot these applications on the classic hype curve, patterns emerge:

Plateau of Productivity: Customer service chatbots hit 62% adoption. They work reliably for basic questions and companies actually know how to implement them. Safe territory for investment.

Slope of Enlightenment: Text generation tools sit at 45% adoption and climbing. The initial disappointment with generic AI copy is giving way to smarter applications. Companies are figuring out what actually works.

Trough of Disillusionment: Video generation is losing adoption after massive early hype. Turns out the technology isn't quite ready, and companies are pulling back after disappointing pilots. Or it's too expensive for the current results.

Peak of Inflated Expectations: AI agents and advanced automation are generating huge excitement, but most implementations are still experiments.

Here's the thing: each of these will follow its own path, and being early or late to each curve creates completely different strategic implications.

Why Measurement Matters More Than Headlines

Remember that viral MIT study claiming 95% of AI pilots fail? It's the perfect example of the measurement problem. The study defined "success" as direct P&L impact within six months and completely ignored efficiency gains, cost reductions, or customer experience improvements.

At first I was dismayed because it made it seem like AI is overhyped. I don't believe that it is. It's not perfect yet, but different aspects are getting better every week.

And when was the last time you started using new enterprise-level tech and it had an impact in six months? Does this count the 2-3 months it takes to get the paperwork through legal, procurement, and finance? 😆

But if you understand hype cycles, a 95% "failure" rate actually makes sense for emerging technology. Most pilots are probably testing tools in the trough or early slope phases. Of course they're not delivering immediate ROI.

The real question isn't whether AI is working. It's whether you're measuring the right things at the right time:

• Plateau tools: Measure efficiency and cost savings
• Slope tools: Track adoption and quality improvements
• Trough tools: Focus on learning and capability building
• Peak tools: Set expectations around experimentation, not results

Plus, as Paul Roetzer mentioned, the MIT study is not statistically valid due to the small number of orgs surveyed.

The Competitive Advantage

While your competitors are either writing off AI entirely or chasing every shiny new announcement, you can build real advantage by understanding where each technology actually stands and where they are going.

This matters for internal politics too. When executives ask about your AI strategy, you can give intelligent answers about which applications are ready for investment versus which ones need more time. That credibility matters when the next wave of AI tools hits peak hype.

It also helps with budget allocation. Instead of spreading money across random AI experiments, you can focus on tools that are ready to deliver while keeping smaller bets on emerging stuff. We've moved past the "just spend money on AI" phase of marketing.

The Long Game

The companies that will win with AI aren't the ones with the most pilots or the biggest budgets. They're the ones with the clearest picture of where each technology sits on its maturity curve, and strategies that can adapt accordingly.

The alternative is discovering breakthrough AI capabilities six months later and wondering how much competitive advantage you missed.

I know companies tend to be risk adverse. But it's more risky to wait and see how things shake out.

How are you tracking AI maturity in your organization? Do you have a system for evaluating which tools are ready for investment, or are you just experimenting with whatever gets the most buzz?

Thanks to Courtney Trudeau for sharing the ChiefMarTec AI report.

I'm running a little survey. This week, I'm asking 4 simple Canva questions regarding your use.

You can view the survey here.

Each person who completes the survey and leaves their email address will be entered to win one of three digital copies of my book, The Visual Marketer.

If you missed either of the previous surveys, you can find them here:

• Part 1
• Part 2
• Part 3

Thank you for helping!