When AI threatens to eat your company, only the founder can rewrite the menu.
There’s a moment in the life of every successful B2B company where the founder steps back and says: “I built this thing. It works. Let the professional operators take it from here.” We all think about it. Or almost all of us.
It’s the classic playbook. You take it from $0 to $1B ARR, you IPO, you hire a “been there, done that” operator CEO from Google or SAP or Microsoft, and you slide gracefully into an Executive Chairman role. You focus on product. On vision. Maybe you start a venture fund. You’ve earned it.
And then AI shows up. Not incrementally. Not “we added a chatbot to the help page.” But existentially. The kind of disruption that makes your entire category — your entire business model — suddenly look like it might have an expiration date.
And suddenly, the professional operator isn’t enough. The board isn’t enough. The strategy consultants definitely aren’t enough.
They need the founder back.
We’re watching this play out in real-time at two of the most important enterprise software companies in the world — UiPath and Workday — and the pattern is too striking to ignore.
Daniel Dines, UiPath: The 4-Month Retirement
Here’s a story that would be unbelievable if it weren’t true.
Daniel Dines co-founded UiPath in a Bucharest apartment in 2005. He spent ten years grinding before the company hit its first $1M in ARR. Ten years. Most founders would have quit five times over. He didn’t. He built UiPath into the fastest-growing SaaS company of its era, took it public, and grew it past $1.3 billion in revenue.
In July 2023, Dines announced he’d step down as co-CEO, ceding the role to Rob Enslin — a longtime SAP executive who’d come over from Google Cloud. Dines would become Chief Innovation Officer. Focus on product. The classic transition.
On January 31, 2024, Enslin officially became sole CEO.
By June 1, 2024 — four months later — Dines was back as CEO. Enslin was out.
Four months. That might be the fastest boomerang in SaaS history.
What happened? Generative AI happened. The explosion of LLMs and agentic capabilities didn’t just represent a new feature for UiPath to bolt on — it threatened to make the company’s entire core product, robotic process automation, look like yesterday’s technology. If AI agents could reason, adapt, and execute workflows autonomously, why would you need rigid software bots that follow pre-scripted rules?
This wasn’t a question a professional operator could answer. This was a question only the person who invented the category could answer.
Dines was characteristically candid when he returned. He acknowledged the company had become siloed, too bureaucratic, too focused on acting like a big company instead of the scrappy innovator that built the business. He talked about bringing back “the joy of working together.” He told analysts the company needed a “re-invigorated customer-centric approach.”
Translation: we got comfortable, and the world just changed underneath us.
And here’s the thing — it’s working. Under Dines’ renewed leadership, UiPath has pivoted hard into what it calls “agentic automation.” The company launched Agent Builder, an orchestration layer called Maestro, and a technology called ScreenPlay that lets AI agents interact with any application interface visually — the way a human would. Revenue growth re-accelerated to nearly 16% in Q3 of fiscal 2026. ARR hit $1.78 billion. The stock rose 29% in 2025 — and more than 50% off its lows. Over 950 companies are now building AI agents on UiPath’s platform.
Dines isn’t just steering the ship through rough waters. He’s rebuilding the ship while it’s sailing.
Aneel Bhusri, Workday: The $1.3 Billion Reason to Come Back
If Dines’ story is about a founder who couldn’t stay away, Bhusri’s is about a founder who really didn’t want to come back — but the math made it impossible to say no.
Aneel Bhusri co-founded Workday with his mentor Dave Duffield in 2005. He served as CEO (or co-CEO) for fifteen years, building it into one of the most important enterprise software companies in the world — $8.4 billion in revenue, the dominant player in cloud HR and finance.
Bhusri was never the prototypical hard-charging tech CEO. He once said publicly that he didn’t have training for running a multibillion-dollar corporation. At a 2024 keynote, he paused mid-speech to say hello to his mother in the audience, celebrating what was supposed to be his final keynote as CEO. He wanted to get back to building products and setting strategy.
He’d hired Carl Eschenbach, a former Sequoia Capital partner and VMware veteran, as co-CEO in late 2022. Bhusri stepped into the Executive Chairman role in February 2024. The succession was complete. The plan was working.
Then came the SaaSpocalypse.
In late January and early February 2026, roughly $285 billion in market value was wiped from software stocks in a matter of days. The catalyst: the growing realization that agentic AI platforms from companies like Anthropic and OpenAI weren’t just interesting demos — they were beginning to threaten the per-seat licensing model that has powered SaaS economics for two decades. Workday’s stock had already been sliding. From its all-time high market cap near $80 billion, the company had shed roughly $40 billion in value. Bhusri personally watched his 8-million-share stake lose approximately $1.3 billion in paper value in less than two years.
On February 9, 2026, Workday announced Eschenbach was stepping down. Bhusri was returning as CEO, effective immediately.
This wasn’t a placeholder appointment while the board searched for a replacement. Workday confirmed to TechCrunch that Bhusri was coming back permanently. And they paid accordingly: a $138.8 million compensation package, with $75 million tied to undisclosed stock price targets over five years, and another $60 million in restricted stock that vests simply by Bhusri staying at the company for four years — no performance targets required.
That structure tells you everything about how the board sees the situation. Half the package is a bet that Bhusri can actually turn the stock around. The other half is essentially a retention payment — an acknowledgment that even the founder might not be able to fully overcome the headwinds bearing down on SaaS, but that he’s still the best option available.
Bhusri’s own framing was telling: “AI is a bigger transformation than SaaS — and it will define the next generation of market leaders.”
He’s not wrong. And he’s the person with both the vision and the voting control to make hard choices. With majority voting power and operational authority as CEO, Bhusri has more power to reshape Workday than any outside hire ever could.
The Pattern: Why Founders Come Back
Step back and look at the broader trend, and it’s unmistakable.
Steve Jobs returned to Apple in 1997 when the company was 90 days from bankruptcy. Michael Dell came back to Dell in 2007 when the PC business was cratering. Jack Dorsey returned to Twitter in 2015 when the platform was stagnating. Howard Schultz came back to Starbucks — twice. Sam Altman was reinstated at OpenAI within days of his ouster in November 2023.
Academic research from the University of North Carolina actually suggests boomerang CEOs tend to underperform first-timers on average. But that statistic misses the point. Boards don’t bring founders back because they expect average results. They bring founders back because the company faces an existential threat that requires someone who can make decisions with the authority, speed, and conviction that only a founder possesses.
Professional operator CEOs are optimizers. They’re brilliant at scaling what works, improving margins, running efficient sales organizations. But when the ground shifts beneath you — when the entire category you compete in is being questioned — you don’t need optimization. You need reinvention. And reinvention requires someone who was there at the beginning, who understands the company’s DNA at a cellular level, and who has the credibility (and the equity stake) to make bold, potentially unpopular bets.
Dines could walk into UiPath and say “we’ve become too siloed, too bureaucratic, we’ve lost our edge” and people listen — because he built the company from zero. Bhusri can reshape Workday’s entire go-to-market and product strategy because he has both the institutional memory and the voting control to do it.
An outside CEO saying those things gets a polite nod and a 6-month “transition period.”
The Other Boomerangs We Shouldn’t Forget
Dines and Bhusri are the most dramatic recent examples in enterprise SaaS, but they’re part of a broader pattern of founders being pulled back into the arena by AI:
Michael Dell, Dell Technologies — The OG Boomerang (2007), Now Thriving. Dell stepped down as CEO in 2004 and came back in 2007 when the PC business was cratering. The stock fell another 40% after his return, and he ultimately took the company private in 2013. But that privatization move is exactly what positioned Dell to pivot quietly into enterprise infrastructure — and now AI servers. AI server orders hit $12.3 billion in a single quarter in late 2025. Dell’s stock is up roughly 35% year-to-date. The man who came back to save a dying PC company is now running one of the most important AI infrastructure businesses on the planet. That’s the boomerang playbook at its absolute best — but it took him 18 years from his return to get here. Not every founder has that patience or runway.
Sam Altman, OpenAI — The 5-Day Boomerang (2023). Technically not a public B2B company yet, but too important to ignore. Altman was fired by OpenAI’s board on a Friday in November 2023 and was reinstated as CEO by the following Wednesday. Hundreds of employees threatened to quit. Investors revolted. The board capitulated. It was less a “boomerang” and more of a “rubber band snap.” But it reinforced a critical lesson: when a founder is the product vision, removing them creates an existential vacuum that no board process can fill.
Sergey Brin, Google/Alphabet — The Quiet Re-Engagement (2023). Brin isn’t technically a boomerang CEO — Sundar Pichai still runs the show — but he effectively came out of semi-retirement in 2023 after ChatGPT forced Google into “Code Red” mode. He’s been personally recruiting AI researchers and re-engaging deeply with the technical side of the business. It worked: 20% of Google’s AI software engineer hires in 2025 were returning former employees, and Gemini 3 has made Google competitive again. Sometimes the founder doesn’t need the title — they just need to show up.
The Ones Who Never Left — And Why That Matters. It’s worth noting the B2B founders who didn’t step back and are now navigating AI from the CEO seat directly: Marc Benioff at Salesforce is doing a boomerang reinvention without the boomerang — overhauling leadership, cutting thousands of roles, and betting everything on Agentforce. Jensen Huang never left Nvidia. Satya Nadella isn’t a founder but has functioned like one since transforming Microsoft. These “continuous founders” avoided the succession problem entirely, but they’re facing the same existential question: can they reinvent their companies fast enough?
Who’s Next?
This is the question every B2B board should be asking right now. If the SaaSpocalypse deepens — if AI agents genuinely start cannibalizing seat-based revenue at scale — which other companies will need to pull the founder back in?
Here are the dynamics to watch:
- The AI Vulnerability Test: The more a company’s revenue depends on per-seat licensing of workflows that AI could automate, the more vulnerable it is. Think about it — if 10 AI agents can do the work of 100 humans, you don’t need 100 seats anymore. You need 10. That’s a 90% revenue hit for the same work output.
- The Founder Availability Test: Is the founder still alive, engaged, and willing? Not every company has this option. Some founders have moved on entirely. Others are running new companies. The lucky ones have founders who stayed close — as Executive Chairs, board members, or “Chief Innovation Officers” — close enough to come back quickly.
- The Urgency Test: Subscription revenue growth is slowing across the board in enterprise SaaS. Companies that were growing 20%+ are suddenly posting mid-teens or single digits. The pressure on boards is mounting. Activist investors are circling. The longer a company waits, the harder the turnaround becomes.
Some names worth watching:
Salesforce still has Marc Benioff at the helm, so the founder-return question doesn’t apply — but the AI pressure is intense. The company just reshuffled its entire leadership team, laid off thousands, and is grappling with how Agentforce fits into a world where customers want fewer seats, not more. Benioff is effectively doing a “boomerang reinvention” without the boomerang — pivoting the company he founded in real-time.
Atlassian, which has shed enormous value in the SaaSpocalypse, still has co-founders Scott Farquhar and Mike Cannon-Brookes deeply involved. If things get worse, expect them to tighten their grip.
Intuit, where founder Scott Cook remains on the board, has seen its stock punished by fears that AI agents could automate the tax preparation and small business accounting workflows that drive its business.
And then there are the dozens of mid-cap and growth-stage SaaS companies — the $500M to $5B revenue range — where founders stepped back in the last few years and handed the reins to professional operators who are now staring down an AI disruption they didn’t sign up for.
The Cautionary Tale: Asana and What Happens When the Founder Doesn’t Come Back
If Dines and Bhusri are the boomerang success stories — or at least the boomerang hope stories — Asana is the cautionary tale of what happens when a founder walks away right when AI demands founder-level conviction.
Dustin Moskovitz co-founded Asana in 2008 after leaving Facebook, where he’d been Mark Zuckerberg’s roommate at Harvard and the company’s first CTO. He built Asana into a respected work management platform with over 150,000 customers, more than $700 million in annual revenue, and penetration into 85% of Fortune 500 companies. For 15 years, he was the CEO.
Then, in March 2025, he announced he was retiring.
The stock dropped 25% overnight.
It wasn’t that the quarterly numbers were terrible — Q4 revenue was up 10% year-over-year and the company was approaching breakeven. It was the signal. In the middle of the most consequential technology shift since cloud computing, the founder was walking away. And his stated reasons made it clear this wasn’t a strategic succession — it was exhaustion. Moskovitz later told the Stratechery podcast that the CEO role was “exhausting” and fundamentally ill-suited to his personality as an introvert. He said he’d never intended to be CEO in the first place, that he’d wanted to be head of engineering, and that “one thing led to another.”
He left to focus on philanthropy — specifically, ironically, on AI safety through Open Philanthropy.
The market’s reaction told you everything. The day of the announcement, the revenue guidance for fiscal 2026 came in at $782–$790 million, roughly $13 million below consensus. Dollar-based net retention was stuck at 96% — below 100%, meaning the existing customer base was shrinking on a net basis. Revenue growth was decelerating from 11% to the high single digits. Jefferies cut their price target and warned of “uneven macro and tech sector headwinds.”
And then it got worse.
The board hired Dan Rogers — a solid operator with stints at ServiceNow, Rubrik, and LaunchDarkly — as the new CEO, starting July 2025. A professional operator. The classic playbook. But Rogers walked into a buzzsaw: the SaaSpocalypse, collapsing SaaS multiples, and an AI-native competitive landscape where Microsoft, Monday.com, and a dozen AI-first startups were all converging on Asana’s core territory.
The numbers tell the story. The stock that once traded at $142.68 at its November 2021 peak closed at $12.93 when Rogers was named CEO. It hit an all-time low of $7.15 on February 12, 2026 — right in the teeth of the SaaSpocalypse selloff. That’s a 95% decline from peak. Market cap shrank to under $1.8 billion. Revenue growth continued to decelerate to around 9%. NRR stayed below 100%.
Moskovitz, to his credit, kept buying shares throughout the decline — he owns over 50% of the voting power and roughly 39% of outstanding shares. He’s not abandoning the company financially. But he did abandon the CEO seat at the worst possible moment.
Here’s the brutal contrast: Daniel Dines saw that AI threatened to make UiPath’s core product obsolete and ran back into the building. Aneel Bhusri watched $1.3 billion in personal wealth evaporate and ran back into the building. Dustin Moskovitz saw the same existential threat — he literally funds AI safety research — looked at the burning building, and said “I’m an introvert, this job is exhausting, good luck to the next person.”
That’s not a criticism of Moskovitz as a person. He was honest about his limits. Not every founder is wired to be a wartime CEO. But it is a cautionary tale for every board in enterprise SaaS right now: when the platform shift arrives, the founder’s departure isn’t just a leadership transition. It’s a signal — to employees, to customers, to the market — that the person who cared most about this company’s existence has decided the fight isn’t worth fighting anymore.
Asana may yet recover under Rogers. The AI Teammates product is showing early traction. Margins are improving. But the stock price tells you the market’s verdict: without the founder’s conviction at the helm, investors aren’t willing to pay up for the turnaround.
In the Age of AI, founders don’t get to retire gracefully. They either come back, or they watch from the sidelines as someone else tries to reinvent the thing they built.
When The Playbook Breaks, The Founders Have to Rebuild It
Here’s what boards don’t want to admit: the traditional CEO succession playbook — founder builds, operator scales, everyone gets rich — was designed for a world of steady-state competition. You optimize the machine. You expand internationally. You move upmarket. You cross-sell.
That playbook breaks when the entire machine needs to be rebuilt.
AI isn’t a feature. It isn’t a product cycle. It’s a platform shift on the scale of cloud computing itself — and Bhusri is right that it may be even bigger than SaaS was. When the market wipes $285 billion from software stocks in 48 hours, when Forrester is writing about the “SaaSpocalypse,” when Jason Lemkin at SaaStr is pointing out that every dollar going to AI infrastructure is a dollar not going to another SaaS seat — this isn’t a drill. This is real.
And in moments like this, the people best equipped to respond aren’t the ones who learned how to run the company. They’re the ones who imagined it in the first place.
Dines imagined a world where software robots could free humans from repetitive work. Now he has to reimagine that vision for a world where AI agents can think, reason, and act. Bhusri imagined a world where cloud software could transform how enterprises manage people and money. Now he has to reimagine that vision for a world where AI might make the seat-based model obsolete.
They didn’t choose to come back. The world chose for them.
Who’s next?
The boomerang founder era isn’t nostalgia. It’s survival.
