I pulled the old Design a Better Business spread on Design Criteria back up this week. Two pages…yay. A short paragraph on why criteria matter, a fill-in-the-blank template, and a callout box about staying flexible. That's it. It sits right next to a spread on telling a good cover story, like the two belonged in the same weight class.

They didn't then and they really don't now!

This all started when I began to redesign a business model innovation course I have been teaching for the better part of 12 years. Like I do every year, I dug into the vision/goals and theme of this course, which are still quite sound. And…as I dove into the specific day-to-day curriculum I noticed that the first third of lessons are heavily weighted toward human-born ideation techniques along with a sort of “pick your favorite idea” as a next step.

My spidey sense immediately caused me to spiral a bit, as “picking favorites” when it comes to ideas is just plain lazy. Yikes. I personally don’t think that’s a facilitation problem. That's a criteria problem.

Read on below…

— Justin Lokitz

Last issue I made the case that the constraint in design work moved. Options used to be expensive, so the whole method, like sprints, Crazy 8s, walls of sticky notes, existed to manufacture volume. AI made volume free, and the bottleneck slid downstream to judgment.

Fair question, and I got it a few times: judgment based on what, exactly? This is the answer. Criteria are the mechanism. Not a mood, not a gut check, not "does this feel like us." An actual mechanism. Full f-ing stop.

Most teams that I have worked with for the last 25-ish years have adjectives that don’t really stand on their own as criteria. Delightful. On-brand. Intuitive. Feels right. Elevated. I want to be clear that these aren't stupid words, and the people who reach for them aren't being lazy. They worked, for a looooong time, because the field they were judging was small. Five sketches, a team that had spent eighteen months building shared taste together, a facilitator who could read the room and know that "on-brand" meant something specific to these six people on that day, in that room.

Adjectives are compressed shared context. They only decompress correctly if everyone in the room already agrees on what they mean.

Forty orphaned options blow that up immediately. Nobody built shared context with a model. Nobody spent three weeks arguing about what "elevated" means for this product, because nobody had three weeks, because the options showed up in ninety seconds. You're left applying a vibe at a volume vibes were never built to handle, judged by whoever's most confident before lunch.

A criterion is a test rather than a value statement. The gap between the two is the whole issue. A value statement describes a mood. A test tells you what fails.

"We want this to feel premium" is a mood. After all, what is “premium”? Is Stella Artois a premium beer? Only if you’ve ever had Bud Lite as a choice (sorry/not sorry for the beer analogy…it’s hot here in Bend, Oregon). Likewise, every single one of your forty options can plausibly claim it, because nothing about that sentence can say no to anything.

"The unboxing has to work in under 15 seconds without instructions" is a test! Half your options just failed, right there, and nobody had to be the smartest person in the room to make that call. The test did it the work for you.

That's the line I'd draw for you if you're trying to tell a real criterion from a decorative one:

Practically, I don't think a real set is one criterion. It's usually three or four doing different jobs, and most teams are probably missing at least half of them:

A viability line. Does this actually make the business more money, or does it just look good in the deck? This is blunt on purpose. Most teams assume someone else owns this question and end up with nobody owning it. Designers (especially) cannot afford to make this mistake any longer.

A user-truth line. Does this solve something real, or does it solve something a synthetic persona was happy to agree with? I'm not joking…and if you read last month's teaser on AI-mediated research, you know why that line's getting sharper, not softer.

A feasibility line. Can you actually ship this with the team and the timeline you have, not the one you wish you had.

A cost-later line. What does this option owe somebody down the road who isn't in the room right now; this could be a user, a regulator, your own future team cleaning up the mess. This one's likely new for most teams, though it’s one that many product managers know all to well. It didn't used to need a line of its own because velocity capped how much damage you could do before someone noticed. Velocity doesn't cap anything anymore.

Four lines, not forty adjectives. Each one has to be able to kill an option on its own.

I didn't just want to argue this as an academic. I wanted a place to write it down so that we (all of us) could test it in the wild. So I went back to the Must, Should, Could, Won't (MuSCoW) canvas I've used since before this book existed and updated it for what it actually has to survive now. Same four boxes, because the structure was never the problem. What's new: a lock-date field, so nobody's filling this in after they've already picked a favorite. A tag next to every line, V, U, F, or C, for which of the four lines above it that entry is defending, so a band full of empty tags tells you exactly where your team's judgment has a hole in it. And a pass/fail grid down the side, so you're running real options against real criteria instead of eyeballing forty things at once and calling it a decision.

You can download the PDF at the bottom of this issue ;-).

I know that someone's going to tell me this is just what strategy already does, and I believe they're half right. Good strategy has always run on something like criteria. What's changed, at least in my purview, is where that lived. It used to live implicitly, baked into who was in the room, what they'd fight for, what a good facilitator could sense without anyone stating it out loud. That was fine when the room itself was doing the filtering, because the room and the criteria were basically the same thing.

The room just isn't filtering anymore. The options don't pass through six calendars and a Thursday before they reach you, so there's no slow, human, argumentative process quietly encoding your standards along the way. Whatever used to live implicitly now has to live on paper, written before anyone opens the chat window, or it doesn't exist at all.

That's the actual argument for why this is a design leadership skill and not a strategy-team handoff. Design leaders have always been the ones translating fuzzy conviction into something a team could act on. Criteria are that same job, just moved earlier in the process and made a lot more explicit than most of us are used to.

I think this is the artifact now. For a long time the thing a design leader owned was the deck: the synthesis, the narrative, the "here's what we heard and here's what we're building." I do believe that still matters. But…the deck comes after the options exist. Criteria come before. They're the part of the job that decides what even gets a fair shot, and right now most teams have delegated that decision to whoever talks first.

Write the criteria before you generate anything. Make each one able to kill an option by itself. If a criterion can't say no to your best-sounding idea, it's not a criterion yet.

One thing worth sitting with before I let this go: criteria are only as good as what's feeding the judgment behind them, and I haven't even touched that yet. Good criteria still assume the research underneath is telling you the truth. That assumption's getting shakier by the month, and the thing lying to you these days doesn't sound like it's lying. It sounds like a research summary with quotes attached. That's next!

PLEASE send this to somebody who needs to read it. If your team just watched a decision get made on vibes because nobody wrote down what "good" meant first, forward them this one. If you know another design leader watching that same twelve-minute vote happen in their org, send it their way too. This argument works better out loud, in a room, with the canvas open, than it does read alone. Also, I selfishly would LOVE (x10) some real-time feedback on this new canvas. Please send it my way at [email protected].

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