I get it. I’ve even said it. Buying a license for every employee feels decisive. It’s measurable, it’s fast, and it lets you tell your friends that you’re “doing AI transformation.” But handing every employee a chatbot and calling it transformation is like buying everyone a computer in 1995 and declaring yourself a technology company.

We’ve Been Here Before

When computers arrived in the workplace, companies didn’t win by buying everyone a desktop. When the internet became ubiquitous, the businesses that thrived weren’t the ones who simply built a website (though that seemed to be enough to raise venture capital in the late 90s). The winners were the ones who fundamentally rethought their industry, organization, and processes, then built capabilities that competitors couldn’t easily replicate.

The same principles that separated winners from followers in those eras are true today. The technology changes. Human behavior and sound business fundamentals don’t.

What made Amazon formidable wasn’t that its employees had access to the internet. It was that the company rethought what a bookstore could be, built proprietary systems, codified operational knowledge, and obsessively redesigned processes around the technology. What you build with the tools is the strategy. The tools themselves are table stakes.

AI is no different. And yet, here we are, watching company after company conflate access with advantage.

The Adoption Problem Nobody Wants to Admit

Even setting strategy aside, the “give everyone an AI” approach fails because most people don’t actually use the thing.

When you drop an AI assistant on someone’s desk, the most common reaction isn’t excitement—it’s a blank stare. People don’t know where to start. They try it once, ask it to summarize an email, shrug, and go back to working the way they always have. (Read: Canva Paid 5,000 People to Spend a Week Learning AI. They Couldn’t Figure Out How to Start.)

So you end up with the worst of both worlds: you’ve spent the money, you’ve announced the initiative, and six months later you have no measurable ROI and a workforce quietly convinced AI is overhyped.

There’s an enormous gap between AI adoption and AI value, and the gap is almost always human, not technical. (Kaufman Rossin’s State of AI in the Middle Market is worth your time here.)

Counter-Argument

The counter-argument here is worth acknowledging: some companies will say that early, broad access is the right starting point — that you need to let a thousand flowers bloom before you know which will survive. There’s some truth to that. Organic experimentation does surface unexpected use cases. But experimentation without a framework isn’t a strategy either. It’s hope. And hope doesn’t scale.

The Foundations of a Successful AI Strategy

A real AI strategy has to address four distinct dimensions. Skip one and you’ll have a very expensive Slack channel where people occasionally ask a bot to fix their grammar.

1. Prepare the People

Most of your employees don’t see AI as a productivity tool. They see it as the thing that’s coming for their job. If you deploy AI tools without addressing that anxiety directly, you’ll get passive resistance at best and active avoidance at worst.

Start with surveys. Understand where your employees are emotionally and practically. What tasks do they find tedious? Where do they feel AI would help versus threaten? This data shapes everything downstream — which use cases to prioritize, what training looks like, and how you communicate the change.

Then train. Actually train. Not a 45-minute webinar. Not a link to a vendor’s YouTube channel. Structured, role-specific training that answers the question every employee is actually asking: What do I do with this thing tomorrow morning?

Pro Tip: Host AI Office Hours each week to give employees the opportunity to showcase the problem they’re solving and get advice from the AI experts in your business.

2. Prepare the Processes

LLMs are increasingly good at understanding general business context. What they cannot intuit is your business: the unwritten rules, the edge cases, the “we never approve that without legal” knowledge that lives in your veterans’ heads.

This is process work. It’s not glamorous. It requires people to articulate things they’ve been doing on autopilot for years. But it’s foundational. An AI operating without this context is not your AI — it’s just the same generic model your competitors are also running. You’ve automated guesswork.

3. Secure Your IP, Build a Harness

Security, data privacy, and governance are core to the responsible adoption of AI. Developing an AI harness that offers easy access to vetted tools, along with protections for Personally Identifiable Information (PII), ensures that AI is used safely and effectively. This infrastructure not only protects the company but also instills confidence among employees and stakeholders. Without a harness, you have a fragmented ecosystem of personal tool preferences, inconsistent outputs, and genuine compliance exposure. With it, you have a platform your entire organization can build on confidently.

The harness is also where Continuous Quality Improvement lives — the testing, monitoring, and iteration that keeps your AI strategy calibrated as the underlying models evolve, as your business priorities shift, and as better tools emerge. Because they will emerge. And a strategy that only works with today’s technology isn’t much of a strategy.

4. Build the Company Brain

This is where competitive differentiation actually lives, and it’s the piece most organizations never get to because they stall out on the earlier steps.

The frontier AI models — the Claude and GPT and Gemini releases that make headlines — are available to everyone. Your competitors have access to the same model you’re using. What they can’t access is your proprietary data, your institutional knowledge, your accumulated context about your customers and your market. When you connect AI to that, it stops performing like a generic assistant and starts performing like your best employee—one who’s read everything your company has ever known.

• Proprietary data

• Institutional knowledge

• Documented workflows

• Historical decisions and the reasoning behind them

• Customer context accumulated over years

• The specific expertise of your best people, captured and made accessible

These are what transforms a general-purpose AI into something that performs on behalf of your business in a way that nobody else can replicate.

The companies building this now — systematically capturing and structuring their organizational knowledge — are creating something genuinely defensible. The companies waiting for the next model release to save them are building on sand.

“How do I Develop an AI Strategy?”

When your leadership team next discusses AI strategy, the question shouldn’t be “How do we get AI into everyone’s hands?” It should be ”What specific outcomes do we want AI to produce, for whom, and what needs to be true for that to happen?”

Those questions force the real work. It reveals where your processes are under-documented. It surfaces the employee anxiety that needs to be addressed. It identifies the proprietary knowledge that needs to be captured. It makes clear that tooling is the last thing to solve, not the first.

Access is not advantage. Giving everyone a tool is not a strategy. Asking the question above over and over until simple problems are solved and you start to see new capabilities emerge that simply weren’t possible before.

That’s how you differentiate your business.

That’s how you determine your strategy.

That’s not going to happen.

Not because the technology isn’t capable. The reason is something borrowed from chemistry: the diffusion rate.

The Diffusion Rate

In chemistry, a diffusion rate describes how quickly one substance blends into another. Pour cream into coffee and it doesn’t become a latte instantly. The molecules need time to intermingle, to find their way through the medium. That blending speed limits how fast any reaction can happen, no matter how reactive the chemicals are.

The same principle applies to AI, and it operates at every level.

Individuals need time to develop intuition about what AI does well and where it falls apart. Companies need time to rethink workflows, retrain teams, and rebuild trust in new processes. Entire economies need time to adapt regulatory frameworks, reshape labor markets, and redefine what “work” even means. Each layer constrains the next. A company can’t transform faster than its people can learn. An economy can’t restructure faster than its companies can adapt.

That cascading diffusion is why 12 months is fantasy. The technology may be ready. The world isn’t.

A Different Kind of Intelligence

Most AI predictions treat it like a faster human. It isn’t.

AI excels at pattern recognition across massive datasets, generating text and code at speed, and processing information that would take a team weeks to review. But it fails in ways humans never would. It hallucinates confidently. It lacks contextual judgment. It can’t read a room or sense when the “right” answer on paper is the wrong answer in practice.

These aren’t bugs. They’re characteristics of a fundamentally different kind of mind. Working effectively with a different kind of mind takes time and experience, both for individuals and for the organizations and systems they operate within.

Amara’s Law

Roy Amara put it well: “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”

In the short run, AI will not automate most white-collar jobs. The diffusion rate won’t allow it. In the long run, AI will transform how we live and work far more profoundly than most people imagine. But that transformation won’t look like mass automation. It will look like entirely new kinds of businesses, services, and capabilities that didn’t exist before.

Differentiation, Not Automation

Here’s what actually matters: AI is a new material. Not a replacement for existing materials.

When carbon fiber first appeared, manufacturers used it to make lighter versions of existing parts. The real breakthroughs came later, when engineers developed enough intuition about the material to design things that couldn’t have existed without it.

Most companies today are in the “lighter car parts” phase of AI. Using it to speed up existing processes. Saving money. That’s fine as a starting point, but it misses the bigger opportunity.

I’ve seen this firsthand. At Mark My Words Media, a marketing company we acquired last year, we started by using AI to accelerate analyst work. Faster research. Quicker reports. Incremental gains.

Then something shifted. As we developed deeper intuition for what AI could and couldn’t do, we realized it enabled an entirely new type of service: a holistic Revenue Flywheel combining SEO, paid ads, lead verification, and ongoing optimization into one integrated offering. Before AI, that level of service was unaffordable for most small businesses.

Since we launched it, inbound interest has surged. Not because we automated. Because we built something genuinely new. Something different.

That’s the real lesson. The path forward isn’t racing to automate. It’s using AI to become different in ways that weren’t possible before. The most successful people and companies in five years won’t be the fastest automators. They’ll be the ones who discovered new forms of differentiation that AI made possible.

What to Do About It

You have time. Use it wisely. Don’t panic. (And bring a towel. IYKYK)

Start experimenting now. Not to automate your job, but to understand this new material. Spend real time with AI tools. Learn where they’re brilliant and where they break down. That intuition is going to be your most valuable asset.

Then focus on what makes you different. AI will commoditize the generic, the safe, the average. Whether you’re an individual rethinking your career or a company rethinking your product, the question is the same: what can you build with this new material that couldn’t exist without it?

The diffusion rate gives you breathing room. Don’t waste it.

Then the Fed raised rates, growth multiples collapsed, and the entire VC scaffolding trembled. As CFOs everywhere yanked "growth at all costs" budgets, I stared at our fresh burn-rate spreadsheet and thought:

Why did we build every capability from scratch?

We could have bought profitable regional delivery firms—companies already cash-flowing—and infused them with automation, AI routing, and modern leadership. Same market impact, radically lower risk, infinitely faster path to scale.

That moment trashed 20 years of "Silicon Valley default." I stopped believing the next great tech story begins in a garage. Today, I'm convinced the biggest upside lives on Main Street—in service businesses ready for a tech metamorphosis.

Shortly after I began exploring this thesis, I teamed up with David Dripps, who brings a decade of experience running major teams at Meta, including Pages, along with a rare combination of banker's discipline and operator's intuition. In 2024, we bought our first company (with our own money) and have proven the effectiveness of this model. Now we are excited to share it with you.

The Valley's Great Delusion: Growth-at-All-Costs in a Cash-Costs-Money World

Silicon Valley perfected the art of subsidizing unprofitable hyper-growth with cheap capital. Between 2008 and 2021, money was nearly free. Valuations hinged on theoretical futures, not present profit.

But the post-ZIRP world plays by 1990s rules:

Cash has an opportunity cost. Equity dilution feels expensive to founders and investors. Public markets reward durable margins over abstract "land grabs."

Traditional VC remains vital for deep-tech moonshots that cannot bootstrap (fusion, quantum). Yet for the vast majority of software plays, the model now destroys more value than it creates. Capital efficiency—not "blitz-scaling"—has become the competitive moat.

The Productivity Paradox: Why Tech Hasn't Moved the Needle

We like to boast that software "eats the world," yet U.S. productivity growth limps along at 1.4%. How can that be? Doesn't cloud infrastructure and generative AI promise god-mode leverage?

The answer: distribution of gains.

Platform startups create enabling tech, but value capture happens only when someone applies those tools to gnarly, domain-specific problems. Your HVAC contractor in Ohio sees marginal benefit from AI until a tech-literate operator rewires their entire workflow. From sales and quoting to dispatch and preventative maintenance—everything needs to be rebuilt around algorithms and data.

That last mile of implementation is messy, boring, "unscalable", and absurdly lucrative for anyone willing to wade in.

Buy → Transform → Amplify: The Tinman Playbook

Tinman Group's playbook is simple but compounding. Each steps gives us another “bite at the apple” along a different dimension:

1. BUY - Acquire profitable service businesses at 6-8x EBITDA. Value Bite: Multiple Arbitrage

2. TRANSFORM - Install tech-native leadership, ownership culture, and AI/automation stack. Value Bite: Margin Expansion

3. AMPLIFY - Use the new data insights and capabilities enabled by our change to launch into adjacent products and markets. Value Bite: New revenue.

4. PARTNER (Bonus Bite!) - Where possible, co-invest into upstream software vendors that we rely on, then turn ourselves into lighthouse customers. Value Bite: Strategic Equity Upside

Case in point: Mark My Words Media.

A 20-year-old flat-fee digital-marketing agency serving law firms and local trades. Purchased at a single-digit EBITDA multiple with seller financing.

Our first moves were straightforward: implementing basic systems and rolling out professional management practices. These fundamental improvements laid the groundwork for what came next.

When our analyst quit, we didn't just replace them—we built an internal AI research assistant. This technology, combined with a strategic repositioning of the product, allowed us to reestablish key relationships and strengthen our market position.

The result? YoY revenue growth, rising margins, and a pipeline 3× larger than at acquisition. The transformation wasn't a single silver bullet—it was a methodical sequence of operational improvements emphasized by technological leverage.

Talent First: The Leadership Arbitrage No Spreadsheet Shows

Great tech culture isn't "foosball + AWS credits." It's a disciplined ownership model: clear goals, radical transparency, self-managing teams. We have learned—sometimes painfully—that transformation lives or dies with the general manager.

Going forward, we start each vertical search with the leader, not the target company.

Who are we recruiting? Experienced founders and startup VPs who have seen a thing or two and are craving real operating leverage. Leaders precisely like David—who built high-performing teams at both startups and Fortune 100 companies with a distinctive blend of tech fluency and strategic rigor. Our platform handles deal sourcing, financing, and the AI enablement toolkit. What these leaders bring matters more than any technological upgrade.

(If this describes you, please reach out!)

Why Investors Should Care

The capital structure we've built offers something increasingly rare: asymmetric returns with genuine downside protection.

At acquisition, we require a 15%+ unlevered yield. This creates a cushion that makes even base-case scenarios attractive. But our model isn't designed for singles and doubles.

When we layer in technological transformation, those 15% IRR base cases morph into 40%+ outcomes. We've watched it happen repeatedly:

• A scheduling algorithm reducing technician downtime

• A pricing engine capturing willingness-to-pay across customer segments

• An AI-powered recruiting funnel slashing hiring costs while improving retention

Each innovation compounds until the business hardly resembles its original form—except in its economic fundamentals.

We're structured as an evergreen holding company, not a forced-exit fund. Yields come from distributions, recapitalizations, and strategic exits – when timing makes sense, not because a prospectus dictated liquidity seven years from signing. This permanent capital approach lets us build for decades, not quarters.

Perhaps most compelling in today's macroeconomic climate? These businesses serve as inflation hedges. We sell concrete outcomes indexed to the real economy. When inflation rises, so do our prices, because we're selling necessities, not luxuries.

In short, we produce PE-style risk with VC-style opportunity without praying for IPO windows to reopen. It's a model built for the 2025 reality, not the 2021 fantasy.

Acquisitions are funded through a carefully calibrated capital stack: seller notes for motivated founders who believe in our vision, debt secured against predictable cash flows, and—starting Q4 2025—select outside equity for partners who want to ride alongside us.

Every deal begins with a leader already identified and ready to execute the playbook.

Future Verticals

Marketing is just the beginning. There are many service-heavy industries with large target markets, fragmented market share, value-based offerings that are ripe for a transformation like this. As AI matures even more industries will become eligible.

Our goal is to open roughly one new vertical per year, following the frontier of AI as it matures and enables more industries to transform. Each led by an experienced operator with deep knowledge of industry paired with our playbooks and unique technology platform.

Over the next decade we believe we can not only create billions of wealth with this strategy but also demonstrate to the world what good looks like in terms of how people and AI will thrive together.

We chose the Tinman from the Wizard of Oz because he was a machine with a heart. Like the Tinman, the best companies that generate best returns will likewise combine people and agents in a way that amplifies their unique strengths.

A Manifesto for Builders

Think the only interesting problems live inside VC-funded software companies? Look beyond Sand Hill Road's limited imagination.

Consider Nivoda. As one of their first investors, I worked closely with this jewelry marketplace as they transformed the sleepy B2B gemstone trade. Though they were a new venture, not an acquisition, their approach exemplifies our thesis perfectly: they brought modern marketplace dynamics to an industry operating like it was 1975. Within three years, they connected diamond suppliers in Antwerp and Tel Aviv with retailers in 100+ countries. We've incorporated elements of their playbook in our own portfolio. Proving digital transformation doesn't need a Series A. Today the company is venture funded and growing fast, but off the back of a proven business model.

Or look at Cava. Beginning as a Mediterranean restaurant chain, they evolved into a tech-optimized operational juggernaut. Their data-driven approach to menu design, store location, and throughput optimization turned what could have been a commoditized food service model into a company that exited at a $4.7B IPO valuation. Not a tech company by Silicon Valley's narrow definition, but their tech-forward operations created more shareholder value than 99% of software startups founded the same year.

These aren't isolated examples. Across America, hundreds of HVAC contractors, logistics providers, legal firms, and specialty-care clinics quietly compound at 25%+ annual growth with zero venture backing. They apply technological leverage to boring, essential services that no one wants to Instagram but everyone needs to consume.

The frontier has migrated. It's no longer about code alone, but code tailored to and multiplied by operations. That's where the next generation of fortunes will be minted. While others chase the latest AI hallucination feature, we're applying these tools to businesses that move physical things and serve real people.

The upside isn't theoretical. It shows up in your P&L.

A Call to Three Audiences

This model needs three distinct groups to succeed. Consider this your personal invitation from David and me.

OPERATORS. Can you lead a P&L, inspire teams, and bring a tech friendly mindset with deep knowledge of your industry, and now looking for the next opportunity that will help you ride the wave of AI with enough leverage on your skills to be worth your time? We offer genuine autonomy and resources to match.

We'll hand you the keys to a vertical with the mandate to transform it. Your sandbox will be a profitable company on Day 1, augmented by our transformation toolkit and cross-portfolio expertise. The upside isn't a lottery ticket. You receive a direct share of the value you create through phantom equity that pays real cash when hitting predefined milestones.

BUSINESS OWNERS. Considering an exit? We represent a distinct alternative to both strategic acquirers and financial sponsors.

We buy for the long-term and maintain founding legacies. Your company becomes the foundation of a tech-enabled dynasty, not a line item in a roll-up spreadsheet destined for "cost synergies." Many of our sellers stay involved as advisors, board members, or collaborators. They take chips off the table while watching their life's work evolve into something even more valuable.

INVESTORS. Are you hunting asymmetric returns grounded in real cash flow? Tinman Group represents a new category. Somewhere between private equity and venture capital—deliberately designed to capture the best elements of both.

We're opening our first external capital tranche this year after proving the model with proprietary capital. You won't wait seven years to see if a theoretical exit materializes. Distributions from profitable operations will offer liquidity within the first 24 months.

Email us at MachinesWithHeart@tinman.group or DM on X (@tinmangroup). Let's compound real value—out where the servers meet the street.