The “90-Minute CMO” Is a Lie

Let me tell you the thing I've been getting wrong.

When people ask what AI has done for my work, I usually say some version of: "I compressed a 6-hour workday to about 90 minutes." Which is true. I run two marketing operations — one a high-volume SMS fundraising practice, one a senior transition and real estate company — and the time I spend on each has dropped by roughly 70% since I started building AI workflows in earnest about 18 months ago.

The reaction I get is usually some combination of envy and skepticism. "90 minutes? Really?" And then the conversation goes down the productivity rabbit hole: what tools, which prompts, how exactly, can you walk me through it?

And that's where I've been making the mistake. Because the 90-minute story isn't about productivity. It's about leverage. And those are not the same thing — not even close.

What I Mean When I Say "Leverage"

A few months ago, I was sitting on my couch. Kids on me, TV on in the background, the usual chaos. A stray thought surfaced: How many real estate deals did we lose last year where the house eventually sold to someone else anyway?

I described the question out loud to my phone. Three minutes later — I actually timed it — I had a dashboard. 332 closed-lost records from our CRM, cross-referenced against public property sale records. 22 confirmed sales we didn't close. Combined transaction value: $36 million. Interactive map, rankings, links back to Salesforce.

Kids were still on me. I hadn't moved.

Now — here's the question you should ask: Did that save me time?

Sort of. Before AI, I could have asked a data analyst to run that analysis. It would have taken two days minimum, assuming we even had that person. But here's the more honest answer: I never would have asked for that analysis at all. The friction cost — defining the task, writing the brief, waiting days for a result, reviewing a static spreadsheet — was high enough that the idea would have died in my head before reaching anyone else's.

The $36M dashboard didn't save me two days. It made possible something that was previously impossible. That's not efficiency. That's a category shift.

The Lie Hidden in the Efficiency Story

The reason the 90-minute framing is dangerous isn't that it's inaccurate. It's that it sends people optimizing for the wrong thing.

If you hear "AI compressed my workday to 90 minutes," you probably think about your existing workflows and start asking: where can I apply AI to do what I'm already doing, but faster? That's a reasonable question. But it's the second-order question. The first-order question is: what was I not doing at all — because the economics were wrong — that I should be doing now?

There's a principle that the direct-response world understands intuitively, but most marketing organizations don't: the constraint is rarely effort. It's economics.

Every sales team has brainstormed what I call "conference room tactics" — strategies that sound brilliant for about 45 minutes until someone does the math. The obituary monitoring play for estate services (monitor death records, reach families in the 72-hour window before real estate agents do). The court-records prospector (scan for competitor clients with active complaints — reach them before they shop around). The closed-lost resurrection engine (match every lost deal against subsequent transactions; re-engage at the right moment).

These weren't rejected because they were bad ideas. They died because the execution economics were wrong. Run the obituary scraper properly: you need someone monitoring feeds daily, a researcher for property data, a copywriter, mail coordination, response tracking. For maybe three qualified leads a week. The ROI calculation is a disaster before you write the first letter.

We built that system on a Saturday afternoon. $47 a month to run. It surfaces 3–8 qualified leads per week, unsupervised. The idea didn't get better. The constraint changed.

The Productivity Frame Doesn't Ask This Question

Here's the question the efficiency story never asks: if you compress a 6-hour workday to 90 minutes, what do you do with the other 4.5 hours?

The productivity optimization answer: get more done. Ship more campaigns. Write more copy. Handle more clients. The same type of work, more volume.

The leverage answer is different. The 4.5 hours you freed up isn't for doing more of the same. It's for thinking at a level you previously couldn't afford. For running analyses you previously couldn't justify. For building systems that replace repetitive human decision-making — including yours — at scale.

I've run SMS fundraising campaigns for political organizations for fifteen years. In that time, I've produced thousands of creatives, tested hundreds of angles, and built an intuition about what works. That intuition is valuable. What's also true is that I was spending roughly 70% of my time translating that intuition into finished output — writing briefs, producing variants, formatting sends, analyzing results.

The 90-minute day isn't me working less. It's me working at a different layer. The translation work is mostly delegated. What's left is judgment — which is the only part that was irreplaceable to begin with.

What "Campaign-Grade" Actually Means

In the first essay in this series, I wrote about what running digital for a presidential campaign taught me about operating under pressure. The three principles: speed-to-market, ruthless measurement, message discipline.

What I didn't explain clearly enough is what those principles are for. They're not productivity tactics. They're a framework for making leverage decisions.

Speed-to-market isn't about going fast. It's about keeping strategy as the only bottleneck. When execution is slow, you spend your cognitive resources managing delays instead of making decisions. When execution is fast — when idea-to-market is measured in hours — your attention can operate at the level of judgment instead of coordination.

Ruthless measurement isn't about tracking more metrics. It's about having a system that kills pet projects — including your own — based on data instead of confidence. On the campaign trail, we didn't have the luxury of gut-feel decisions at scale. Every dollar tracked to a conversion. Everything else was noise. AI makes this possible for any organization, not just the ones with a war room and seven-figure analytics budgets.

Message discipline isn't about consistency. It's about making your constraints part of the system — so the constraint doesn't require judgment every time someone produces a piece of creative. Merge fields, character limits, claims, urgency framing: encode them in the workflow, not the approval process.

These three things together are what I mean by Campaign-Grade. Not faster. Not cheaper. Structurally different — in a way that allows good judgment to operate at higher volume with lower coordination cost.

The Actual Claim I Should Be Making

When I say "90 minutes," the honest version sounds like this:

I restructured the way marketing work gets done — not to do more of the same work faster, but to operate at a fundamentally different level. The work that previously consumed 5–6 hours was mostly translation: taking judgment and converting it into finished output. AI now handles the translation. What's left is the judgment. And I can operate at the level of judgment for 90 minutes and produce the same output — often better output — that previously required a full workday.

That's a different claim. It's less catchy. But it points people toward the right question: what is the highest-leverage layer at which you can operate, and what would it take to get there?

The productivity frame asks: how do you do what you do faster? The leverage frame asks: what do you do when you're freed from doing what you do?

The first question optimizes your current workflow. The second question redesigns it.

A Note on Who This Is For

I write about this at Hello AI. Not for AI researchers or people building models. For operators — marketers, sales leads, business owners, CMOs — who are running real things and want to understand what AI actually changes, not what it's supposed to change in theory.

The AI productivity story is real. The time savings are real. But the most important version of this isn't the one where you get your existing work done faster. It's the one where you realize you've been producing the wrong outputs at the wrong level of abstraction — and now you have a way to fix that.

I'm not there yet, fully. I suspect nobody is. But the companies and operators who figure out the leverage version of this — rather than just the productivity version — are going to have a structural advantage that's very hard to explain from the outside.

That's what I'm trying to document, one build at a time.