Why SaaS Companies Require a Bigger Mission, Not a Smaller Headcount with AI
Jack Dorsey’s bold move at Block tells us a lot about SaaS, and highlights what too many leaders are still missing.
By now you’ve heard the news. Block, the parent company of Square and Cash App, announced it is cutting nearly 4,000 jobs (almost half its workforce) even as it reported Q4 gross profits of $2.9 billion, up 24% year over year. CEO Jack Dorsey has been unambiguous about the rationale: AI-driven productivity gains mean smaller teams can do more. His CFO, Amrita Ahuja, put a number on it. Gross profit per employee was roughly $500,000 in 2019. It hit $750,000 in 2024, $1 million in 2025, and if Block hits its 2026 targets, it will reach $2 million, almost double last year’s level.
That’s not a rounding error. That’s a genuine transformation story, and Dorsey deserves credit for executing it with unusual clarity and conviction.
But here’s my question leaders like Dorsey:
Are you building a more efficient company, or you building a more significant one, that will last the test of time?
Because those are not the same thing. And the difference between them may be the most important leadership question of our time.
Let’s Talk About Railroads
In 1960, Harvard Business School professor Theodore Levitt published what would become one of the most cited business essays in history: Marketing Myopia. His central argument was deceptively simple. The American railroad industry didn’t decline because people stopped needing transportation. It declined because railroad companies believed they were in the railroad business — not the transportation business.
They optimized their trains. They cut operational waste. They were ruthlessly efficient at doing exactly what they had always done. Meanwhile, automobiles and airlines ate their market, their customers, and eventually their future. The railroads were so busy getting better at their existing mission that they never stopped to ask whether that mission was still the right one.
Here’s the parallel: Block is, at its core, a financial software company. Square processes payments. Cash App moves money. The AI agent “goose” drafts emails and automates workflows. These are genuinely impressive capabilities. But if Block’s highest ambition is to run those capabilities with fewer people and higher margins, then it is, in a very real sense, in the railroad business.
The efficiency is real. The myopia might be too.
What Business Is Block Actually In?
Let’s think about who Block actually serves. Square is the financial infrastructure for millions of small and micro-businesses — the food truck owner, the independent nail salon, the pop-up boutique at a weekend market. Cash App has become a primary banking interface for millions of Americans who are underserved or entirely excluded by traditional financial institutions.
That’s not a payments company. That’s potentially a financial inclusion platform at civilizational scale.
The question Dorsey and his team have answered is: How do we do more with less? That’s a fine question. But the question they haven’t answered loudly enough is: What does ‘more’ actually mean for the people who depend on us?
Imagine a Block that used its AI capabilities not just to improve margins, but to actively reshape the financial lives of underserved communities. Imagine AI-powered financial coaching embedded in Cash App that helps a first-generation college graduate understand their credit score. Imagine Square evolving from a payment terminal into an AI-powered small business operating system that does for the owner of a neighborhood restaurant what a $200,000 CFO does for a Fortune 500 company.
That’s a much bigger company. A harder company to build. And one that would create a far more durable moat than operational efficiency ever could.
The IBM Lesson Nobody Remembers Correctly
People love to tell the IBM turnaround story as a technology story. It’s actually a mission story.
When Lou Gerstner arrived at IBM in 1993, the company was collapsing under the weight of its own hardware legacy. Conventional wisdom at the time was to break IBM apart and sell its divisions. IBM had been an extraordinarily efficient manufacturer of mainframes for decades.
Gerstner refused. Not because he was sentimentally attached to the hardware. But because he looked at IBM’s customer relationships, including deep, decades-long partnerships with the world’s largest enterprises, and asked a different question than his predecessors had been asking. He didn’t ask, How do we sell more hardware? He asked, What do our customers actually need?
The answer was integration. They needed someone to help them make sense of their increasingly complex technology environments. IBM’s mission shifted from making computers to solving the world’s hardest technology problems. IBM Global Services was born. And a company that was weeks from irrelevance became, again, one of the most important enterprises on the planet.
The lesson isn’t that you pivot away from your core. It’s that your core capability, including your people, your relationships, your institutional knowledge, is the raw material for a much grander mission, if you’re willing to ask the bigger question.
Ping An Didn’t Just Get More Efficient. It Got More Ambitious.
The most instructive counterexample to Block’s current trajectory isn’t a tech company at all. It’s Ping An Insurance, profiled in my upcoming book, Hyperadaptive: Rewiring the Enterprise to Become AI-Native. Headquartered in Shenzhen, China, it has used AI to expand its definition of what it is.
Ping An started with a clear and humble goal: better insurance underwriting. That’s about as unglamorous as it gets. But instead of using AI to simply process policies faster and cut underwriters, the company asked a different question: What could we become if AI let us understand our customers more completely?
The answer transformed the entire organization. Data from connected vehicles informed insurance pricing. Medical diagnostics enhanced risk assessment. Customer conversations with AI-powered voice robots continuously refined service delivery across all channels. A person checking their health metrics through Ping An’s app might now receive tailored financial planning suggestions. Scenario-based marketing reaches customers at precisely the moment a product would be most relevant.
What began as an insurance company became, in the words of Hyperadaptive, an orchestrator of customer life journeys.
The business results reflect the ambition. Customers using Ping An’s digital ecosystems hold an average of 2.9 contracts, compared to 1.2 for non-ecosystem users. Average revenue per ecosystem user is $5,288 — nearly four times the $1,399 generated by non-users. When your AI systems work together this seamlessly, customers don’t just buy products. They buy into an entire ecosystem that genuinely improves their lives.
Ping An didn’t shrink its way to significance. It used AI as a launchpad to become something no one had imagined an insurance company could be.
The Danger Hiding in the Headcount Numbers
There is a quieter risk in what Block is doing, and it deserves naming directly.
In Hyperadaptive, I describe what happens when organizations over-rotate on head-cutting: “Institutional knowledge walks out the door. You won the technical battle, but lost the cultural war.”
Block is cutting nearly half of its workforce. That is not a reorganization. That is a wholesale renegotiation of the organization’s relationship with its own accumulated knowledge. The people leaving don’t just hold job descriptions. They hold customer relationships, institutional memory, hard-won context about what went wrong in 2021 and why, and the quiet, undocumented understanding of how things actually get done.
AI tools are remarkable. But they learn from people. The people who lived the past, who understand the nuances of why Block’s lending product works in one market and struggles in another, who know which enterprise client needs a phone call and which one wants to be left alone. Those people are not replaceable by a language model, at least not yet, and not without significant loss.
The CFO’s productivity data is compelling. But productivity metrics measure what we can count. Institutional knowledge is precisely the thing that resists counting until the day it’s gone.
The Leadership Mindset We Need Today
What separates the railroad operators from the Ping Ans of the world isn’t intelligence or resources. It’s the question they start with.
Efficiency-oriented leaders ask: How do we do what we do with less?
Mission-oriented leaders ask: Given everything AI now makes possible, what should we become?
In Hyperadaptive, I describe this as the difference between purpose and process.
You can revolutionize every process in your organization and still watch momentum stall, because without a clear and elevating purpose, you’ll just end up doing the wrong things more efficiently.
The leaders in your organizations who are currently framing their AI transformation as a headcount exercise are, in the most generous interpretation, solving for the wrong variable. Efficiency is a byproduct of great strategy, not a substitute for it. Margin expansion is a result of serving customers better, not of serving fewer of them.
The question for every AI transformation leader reading this is whether you are using AI to become a smaller, cheaper version of what you already are, or to become something genuinely more important to the people and communities you serve.
I sincerely hope it is the latter.
Before You Open That Org Chart
If you are an leader in charge of AI (and if you’re reading this, you likely are) here is the exercise I’d ask you to do before your next workforce planning conversation:
Write down what your best customers would genuinely grieve if your company disappeared tomorrow. Not the features. Not the integrations. The actual gap in their lives. The thing they’d have to figure out on their own.
Now ask: Is your AI strategy making that thing better, bigger, and more central to everything you do? Or is it making your cost structure leaner while that core value quietly erodes?
Dorsey isn’t wrong that AI changes the economics of running a company. He’s making a rational, data-driven decision by the numbers he has in front of him. But the most transformative companies of the next decade won’t be the ones that used AI to cut the most efficiently. They’ll be the ones that used it to see further, serve more deeply, and become indispensable in ways their competitors can’t easily replicate.
Ping An didn’t start with a plan to orchestrate customer life journeys. It started with a commitment to understanding its customers better than anyone else, and it let that commitment compound.
That’s the leadership mindset that changes everything. The efficiency will follow. It always does.
— Melissa
P.S. — Something new is happening on the paid side of this Substack. I’m launching monthly peer roundtables. These small, candid, off-the-record conversations with enterprise leaders who are deep in the AI transformation weeds are where we can all learn together. Plus short ‘conversation roundup’ videos where I share what I’m actually hearing from the front lines. The first roundtable is April 16th on leadershp alignment. If you’ve been on the fence, things are about to get a lot more interesting over here.
Melissa M. Reeve is the author of Hyperadaptive, a research-backed guide for leaders building AI-native organizations. The Hyperadaptive Model provides a five-stage pathway from AI pilots to full organizational transformation. Learn more and take the free Hyperadaptive Discovery Survey at hyperadaptive.solutions/discover.