Almost exactly 10 years ago, Design a Better Business, a book that I was the lead author on…and one that I continue to be incredibly proud of, was released. With the luxury of (summer) time, I began to work my way through Design a Better Business again, chapter by chapter, marking up what needs to change for a potential update…or maybe even an entirely new book.

I went in knowing there would be a bunch of exciting stuff to add for prototyping, which AI has broken wide open. I also expected the case studies to be a big problem. Oh…they are! The accelerator section calls accelerators "the new hunting ground for scale," which was a very 2016 thing to believe and did not survive 2022. Autodesk's story is a decade old. Netflix and Qwikster happened in 2011, which is now closer to the Sopranos finale than to today.

Fine. Case studies age. Super simple to swap them out for shiny new ones.

What actually gave me pause was a two-page spread called Wall of Ideas. It tells you that at the end of a good ideation session it's not uncommon to have 500+ sticky notes on the wall. We wrote that as a win. That was the applause line…in 2016.

I read it last Tuesday and sort of just sat there, because I'd spent that morning getting forty distinct business model variants (for Traction) out of a chat window in about ninety seconds. Nobody stood up. Nobody uncapped a Sharpie. In fact…I have to walk to another room to grab a Sharpie! If the wall builds itself now, what exactly is the workshop for?

Read on below…

— Justin Lokitz

Here's the thing I want to be careful about: this isn't a mea culpa or "my old book was wrong" piece. Design a Better Business was very right for 2016 and still stands the test of time in many cases. However, The ground, as we ALL know, moved. And…not just a little. We’re talking Pangea-to-seven-continents scale shifts.

What’s more it didn't move under my book specifically. It moved under nearly every method any of us were trained on and that I wrote about a decade ago.

Just look at what all those dyed in the wool design methods quietly assume. A classic Design Sprint reserves an entire day for sketching, and Crazy 8s exists to force volume out of people who'd otherwise produce one safe idea. The Double Diamond's first diamond is divergence, treated as an act of deliberate effort. Osborn's brainstorming rules (i.e., defer judgment, go for quantity) got handed down through IDEO and the d.school and into basically every workshop deck you've ever sat through (or that I’ve ever created).

Every single one of those is machinery for manufacturing quantity. We built that machinery because quantity (and breadth) was genuinely hard to get. Getting fifty real options in front of a team used to cost you/me a room, a facilitator, six calendars, catering, and most of a Wednesday.

Now (hahah…nervous laugh) it costs a prompt.

I can sit with Claude and have forty positioning angles before my coffee's cold. I can hand Lovable, Framer, or Claude Code a paragraph and get a clickable prototype that would have been two sprints of front-end work in 2019. Figma Make will take a rough idea and give me something to react to in minutes. None of this is a hypothetical near-future thing. It's Tuesday, with time midday to walk the dog.

So the cost didn't just drop. It moved. This is the part I think a lot of design leaders are still under-reacting to. We keep celebrating the collapse of production cost as pure upside (look how much faster we can explore!) without noticing that the constraint just relocated downstream and landed on the part of the process we invested the least in.

The old scarce thing: getting fifty credible options in front of the team. The new scarce thing: knowing which of the fifty deserves the next six weeks of your team's life.

Those, my fellow design friends, are not the same skill. They're barely related skills. And we spent fifteen years building a profession, a workshop economy, and an entire genre of book (yay) optimized for the first one.

And there's a second-order problem the future never warned us about. A wall of AI-generated options is worse than a wall of human-generated ones in one very specific way, and it took me a while to name it.

When someone on your team spent twenty minutes sweating over a sketch, that sketch showed up to the conversation (or even better…a design critique) with an advocate. Somebody would defend it. Somebody would be quietly disappointed when it got voted off. That effort was largely both a commitment device as much as it was a production cost. It was skin in the game, and it made the conversation real.

Forty AI options have zero advocates. Nobody's disappointed. Nobody can tell you why option 17 exists, because nobody made option 17. You're not running a design conversation anymore, you're running a review of output that arrived orphaned.

The sticky notes were never really about idea supply. Effort was doing the filtering for us, invisibly, the entire time. Take the effort out and you take the filter out with it, and I don't think most teams have noticed that's what happened.

Which brings me to the tools we use to converge…and how badly outgunned they now are. Dot voting. Affinity mapping. 2x2s. The Innovation Matrix in my own damn book, which asks a team to plot ideas on cost versus revenue and pick the top-right quadrant.

Every one of those was designed to cut a pile of forty down. Not four hundred. And more importantly, every one of them was designed to cut down a pile where the effort filter had already done half the work before anyone picked up a dot sticker.

Dot voting on machine output is a popularity contest over things nobody authored, decided by whoever in the room is most confident before lunch. I've watched it happen. I've facilitated caused it to happen. No, really: I am extremely complicit in all of this. I wrote the spread in this best-selling book that told you a five-hundred-note wall was the goal. And…I have personally facilitated thousands of workshops to the same end.

So what's the workshop for? Interesting (at least to me), I don't think the answer is "nothing." I think the answer is that the two halves of every method we use just switched weight classes, and we haven't redistributed the time.

Divergence used to be the expensive half and convergence was the quick vote at the end. Now divergence is nearly free and convergence has to carry the entire load…with tools that were designed as a tiebreaker.

If you got here from "When Everyone Can Build, Taste Becomes the Advantage," this is the structural version of that same argument. Taste isn't a vibe. In a working team it has a shape, and the shape is criteria — the explicit, written, argued-over statements of what a good answer has to do, agreed before anyone sees the options.

That's the thing I most under-wrote in 2016. Design Criteria got two pages in the book, tucked next to storytelling, treated as a nice-to-have you'd fill in if you had time before lunch. It's the one tool in the entire method that gets more valuable as options get cheaper, because it's the only one that scales with volume. Forty options, four hundred, four thousand; a real criteria set handles all of them identically. A dot-voting session absolutely does not.

Next week (or the week after) I'm going to take that apart properly: what a criteria set actually looks like when it's load-bearing, why "delightful and on-brand" collapses the second you point it at machine output, and why I think criteria are becoming the artifact a design leader owns instead of the deck.

For now, the thing worth sitting with is this: the shortage is over. Nobody is short on options anymore, and nobody's coming back to that constraint.

We spent a decade teaching teams how to make the pile bigger. The next decade belongs to whoever can say, out loud and in advance, why anything gets to leave it. Onward!

Subscribe to Design Shift for more conversations that help creative professionals grow into strategic leaders.

If you’re reading this…or expecting a custom Post-It Note, leave some feedback above and I’ll know to add it back in next time ;-)