Welcome back to the Launch Key 🚀

One of my current projects is purging old files to create one clean Notion knowledgebase. It includes newsletter topics, product ideas, personal reference material, books and old work history. I’ve kept summaries, notes, case studies, business stories and am slowly throwing away detrius while finding interesting connections.

A topic that keeps resurfacing is Innovation. I’ve adapted best practices, changed my attitude along the way and have written about it here before. But it’s time for an update.

This is the first in a series on innovation — examples, what works, what doesn't, and why the playing field has shifted in a direction that finally favors people like you.

Let's get into it.

Two years ago I wrote an issue on Innovation titled Don't Listen to the Bozos.

It got solid feedback. A few of you replied with your own corporate-innovation-lab horror stories. One of you sent a photo of actual yellow sticky notes on glass wall. One of you told me to write a series on it. I have updated it, but tabled the series until now.

Then I spent the last few months creating the knowledgebase of my own digital history — Evernotes, links, Vialogix case studies, product features I helped ship at Empowered Benefits, data-product redesign at EDA — and I realized I'd buried the lede.

The Bozos issue (and Guy Kawasaki) was right about what doesn't work. Brainstorming on demand, designated innovators, quarterly innovation pushes — all theater. Still true in 2026. More true in 2026 actually, now that "AI innovation lab" is the new sign on the same room.

What I didn't say then is what does work. I had a thirty-year answer of examples I didn’t use.

The corporate innovation problem hasn't gone away. It's just got new tools.

In 2018 the Bozos were running design-thinking workshops. By 2024 they were running hackathons. In 2026 they're running AI accelerators — same room, same yellow stickies, now with a Claude tab open and a vibe-coded demo by Friday.

The instinct underneath is identical: we need innovative ideas for product so-and-so by next quarter, and (insert latest) shiny tool will get us there.

It won't. Not because AI isn't powerful — it absolutely is — but because AI lets you build the wrong thing faster.

That's the 2026 trap nobody's naming clearly enough. Vibe-coding a feature in an afternoon does not mean the feature is what your customer needs. The 24-year-old shipping at speed has no idea whether what they shipped removes friction or adds it. They just know it works in the demo.

Adding features, by the way, is the least valuable of Doblin's Ten Types of Innovation. The frame I borrowed from Larry Keeley in the Bozos issue still holds. AI didn't invalidate it — AI just made the bottom-right corner of that matrix cheaper to operate in, while the more valuable corners (Profit Model, Network, Channel, Customer Engagement) still require something AI can't fake…

Judgment about what's worth building.

At Vialogix, we built a usability lab with a one-way mirror at our Charlotte offices in 1997 — at a time when most web design agencies were still pushing digital brochures. I admit we did it to differentiate from other agencies.

But the lab let us literally watch customers use our designs in real time. It led us from pretty HTML to user interface design and ultimately application development. But those interactions taught us that ‘design’ needed to not get in the way of customer needs.

We learned something every week that the client's VP of marketing did not know.

Creatas was a stock-photography redesign — we shipped a custom "lightbox" that surfaced in customer interviews, and sales went up 147% because we removed friction for customers. The redesign won a SXSW award and ended up in a Nielsen Norman ROI case study.

EDA was a data-products redesign I led in the mid-2010s. Discovery interviews showed us customers didn't want more data — they wanted easier access to the data they were already buying. We redesigned around that. 20% sales lift in year one. And 25% more in year two.

Different industries. Different decades. One repeating process.

I bought my first Mac in 1987.

I was working as a headhunter at the time — a job that did not, in any meaningful sense, need professional layout tools. But my Mac came with PageMaker, Illustrator and HyperCard. Suddenly I could organize all my paper files into a simple digital database and print candidate one-pagers that looked like a finished product.

That Mac sits at the head of every career move I made afterward. Grad school in 1989, specifically to put me near the creative process. Corporate communications at First Union, where I started as an unpaid intern. Vialogix in 1995. Empowered Benefits in 1999. EDA. EnergyXchain. National Science Foundation grants. Blockchain patents. AI agents. Everything points back to that little Mac.

And that Mac existed because a guy named Don Valentine had bet on Apple a decade before my fateful purchase.

I recently listened to the Acquired podcast's deep-dive on Valentine and Sequoia Capital (Free Knowledge below).

Here's what struck me. Valentine didn't invent semiconductors. He didn't write Apple's operating system. He didn't design the Mac.

What he did was spend two decades in sales-and-marketing roles at Fairchild Semiconductor and National Semiconductor — the foundational chip companies of the early Silicon Valley era. He learned to read markets. He learned what big inevitable demand looks like before anyone else can see it.

Then in 1972, he founded Sequoia Capital. Not because the venture capital industry needed another firm — it barely existed at the time. He built Sequoia because the role didn't exist yet and he was perfectly shaped to create it. He fit the job to himself. He didn't fit himself to a job.

The result: Apple, Oracle, Electronic Arts, NVIDIA, Cisco, Google, YouTube and a chain of subsequent bets that essentially funded the modern internet.

His advantage was not genius. His advantage was a body of pattern recognition that he'd built one chip-customer-meeting at a time over twenty years, then applied to a vehicle of his own design.

And it's the same move available to you.

AI is not going to replicate Don Valentine's read on the semiconductor industry in 1972. It is also not going to replicate the Launch Key reader's knowledge on whatever industry, customer base, or craft you've spent thirty years inside. Your particular pattern recognition — built from particular meetings, particular failures, particular customers — is not in any training set. It cannot be vibe-coded. It is uniquely yours.

The only question is what vehicle you build to deploy it.

Most people don't think of themselves as innovative.

Same way most people don't think of themselves as creative — until you point out they decorated their kid's birthday party from scratch last weekend, or solved a leaking gutter problem with a piece of pool noodle and some zip ties, or rewrote their resume three different ways until one landed an interview.

Creativity got democratized when the tools got cheap. Suddenly everyone could ship a flyer, a song, a video a great picture. People who'd been told their whole lives they "weren't creative" started shipping creative work, and the world got more interesting.

Innovation is going through the same transition right now.

The frameworks (Doblin's Ten Types, Business Model Canvas, lean startup) are free. The tools (AI for prototyping, no-code for building, LinkedIn or newsletter platforms for testing) are cheap. The case studies are everywhere — including, increasingly, in this newsletter.

What's not democratized — what AI cannot give a 24-year-old vibe-coder no matter how many tokens they burn — is the lived experience of watching real customers struggle with real products over multiple decades and across multiple industries.

That's the asset you're sitting on.

The pattern recognition you built by being in the room. The judgment Don Valentine built one Fairchild sales call at a time, applied to a vehicle nobody had asked him to build.

Next week I'll get into the how — what the actual process looks like, where AI earns its keep inside it, and what a one-person Launch Key reader can do next Saturday morning that an entire corporate innovation lab can't.

For now: the Bozos are still Bozos. AI didn't fix them. But the playing field has shifted in a direction that finally favors the operators who've been doing this work for thirty years.

Don't wait for someone to pick you.

Now go launch something 🚀

f you work in fintech or finance, you already have too many tabs open and not enough time.

Fintech Takes is the free newsletter senior leaders actually read. Each week, I break down the trends, deals, and regulatory moves shaping the industry — and explain why they matter — in plain English.

No filler, no PR spin, and no “insights” you already saw on LinkedIn eight times this week. Just clear analysis and the occasional bad joke to make it go down easier.

Get context you can actually use. Subscribe free and see what’s coming before everyone else.

Sign Up Free

🗃 Dealroom Business Success Uncovered : Learn directly from billionaire entrepreneurs on how to grow a business. Join a community of 2,000+ innovators.

📕 MGMT Playbook : Practical management insights straight to your inbox every Wednesday.

🤖 AI Report : 400,000+ business leaders (and teams at IBM, AWS & Zapier) start their day with The AI Report. 5 minutes. Plain English. No hype.

🔖 The StartUp Marketing Newsletter : Your cheat sheet for marketing news, insights & tips tailored for the startup space.