Why the World's Most Successful AI Companies Are Doing Less - And Getting More

‍

In 1906, Italian economist Vilfredo Pareto noticed something peculiar in his garden: 20% of his pea pods produced 80% of his peas.

This observation would eventually revolutionize business thinking worldwide. In 2025, in the age of artificial intelligence, we're witnessing something even more extreme:

Leading companies focus on an average of 3.5 AI use cases, compared with 6.1 for other companies. The leaders anticipate generating 2.1 times greater ROI on their AI initiatives than their peers.

They're doing 43% less. They're achieving 110% more.

This isn't a paradox. It's a pattern and it's hiding in plain sight in every piece of AI success data we have.

The Proliferation Trap

Walk into any enterprise today and you'll find AI initiatives everywhere. Chatbots in customer service. Machine learning in supply chain. Predictive analytics in sales. Natural language processing in HR. Computer vision in quality control.

It feels like progress. It looks like innovation. It's actually a form of organizational suicide.

About 95% of AI pilot programs fail to achieve rapid revenue acceleration, while only 5% succeed in driving tangible results.

Why such catastrophic failure rates? The answer isn't in what these companies are doing. It's in how much they're trying to do at once.

The Mathematics of Dilution

Let's do the brutal math that most organizations refuse to calculate.

Imagine you have 100 units of organizational capacity—time, attention, resources, change management bandwidth. This is a finite resource, despite what your ambitions tell you.

Scenario A: The Unfocused Organization (6.1 initiatives)

  • 100 units ÷ 6.1 initiatives = 16.4 units per initiative
  • Result: Each initiative is under-resourced, under-managed, and under-delivered
  • Outcome: 74% of companies have yet to show tangible value from their use of AI

Scenario B: The Focused Organization (3.5 initiatives)

  • 100 units ÷ 3.5 initiatives = 28.6 units per initiative
  • Result: Each initiative has 74% more resources than in Scenario A
  • Outcome: 2.1x greater ROI

This isn't theoretical. This is what's actually happening across thousands of organizations right now.

The Hidden Cost of Context Switching

But the real math is even worse than simple division suggests.

AI leaders successfully scale more than twice as many AI products and services across their organizations.

Wait—if they're pursuing fewer initiatives, how are they scaling more?

Because scaling isn't about starting. It's about finishing. And finishing requires something most organizations have forgotten how to do: saying no.

Every additional AI initiative doesn't just divide your resources. It multiplies your complexity:

  • Integration points grow exponentially
  • Governance overhead compounds
  • Change management becomes chaos management
  • Technical debt accumulates faster than value

The survey reveals that companies face numerous challenges when implementing AI initiatives, with around 70% stemming from people and process-related issues.

Every new initiative amplifies these people and process challenges. It's not addition. It's multiplication.

The Venture Capital Principle

The world's best venture capitalists know something most corporate executives don't: Power laws rule everything around us.

In a typical VC portfolio:

  • 1 investment returns 100x
  • 2 investments return 10x
  • 4 investments return 2x
  • The rest return nothing

The entire portfolio's success depends on identifying and nurturing that one massive winner.

AI initiatives follow the same power law, as proven by the data:

62% of AI's value lies in core business functions such as operations (23%), sales and marketing (20%), and R&D (13%).

Just three areas account for nearly two-thirds of all value. Yet most companies spread their efforts across dozens of use cases, treating each as equally important.

The ROI Revelation

Let's dissect what "2.1 times greater ROI" actually means in practice.

Future-built companies achieve 1.7x revenue growth, 3.6x three-year total shareholder return, and 1.6x EBIT margin compared to laggards.

These aren't incremental improvements. These are category-defining differences. The kind that determine whether you're acquiring competitors or being acquired.

How do they achieve this?

Almost 45% of leaders integrate AI in their cost transformation efforts across functions (compared with only 10% of others). More than a third of leaders focus on revenue-generation from AI, compared with only a quarter of other companies.

They don't just focus on fewer initiatives. They focus on the right initiatives—the ones that transform core operations rather than optimize periphery.

The Scaling Secret

Here's what nobody tells you about scaling AI: It's not about technology. It's about organizational physics.

Leaders successfully scale more than twice as many AI products and services across their organizations.

How can doing less lead to scaling more? Because scaling requires three things:

1. Deep IntegrationWhen you're juggling 6+ initiatives, integration is shallow by necessity. When you're focused on 3.5, you can weave them into the fabric of operations.

2. Change AbsorptionGenAI impacts the majority of an enterprise's workforce, both in terms of potential productivity improvements and the extensive training necessary to achieve them. Organizations can only absorb so much change at once. Fewer initiatives means deeper adoption.

3. Learning LoopsOnly 10% of global companies have brought one or more GenAI applications to scale. Why? Because scaling requires learning, and learning requires focus. You can't learn from six experiments running simultaneously—you can only react to them.

The Investment Paradox

You might think companies pursuing fewer initiatives are investing less. The opposite is true.

Future-built companies plan to spend 26% more on IT (representing almost a full percentage point of revenue) and dedicate up to 64% more of their IT budget to AI in 2025.

They're investing more money in fewer things. That's not timidity. That's conviction.

Leading companies allocate more than 80% of their AI investments to reshaping key functions and inventing new offerings rather than smaller-scale, productivity-focused initiatives.

The False Promise of Diversification

"But shouldn't we diversify our AI portfolio to reduce risk?"

This is corporate finance thinking applied to the wrong domain. Diversification works when returns are uncorrelated. But AI initiatives aren't stocks. They're organizational transformations. And transformations are deeply correlated through shared resources:

  • The same team implements them
  • The same infrastructure supports them
  • The same workforce must adapt to them
  • The same leadership attention guides them

When you spread these resources thin, you don't reduce risk. You guarantee failure across the board.

Companies that succeed purchase AI tools from specialized vendors and build partnerships 67% of the time, while internal builds succeed only one-third as often.

Even successful companies aren't trying to do everything themselves. They're focusing their internal efforts where they can create unique advantage and partnering for everything else.

The Competitive Compound Effect

The focus gap creates a compound effect that becomes insurmountable over time.

Year 1: Focused companies achieve 2.1x better ROI, reinvest gainsYear 2: Gap widens as reinvestment accelerates focused initiativesYear 3: The spread in digital and AI maturity between leaders and laggards has increased by 60% over the past three years

This isn't a linear race where laggards can catch up by running faster. It's an exponential game where the leaders are accelerating away.

Leading companies that moved early enjoy outsized benefits across financial and operational fronts, and this performance gap is widening.

The Attention Architecture

Every organization has what I call an "attention architecture"—the invisible structure that determines what gets noticed, nurtured, and scaled.

Leaders have six differentiating characteristics: They focus on the core business processes as well as support functions.

But here's the key word: focus. Not spread. Not diversify. Focus.

When you have 6+ AI initiatives:

  • No single initiative gets executive attention
  • Middle management is overwhelmed
  • Frontline workers are confused
  • Success metrics become meaningless (which of the 6 initiatives drove that improvement?)
  • Failure becomes impossible to diagnose

When you have 3.5 initiatives:

  • Each has a senior champion
  • Progress is visible and measurable
  • Problems get solved, not documented
  • Success patterns become clear and replicable

The Strategic Subtraction Method

So how do you move from 6.1 initiatives to 3.5? You don't add focus. You subtract distraction.

AI's greatest value lies in core business processes where leaders are generating 62% of the value.

Start with a simple question: Which initiatives directly transform core business processes?

Everything else is a distraction, regardless of how innovative it sounds.

Here's the framework the winners use:

The Three-Filter System:

Filter 1: Core Process ImpactDoes this fundamentally change how we create or deliver value?

  • If no → eliminate
  • If yes → proceed to Filter 2

Filter 2: Compound PotentialWill success here make other initiatives more valuable?

  • If no → defer
  • If yes → proceed to Filter 3

Filter 3: Organizational ReadinessCan we realistically execute this given our current capabilities?

  • If no → partner or postpone
  • If yes → commit fully

Most initiatives fail the first filter. That's the point.

The Courage Gap

Here's what really separates the 3.5-initiative companies from the 6.1-initiative companies: courage.

It takes no courage to say yes to every AI opportunity. It takes tremendous courage to say no to good ideas in service of great ones.

Leaders look beyond pure productivity plays and back their ambitions with investment in AI and workforce enablement, making twice the investment in digital, double the people allocation, and twice as many AI solutions scaled.

They're not doing less because they're less ambitious. They're doing less because they're more ambitious. They want transformation, not iteration.

The Customer Value Test

Almost 7 out of 10 employees believe generative AI will assist them to better serve their customers.

But here's the question: How many of your 6+ AI initiatives directly improve customer experience?

Leaders apply a simple test: If the customer can't feel it, it's not strategic.

This eliminates:

  • Internal reporting dashboards
  • Theoretical predictive models
  • "Innovative" pilots with no path to production
  • Technology demonstrations masquerading as business initiatives

What remains? The 3.5 initiatives that actually matter.

The Paradox of Choice

In 2000, psychologists Sheena Iyengar and Mark Lepper ran an experiment. They set up a jam display at a grocery store. Sometimes they displayed 24 varieties. Sometimes just 6.

The 24-variety display attracted more browsers (60% vs 40%).But the 6-variety display generated 10 times more purchases (30% vs 3%).

Your organization faces the same paradox with AI initiatives. More options feel like progress. But they paralyze decision-making and dilute execution.

Few organizations are experiencing meaningful bottom-line impacts from gen AI.

Not because the technology doesn't work. Because they're trying to make it work everywhere at once.

The Network Effects of Focus

When you focus on 3.5 initiatives instead of 6.1, something remarkable happens: network effects emerge.

A third of future-built companies use AI agents, compared with 12% of scalers and almost none of the laggards.

How did they get to AI agents while others are still struggling with basic automation? Because each successful focused initiative creates capabilities that accelerate the next:

  • Initiative 1 builds data infrastructure → Initiative 2 leverages it
  • Initiative 2 develops AI talent → Initiative 3 benefits from their expertise
  • Initiative 3 creates change management patterns → Future initiatives scale faster

This is how doing less becomes more. Not through efficiency, but through compounding capabilities.

The Executive Dilemma

You're facing pressure from:

  • The board wanting to see "AI innovation"
  • Vendors pitching their solutions
  • Competitors announcing new AI initiatives
  • Internal teams proposing AI projects

The temptation is to say yes to everything. To hedge your bets. To avoid looking like you're falling behind.

But the data is unequivocal: Companies that focus on depth over breadth, prioritizing an average of 3.5 use cases, anticipate 2.1 times greater ROI.

Your choice isn't between action and inaction. It's between focused transformation and diffused failure.

The Implementation Blueprint

Moving from 6.1 to 3.5 initiatives isn't about arbitrary reduction. It's about strategic concentration.

Step 1: The Portfolio AuditList every AI initiative. Be honest. That "exploration" counts. That "proof of concept" counts. That vendor trial counts.

Step 2: The Value MappingFor each initiative, answer: What specific business metric will this move? By how much? By when? If you can't answer with precision, it's not an initiative—it's an experiment.

Step 3: The Dependency AnalysisWhich initiatives are prerequisites for others? Which create capabilities others need? This reveals your critical path.

Step 4: The Courage CutEliminate everything that isn't on the critical path. Yes, even the CEO's pet project. Yes, even the initiative that's "90% complete" (it never is).

Step 5: The Full CommitmentTake every resource from the eliminated initiatives and concentrate them on the remaining 3.5. Double down on what matters.

The Uncomfortable Questions

Before you dismiss this as too simplistic, answer these questions:

  1. Can you name the specific ROI of each of your current AI initiatives?
  2. How many of your AI pilots from 18 months ago are now in production?
  3. What percentage of your workforce is actually using your AI tools daily?
  4. If you had to eliminate half your AI initiatives tomorrow, which would you keep?

If these questions make you uncomfortable, you already know the answer: You're doing too much.

The Historical Parallel

In the 1990s, most companies pursued dozens of internet initiatives. Websites for every department. E-commerce experiments everywhere. Digital everything.

The winners? Amazon focused on online retail. Google focused on search. Facebook focused on social connections.

The losers? They did everything and dominated nothing.

We're at the same inflection point with AI. The companies trying to do everything with AI will achieve nothing with it. The companies that choose their battles will win the war.

The False Choice

"But we need quick wins AND transformation."

This is the false choice that leads to 6.1 initiatives. You don't need quick wins and transformation. You need quick wins THROUGH transformation.

Deploy GenAI tools to quickly deliver broad, diffuse productivity gains of 10% to 20% across the enterprise.

Notice: broad, diffuse gains from deployment, not from dozens of separate initiatives. One well-executed initiative can deliver more quick wins than six half-executed ones.

The Competitive Reality

While you're debating which of your six initiatives to prioritize this quarter, your focused competitors are on implementation phase three of their first initiative.

Future-built firms plan to spend more than twice as much on AI compared to laggards in 2025.

They're not spreading that investment across more initiatives. They're concentrating it on fewer. That's not just a different strategy. It's a different philosophy.

The Decision

The math is clear. The evidence is overwhelming. The path is obvious.

Yet most organizations will continue pursuing 6+ initiatives, believing that more activity equals more progress, that more bets equals less risk, that more innovation equals more value.

They're wrong on all counts.

Only 26% of companies have developed the necessary set of capabilities to move beyond proofs of concept and generate tangible value.

These aren't the companies doing the most. They're the companies doing the least, best.

The Bottom Line

In the age of AI, competitive advantage doesn't come from doing everything. It comes from doing the right things completely.

3.5 initiatives. 2.1x returns.

Half as many. Twice the return.

The math isn't complex. The execution isn't easy. But the choice is simple:

You can join the 74% failing to capture value from their scattered AI efforts.

Or you can join the 26% transforming their businesses through focused execution.

The difference isn't in your technology. It isn't in your budget. It isn't even in your talent.

It's in your willingness to do less, better.

Every AI initiative you don't pursue is an investment in the ones you do.

Every resource you don't scatter is power you can concentrate.

Every "no" you say is a multiplier on your "yes."

The organizations winning with AI aren't the ones doing everything.

They're the ones who understood, early and deeply, that in the age of infinite possibilities, the only strategy that matters is choosing what not to do.

Peter Cobabe
Peter Cobabe
03 Nov 2025

Sources

  1. Boston Consulting Group (BCG). (April 15, 2025). "From Potential to Profit: Closing the AI Impact Gap." Retrieved from: https://www.bcg.com/publications/2025/closing-the-ai-impact-gap
  2. MIT NANDA. (August 2025). "The GenAI Divide: State of AI in Business 2025." MIT Research Report.
  3. Boston Consulting Group (BCG). (October 24, 2024). "AI Adoption in 2024: 74% of Companies Struggle to Achieve and Scale Value." Retrieved from: https://www.bcg.com/press/24october2024-ai-adoption-in-2024-74-of-companies-struggle-to-achieve-and-scale-value
  4. Boston Consulting Group (BCG). (September 30, 2025). "AI Leaders Outpace Laggards with Double the Revenue Growth and 40% More Cost Savings."
  5. Fortune. (August 18, 2025). "MIT Report: 95% of Generative AI Pilots at Companies Are Failing."
  6. McKinsey & Company. (March 12, 2025). "The State of AI: How Organizations Are Rewiring to Capture Value."
  7. McKinsey & Company. (January 12, 2024). "Rewired and Running Ahead: Digital and AI Leaders Are Leaving the Rest Behind."
  8. Boston Consulting Group (BCG). (February 4, 2025). "Turning GenAI Magic into Business Impact."
  9. Boston Consulting Group (BCG). (February 5, 2025). "Laying the Foundation for Gen AI Success."
  10. Vena Solutions. (August 5, 2025). "100+ AI Statistics Shaping Business in 2025."