What Mid-Market Leaders and PE Firms Actually Need to Know About AI Right Now

Recently, I sat across from the Chief Information Officer of a mid-market company who told me something I haven't been able to shake. "I thought I'd retire before I had to deal with all this," he said. He's built a profitable business over decades, and now he's staring at a technology shift that feels, to him, like learning a new language while the dictionary keeps changing.

He's not alone. I spend most of my time talking with mid-market executives, operating partners at PE firms, and the people who actually run companies trying to figure out where AI fits. Then I attend AI conferences, where the energy is relentless and the predictions grow bolder by the quarter, and the distance between those two realities is enormous, and growing.

This article is about that gap, and what I think leaders who actually run companies should focus on right now.

The questions mid-market executives are actually asking

At industry events and in AI-focused circles, the conversations tend to center on the latest LLM updates, model architectures, agentic workflows, and autonomous systems. The executives I meet in the field are somewhere else entirely. Their questions are more basic, and more urgent: Which AI tools should we start with? How much will it cost? Is our data safe? Can we trust these systems with sensitive information?
These aren't unsophisticated people. They're experienced operators who know their industries cold. But AI is foreign territory for most of them, and the gap between what they understand and what they're being asked to decide on is growing fast.

The Stanford HAI 2025 AI Index found that 78% of global companies had adopted AI by mid-2025, up from 55% the year before. But adoption doesn't equal readiness. A Harvard Business Review survey of digital leaders heading into 2026 surfaced persistent issues around human and organizational preparedness, even among companies reporting measurable value from AI investments. The gap between "using AI" and "being ready for AI" is wide, and in the mid-market, where internal tech teams are smaller and budgets tighter, it's even wider.

Here's what makes the mid-market situation distinct. A CEO who has spent 25 years running a logistics business knows exactly when a new truck pays for itself. He can calculate the ROI on a piece of equipment down to the month. But when someone tells him to invest in an AI implementation, he has no frame of reference for what success looks like, how long it takes, or what happens if the tool he chose six months ago gets leapfrogged by something new. That uncertainty creates paralysis, and paralysis is expensive.

The problem isn't a lack of interest. It's a lack of translators. People who can take what's happening in the AI world and turn it into decisions that make sense for a healthy, running company doing hundreds of millions in revenue and systems that were built fifteen or 20 years ago.

Why PE firms are rethinking their software bets

The private equity side of this story has its own set of pressures, and they've intensified sharply over the past year.

For the last decade, software was one of private equity's favorite asset classes. SaaS businesses offered recurring revenue, high margins, sticky customers, and clean growth metrics. But AI is forcing a reassessment of that thesis, and fast.

CNBC reported in March 2026 on talks between Anthropic and Blackstone to form a joint venture that would embed AI across PE portfolio companies. The partnership makes strategic sense: AI can drive efficiency and value creation across a portfolio. But here's the tension. The same AI capabilities that PE firms want to deploy are also threatening the value of their software investments. If AI models can approximate what many horizontal SaaS tools do (project management, CRM, basic analytics, HR workflows), then the companies PE firms are partnering with may be devaluing the very assets they already own.

The scale of exposure is significant. According to J.P. Morgan Asset Management, software represented 18% of PE deal value in 2025, up from a 14% average over the prior decade. And as Long Angle documented in its Q1 2026 report, private SaaS transaction multiples compressed from 6.7x EV/Revenue at their 2021-2022 peak to 3.1x by the second half of 2025.

At a private dinner we hosted in LA last month, I heard a version of this concern from several PE investors: "How do you underwrite a five-year deal if you can't predict what happens to software in the next one or two?" It's not that they don't understand AI. It's that even those who do can't see far enough ahead to bet confidently. So some are pulling back. The investors who entered tech because returns were strong, without deep technical conviction, are returning to what PE has always done well: buying and improving businesses with real physical complexity (transportation, manufacturing, healthcare operations, legal services).

But here's the twist. Those same investors now recognize that the portfolio companies they're moving toward are exactly the businesses that could benefit from AI the most. Not AI as a product, but AI embedded into how they operate, price, schedule, and serve customers. The question shifts from "Should we invest in an AI company?" to "How do we bring AI into a trucking company, or a veterinary group, or a regional lender?"

That's where things get practical, and where the knowledge gap becomes a constraint.

The real bottleneck isn't technology

There's a line I keep coming back to: the technology is moving faster than people can absorb it. I don't mean that abstractly. It's an operational fact that affects timelines, budgets, and outcomes.

Harvard Business School professor Raffaella Sadun, speaking on MIT Sloan Management Review, framed this in historical terms. During the ICT revolution, the companies that pulled ahead weren't the ones with the best hardware. They were the ones that reorganized work around the new tools. The firms that couldn't adapt organizationally fell behind permanently, and that gap, she argued, never closed. Her concern about AI is the same: the organizational adaptation is enormous, and most companies are neglecting it. May Habib, CEO and cofounder of Writer, put it bluntly at HumanX in San Francisco a few weeks ago: 'Change management is f***ing hard, and anyone who tells you differently never experienced it.'.

A separate Harvard Business Review analysis from February 2026 reinforced this point, finding that AI initiatives often stall because employees' anxiety about relevance and job security drives surface-level adoption without real commitment. The problem isn't that people refuse to use the tools. It's that they use them tentatively, without integrating them into how work actually gets done.

I see this constantly. A company deploys an AI tool, announces it in an all-hands meeting, runs a training session, and then wonders why adoption plateaus. The mistake is treating it like installing Slack. AI changes workflows, decision-making patterns, and reporting structures. People need time to build confidence, and that time doesn't shrink just because the tools improve every few months.

I sometimes compare this moment to when personal computers entered the workplace 30-plus years ago. Companies that resisted didn't just fall behind temporarily. Many never caught up. The difference now is speed. MIT Sloan's Thomas Davenport and Randy Bean observed in their 2026 AI trends analysis that organizations change much more slowly than AI technology does, making the gap wider every quarter.

The growing gap between technology capability and organizational readiness

That gap is the real bottleneck. Managers who don't know what to ask for, teams that haven't been trained, processes that haven't been redesigned. And unlike the technology, you can't accelerate people with more compute.

Four things PE operating partners and mid-market CEOs should evaluate now

Given everything I'm seeing in the market, here's where I'd focus. These come from the questions I get asked most often and the patterns I see in companies that are making progress.

Evaluate your business's defensibility against AI disruption. If a foundation model vendor ships a new feature tomorrow that replicates your core product, what happens? Said by a tech PE investor to me last month: “businesses that hold up best are those where value lives in domain complexity, customer relationships, regulatory expertise, and physical-world operations”. At Making Sense, we help private equity firms run this kind of assessment as part of tech due diligence, specifically to understand where technology risk sits.

Assess your team's readiness to work alongside AI. Can your people build automated workflows? Do they know when to trust an AI recommendation and when to override it? A U.S. Chamber of Commerce workforce report found that among mid-market firms facing staffing challenges, 61% planned to invest in AI tools, but the skills training to make those investments productive is still the missing piece. We've built an AI Enablement & Learning Program for exactly this reason.

Stress-test how future-proof the business is. What's custom and expensive today might be automatic and low cost, even free in six months. Which parts of your operation are at risk of being commoditized by the next wave of AI? Which parts become more valuable because they're harder to automate? The World Economic Forum has noted that mid-market businesses often have weaker IT infrastructure than larger competitors, making them more exposed to disruption but also positioned to gain more from focused adoption.

Prioritize speed to value. The biggest mistake I see is over-scoping. Companies build comprehensive AI strategies that take months to plan and more months to execute. By the time they ship something, the landscape has shifted. The goal should be to demonstrate measurable impact in weeks, not months. Give teams early wins that build confidence for the harder changes ahead.

Why waiting is the most expensive decision right now

Every time I sit in a meeting where the conclusion is "let's revisit this next quarter," I think about the compounding cost of that delay. Not just in technology terms, but in human terms.

AI tools are going to keep improving. But the organizational capacity to use those tools has to be built deliberately, and it takes time. People need to get comfortable with new workflows. Managers need to learn how to evaluate AI-assisted work. None of that happens if you wait until the technology "settles down."

The FTI Consulting 2026 Private Equity AI Radar, surveying 200 fund and operating leaders, found that 95% of funds reported AI initiatives meeting or exceeding their original business case. Yet adoption across portfolio companies remains low, with talent cited as the primary constraint by 35% of respondents. The value is there. The readiness isn't.

This is the message I deliver to every C-level and operating partner I meet: you don't have to solve everything at once, but you do have to start. The companies that begin building AI fluency today will be the ones that capitalize on the more powerful tools arriving next year. The ones that wait will find themselves trying to sprint when they haven't learned to walk.

At Making Sense, we've spent nearly 20 years helping mid-market companies and PE-backed businesses close the gap between what technology makes possible and what their organizations can execute. We've seen enough cycles (cloud adoption, mobile, digital transformation in PE portfolios) to know the pattern: the winners move early, move with focus, and move with a partner who understands their reality.

The AI conversation doesn't need more hype. It needs honesty about where companies actually are and who can help them get from here to there. If that's what you're looking for, let's talk.

Fernando Florez is the Chief Revenue Officer at Making Sense, where he works with mid-market companies and private equity firms to turn technology into measurable business value. He can be reached at hello@makingsense.com.