The Three Pillars of the Autonomous Enterprise

Shimmy says, shimmy says, shimmy, shimmy, shimmy says, ask me almost anything Hey everyone. " It's great to have you on here. I think I mentioned on my last "Shimmy Says," on the way home from our trip in Europe, stopped in New York, and I caught The Boss, Bruce Springsteen, up there.

I love Bruce Springsteen. Let me give you a little Bruce Springsteen here. We'll talk about the promised land.

Everyone wants to get to the promised land, it seems, when we're talking about the autonomous enterprise. By the autonomous enterprise, I mean an organization that's fully leveraging everything that AI, agentic AI, or automation can give us to really fundamentally change the metrics of your business. So everyone wants to get to this promised land, but no one seems to have the map.

Everyone I speak to, it's fuzzy, squishy marketing words. But I was in Dallas yesterday, actually, and I finally spoke to someone who showed me the map, and I saw that big fat X right there. And like Bruce Springsteen says in that song, now I believe in a promised land.

Before I go any further, let me just admit something up front. The event I was in in Dallas was the Imagine event from Automation Anywhere, and this is my second year covering it. Last year, I met their CEO, gentleman named Mihir Shukla.

Mihir is one of the most intelligent people I've had the pleasure of speaking to in a long time, and he's a heck of a nice guy. Well, last year, Mihir started talking to me about something they were calling APA, autonomous process automation. Think of taking BPA and adding AI and agentics to it.

And he told me that by the time this event this year, APA and agentics would be widespread in use. And I thought directionally he was right, but I thought his timeframe was off by two or three years. So I didn't doubt his vision per se, but I doubted his timing.

I just wasn't convinced that enterprises were operationally ready for it, and they were going to adopt it that quick. I didn't think the technology would mature enough. Frankly, I wasn't convinced organizations understood how much would actually need to change internally to make it successful.

Well, here we are a year later, and I don't think any of us are doubting anymore. It's just the way it is. And now this year, this week, Mihir hit me with what I think is one of the clearest operational frameworks for the autonomous enterprise that I've heard from anybody in the AI world or market so far.

The reason it stood out for me was exactly what I said earlier. It wasn't fuzzy, it wasn't squishy, it wasn't marketing speak. Most AI transformation conversations that I'm involved in today, they sound like consultant Mad Libs.

Remember Mad Libs? I used to love playing that game in the car when I was a kid. But they throw around words, digital transformation, AI native, AI native organizations, future-ready enterprises.

Just a lot of word salad, abstract nouns, very little clarity. But Mihir, he really boiled it down for something that even I can understand. Three simple pillars, and that's what he bases his whole thing on.

You may agree with them, you may not, but this is the first time I've seen someone lay it out so concisely, so logically, that it makes sense to you. Let me go through those pillars with you right now. The first pillar is this: automation.

Right? Almost all organizations I know have been trying to automate so much of our work because at automation, we can do it at scale, we can do it faster, better, cheaper. But sometimes we automate 10% more than we had before, and we pat ourselves on the back and say what a good job we're doing.

Mihir says we need radical automation. We need to automate 80% of the tasks that an enterprise performs. But notice this, he didn't say 100%.

He didn't say to automate everything. Humans still, as we say, have to stay in the loop. Humans still have to govern.

They've got to manage. They got to be there to make the judgment calls, the corner cases. But 80%, that's a substantial, that's a critical mass of what we're doing needs to change and be automated.

Because at that point, automation stops being optimization and really becomes the new operating model for the business itself. As I said before, most companies today are still treating automation like side projects. We do a workflow here, a chatbot there, maybe a little productivity boost in customer support and help desk or internal IT, but that's not what Mihir is talking about.

This is orchestration at an enterprise-wide scale. We've all maybe yearned for that, talked about it before, but this is where this whole agentic age comes in. Because with agentics, systems become different from previous generations of automation technology, BPA and stuff like that.

Agents can increasingly move across workflows, systems, and operational processes in a way that older automation tools simply couldn't. And so now we could really automate the heck out of things. Like in the movie "The Martian" where they're going to science the s**t out of it, well, we could automate the heck out of it now, tooAnd it's possible.

But it's not just IT that needs to be automated. We got to automate finance. We need to automate operations.

We need to automate sales, marketing. It's organizational-wide automation that we need here. So that's pillar one.

Let me get to pillar two, and it may be even more important than automation. We need to three X every employee. Now, mark what Mihir said.

He didn't say do layoffs. He didn't say replace employees. No, we need to make them three times more productive.

And honestly, it's one of the smartest parts of this framework because it cuts through two different nonsense noises at the same time. First of all, it rejects the idea that AI is about headcount reduction. Mihir never mentioned headcount reduction.

Second, it kind of makes real the Silicon Valley kind of fantasy world obsession with the mythical ten X employees. We're going to ten X people. Three X is almost doable across a workforce.

But it would be one of the biggest operational transformations in modern enterprise history at three X. Ten X is sort of fantasy land, in my opinion. And so I think three X is something to shoot for.

But Mihir made another point to me that I think executives, founders, entrepreneurs need to hear. Don't be satisfied with a 5% productivity gain or a 10% productivity gain because those are not transformation. That's optimization theater.

Companies that are spending enormous amounts of money to slightly improve an old operational model while pretending they're reinventing the future, they're not going to become autonomous enterprises. Real autonomous enterprises change leverage fundamentally. They're leveraging change fundamentally.

They are getting not five, ten percent, but 300%, and that's what we've got to shoot for. That means agents handling operational execution as part of that 80% automation. Automation layers removing repetitive work, machine-driven orchestration, decision support systems operating in real time.

Not just better meeting summaries and prettier dashboards. This is important. We're not going to get there in a snap.

We're not going to get there overnight. But it's a goal to shoot for. Three X your employees.

Make that the goal. Let me get into the third pillar. And honestly, this may be the one with the longest interim implications of all, especially for us in the tech industry.

Mihir believes that the SaaS economic model itself breaks at the autonomous enterprise scale. And the more I thought about it, and the more he explained to me why, the more I had to agree with him and not argue against it. We've all gotten real used to our SaaS software.

We pay our monthly licenses. We pay gigantic licensing fees based upon how many seats we bought, how many CPUs we're using, licensing fees that frankly don't make sense in an agentic future. Because if the agents are doing the work, what difference does it make how many seats I have or dashboard views I have of a particular product?

In fact, dashboard views themselves are becoming obsolete. Just the economics of the SaaS model don't make sense. For years, we've all accepted that bigger software stacks automatically mean progress.

Another subscription here, another workflow platform there, a couple of SaaS, we just pay it monthly. It's not as hard to swallow. But when you look at it now, and Mihir made this point, entire IT budgets are getting consumed simply maintaining software ecosystems, or as he said, just keeping the lights on.

80%, 75, 90% of our IT budgets are just keeping the lights on, paying the vig on our SaaS products that are just not built, and not only not built, they're not built for this new age that we're in. If you've been around SaaS and enterprise IT long enough, you know that this is the truth. We all pay the vig on our SaaSes knowing-- I can't tell you deals we've done where I know they're ridiculous, but we have to live within the SaaS boundaries of billing.

So we kind of live a fiction. But when you're paying 80% of your budget to keep the lights on, where is there room for real innovation, for R&D, for growth? So the argument here isn't simply that AI can recreate SaaS applications.

I know some people talk about the SaaS apocalypse, say, "No, we could just have our agent recreate the SaaS software. " It's the entire consumption model is changing once we introduce autonomy and agentics into our workflow. At this Imagine conference this week, I heard from several organizations who discussed enormous operational savings from automation-led redesigns and agentics.

I'm not talking five figures or even six figures. Not theoretical savings, real money being saved right now. One company even talked about nine figures.

And when you're talking that kind of money, that gets your attention. I should mention Mihir just came out with a new book this week. " He's the co-author of it anyway, and the premise is basically that over the next five years, we may compress a century's worth of business transformation into that incredibly small five-year window.

Think about it, 100 years of information in a five-year window. But I got to tell you, imagine the Automation Anywhere conference wasn't the only conference I went to this week. I was at KB4 last week.

I was at the Alteryx conference. I was at SUSE CON before that, and I'm going to be at Cisco Live, and this is where I'm hearing it. I'm hearing the same things over and over again.

I don't believe that it's just hype. AI infrastructure is exploding. Operational models are changing.

They're being disrupted. The economics of software are shifting right before our eyes, under our feet. Agents are moving very quickly from demos into production.

This isn't just another technology curve. This isn't just another wave. The whole team over at Automation Anywhere emphasized this with me, too, and I've heard it before.

This is a once-in-a-lifetime type of change that we're seeing. A lot of enterprises are out here saying, "Well, we still have time. We're going to wait and see.

" And I get it. No one wants to dip their toe in the water or make a mistake. But sometimes not doing anything is a mistake, because once operational leverage starts compounding inside autonomous enterprises, the gap between the leaders and the laggards, the people living above the poverty line and below it, tends to widen really fast.

I spoke about it last week in this AI Minute, right? That's how quick time is moving right now. Another thing that I appreciated hearing this week was the emphasis, though, on responsible AI and making these systems work for humans instead of the other way around.

Because that's the real anxiety around AI. I wrote about that this week as well. People booing Eric Schmidt at a graduation when he mentions AI, but who can blame him?

Jobs, control, human value are all on the table right now. Economic concentration, the disparity between the billionaire tech bro oligarchs and you and I. If autonomous enterprises are going to succeed long term, organizations need to have to figure out how humans remain central to the process, even as automation scales dramatically.

And that part matters, too. So here's my takeaway for this week's Shimmy Says, guys. Everybody says they want an autonomous enterprise and what it represents.

Very few people seem to actually able to explain what that actually means to them operationally. And this week, for the first time, I heard someone lay out a framework that felt concrete enough to actually build on. Automate 80%, 3X every employee.

Stop spending the majority of your enterprise IT budgets feeding outdated software consumption models. They're all simple ideas, but getting there is not easy. It's going to take work.

But if Mihir Shukla is even partially right, the implications are just for enormous, and your business can't afford not to do it. That's it for Shimmy this week. I'll see you next week.

Hey, for those celebrating Memorial Day, have a great Memorial Day holiday weekend. Remember what the holiday is really about, though. It's about those people who gave the ultimate sacrifice so that we can have freedom.

Don't give up your freedom. I'm Shimmy. I'm out.

Shimmy says, Shimmy says. Shimmy, Shimmy, Shimmy says. Ask me almost anything.