I still remember the piercing, digital screech of a 56k modem initiating its handshake with the future. While waiting for QuickTime VR demos to appear on my screen, pixel by agonizing pixel, I’d evangelize to anyone who’d listen about Siteman, a website builder with drag-and-drop functionality that felt like wielding actual magic. Siteman wasn’t just another software tool; it was the inflection point. The tool that would change everything.
Sound familiar?
Over the last 150 years, virtually every generation has coped with “it-changes-everything” technological advancements. The steam engine, electricity, radio, TV, telephony, nuclear power, computers, internet, mobile and social media, to name a few.
We’re doing it again with AI. Our challenge is to heed the lessons from our earlier innovation adoption.
The boom, the bust and the uncomfortable truths
The late 1990s felt electric in a way that’s difficult to convey to anyone who wasn’t there. There were tools that promised to democratize web design and publishing, and there were emerging technologies like QuickTime VR and streaming video(!). Technologies heralding a new age of digital creativity and commerce and were going to change everything.
And some did. Eventually.
Ecommerce exploded. Digital reach expanded exponentially. Access to information and tools was democratized. Mobile and streaming video achieved ubiquity. I held an iPhone in my hands and felt like the future had suddenly and all at once appeared.
But here’s the uncomfortable truth we gloss over as time marches on: we were spectacularly wrong about the timeline and embarrassingly naive about the slog involved in getting there.
Siteman? It didn’t die for lack of vision. Wix and Squarespace prove that. Siteman failed because the infrastructure wasn’t ready, the market wasn’t educated and the business model hadn’t evolved past “build it and they will come.”
QuickTime VR delivered jaw-dropping demos that made everyone nod enthusiastically in conference rooms. But it promptly failed in the marketplace because nobody had figured out the boring stuff: browser plugins, bandwidth, content creation workflows or why a real estate agent would completely restructure their sales process around a technology their clients couldn’t reliably access.
The pattern repeats:
- Revolutionary technology is greeted enthusiastically by early adopters, investors and the press.
- Tepid adoption by the masses because the “boring” stuff hasn’t been addressed.
- Years of quiet iteration before the breakthrough finally arrived in a form we barely recognized.
AI is different. Really!
Fast-forward to today, and the parallels are almost comically identical. Massive investment rounds, breathless “this changes everything” headlines and a collective confusion between impressive infrastructure (OpenAI, Anthroptic, Google, Perplexity, etc.) and practical use cases that actually matter to real businesses.
But here’s where it gets interesting: AI differs in three demonstrable ways.
First, like the great John Madden used to say, “Speed kills.” The speed of adoption and improvement is genuinely unprecedented. Product cycles that took years in the dot-com era now happen in months. The feedback loop between idea, implementation and iteration is compressed to the point where traditional strategic planning feels almost quaint.
Second, the breadth of impact is wider than anything we’ve seen. The dot-com boom primarily affected technical teams and digital-first businesses. AI is touching every role, every department and every decision-making process. Your accountant is experimenting with ChatGPT. Your grandmother is asking ChatGPT increasingly complex questions.
Third, the friction has nearly disappeared. No more waiting for IT to provision servers or budget cycles to approve software purchases. Cloud infrastructure, API-first architectures and plug-and-play integration mean adoption curves that would have taken years now happen in weeks. What’s limiting adoption? Only the imagination of users and the paranoia of lawyers.
So, is Generative AI today’s QuickTime VR? Is it more than impressive demos that flame out? Or are we witnessing the birth of something that rewrites the foundation of how we work and decide?
Probably both.
The myopia trap
Here’s what we consistently get wrong: we mistake tools for outcomes.
In the dot-com era, we fell in love with banner ads before we understood conversion optimization. We built CRM systems (remember Act! and GoldMine?) that promised to revolutionize sales relationships. Then, we watched sales teams ignore them because we’d solved the data problem without addressing the habit problem.
The cognitive trap is always the same. Some humans experience technological change as exponential and emotionally gratifying, so we assume the corporate cultural and organizational experience will be identical. It won’t.
Culture changes linearly, habits change incrementally and organizational inertia is immutable. This feature of successful organizations enables them to repeat success. If you aren’t familiar with Martec’s Law, read up.
The most successful technology adoptions don’t feel revolutionary when they happen.
Salesforce didn’t “win” because it was the most technically sophisticated CRM. It won because it made the sales process feel familiar while quietly transforming everything underneath. Amazon didn’t disrupt retail by reinventing shopping. It made shopping easier, faster and more convenient than the alternative.
Want a lesson in disruption? Look no further than Clayton Christensen, RIP.
From believer to builder
My journey from wide-eyed dot-com dreamer to person who has seen this movie a few times (dot-com, mobile, social, etc.) taught me that the most perilous moment in any technology cycle is when you think you’ve figured it out.
Software adoption trained me to be skeptical of feature lists, over-the-top promises and shiny objects. It taught me to focus on business outcomes. I learned that companies that survive disruption aren’t the ones with the best technology. They are the ones with the best understanding of why people resist change and make new approaches feel inevitable rather than intimidating.
If this sounds pessimistic or like a Luddite perspective, realize that my viewpoint is hard-won over decades of working in tech. It’s the result of watching too many “revolutionary” tools fade away into the night.
After seeing that movie over and over again, you ask different questions. The question is not “What can this do?” but “What will this actually get people to do differently, and why would they want to?”
Or, more simply, “Should we do this?”
The martech hype filter
Where do you go from here? All of the martech vendors you work with or who want to work with you are screaming about AI. If you put AI aside for a moment, how do you make actual decisions to move ahead?
I propose a people-oriented filter to guide marketing software adoptions: the Martech Hype Filter. Use it by answering these three questions:
- What will the use of this software change in human or organizational behavior? Not might change, not should change, but what will it change, given how people and organizations really operate? The difference between these questions strangled more brilliant marketing strategies than lack of funding ever did.
- Which part of our marketing value chain will be disrupted by the software, and how? Disruption is specific, not general. Email marketing automation disrupted postal direct mail operations, and marketing automation platforms disrupted lead qualification processes. AI will disrupt specific workflows and automate human decision-making, not “marketing” as a broad category.
- What optimistic assumptions am I making that haven’t been stress-tested? That’s the killer question. List every assumption your AI strategy depends on, then aggressively try to break each one. Assume your team will resist. Assume the technology will be 50% less reliable than advertised. Assume your customers won’t care about the improvements that excite you.
The tools that survive this filter aren’t necessarily the most impressive but are the most inevitable.
Bottom line: The hard part is the humans
Maybe AI marks an inflection point in how we work, create and decide. Maybe this time really is different. But the lesson from the dot-com boom era isn’t that we were wrong to be excited. The lesson is that excitement without introspection leads to delusional fantasies at best, and career and existential business challenges at worst.
Companies that thrived through the last transition weren’t the ones that moved fastest or spent most aggressively. They were the ones that learned, adapted and remembered. They understood that revolutions always come slower and faster than you expect—slower in their cultural impact and faster in their technical maturation.
Siteman taught me to fall in love with the why behind the technology, not just the wow. The dot-com boom taught me that the most critical question isn’t whether you can spot the wave, it’s whether you can stay upright when it finally crests.
The marketing leaders who thrive in the AI era won’t be the ones implementing every new tool or chasing every shiny object. They’ll be the ones who remember that technology is never the hard part. The hard part is always the humans.
Special thanks to Chris Elwell, Mike Chin, Jacob Sanders, Mike Pastore and Scott Brinker for their editing, perspective and thoughts.
Fuel up with free marketing insights.
MarTech is owned by Semrush. We remain committed to providing high-quality coverage of marketing topics. Unless otherwise noted, this page’s content was written by either an employee or a paid contractor of Semrush Inc.