Five (ish) Exits
Recalling my IPOs and acquisitions
My current employer had a successful initial public offering (IPO) a few months ago. In the vocabulary of venture capitalists, such a development is called a “liquidity event” or an “exit”: it is an opportunity for the investor to reap some or all of their reward, converting the paper wealth of startup stock into cash. By one common rule of thumb, 9 out of 10 startups fail. Accordingly, it’s important for the one out of 10 to succeed in a big enough way to make up for the ones that failed. Another way of thinking about it is to say that startup companies, like songs, movies, and TV shows, are a hit-driven business. A few big winners pay for the vast collection of losers.
This most recent IPO is my fifth exit. That would be a small number if I were a venture capitalist, but is a relatively high number for a technical person like myself, and is arguably an indicator of my good luck. (A careful accounting might say that this is my 4th exit: three are indisputably what a venture capitalist would consider exits, but two of them need to be footnoted with disclaimers and so should only probably count for half each.)
Exit #1: Lotus
My first exposure to this kind of event came when I was at Lotus and it was acquired by IBM in 1995. I had been granted some Lotus restricted stock units as part of my compensation package but didn’t really think very much about them, since at the time the company was not doing terribly well… which in turn was part of the reason why IBM bought them.
I don’t think the Lotus sale itself was enough to make me a “startup guy,” but it was an interesting introduction to the idea that stock-based compensation could turn into money. Before that, I had just assumed that I would be a researcher working for my salary, and that would be how my finances would work for the rest of my life. The amount of money involved was not very much, certainly in retrospect; I think it was about $13,000. It did make real for me the idea that one could be paid in stock, which was then rendered into cash at some subsequent point. When the Internet bubble happened, (which is when I really started being a startup guy) I had at least that modest experience to convince me that there was a plausible payoff.
Exit #2: Sightpath
That said, my motivation for joining my first startup wasn’t money. It was more like a chance to play with the cool kids, doing what the cool kids were doing. I remember that one piece of skepticism when I was interviewing with the CEO of Sightpath was that I had only worked for big companies like IBM or Eli Lilly on my official resume. Fortunately, I had also worked with my mother in a gift shop and engraving business, and accordingly was very comfortable with the real-world issues of small entrepreneurial businesses. I understood the need to make the sale, the need to market relentlessly, the need to pivot as market expectations shifted.
Sightpath was a company founded by people I knew from graduate school at MIT. I was able to step into a role there as a kind of “spare CTO” at roughly the same time that one of the cofounders was winding up their leave to return to their full-time responsibilities at MIT. Of my exits, Sightpath is the winner in terms of the shortest time from when I joined the company to when the exit happened (about 4 months), which is partly a reflection of how frenzied the deal making was during the dotcom bubble.
I was blissfully unaware of anything happening in the lead up to the deal announcement. In retrospect, it’s funny that I went hiking with the company’s CTO the day before the announcement. He asked me a vague question about what I would do if something dramatic happened at the company. I don’t remember anymore what his exact phrasing was, but I do remember that he asked me a question that I interpreted at the time as a random speculation about the future, when it was probably considerably more pointed and focused on the question of whether I was going to head for the exits if the company was sold (which it was, the next day).
Cisco bought Sightpath in 2000 for stock that was then worth $800 million, which was quite a coup. The timing was also excellent, because the crash happened soon afterward. Although Cisco stock was no longer worth what it had been at the point of the deal, it only lost about a third of its value, whereas I knew many people working for other companies where their stock lost 90% to 100% of its value. (I’ve previously written about six-pack stocks and double-six-pack stocks in the context of the AI bubble, which seems to have echoes of the dotcom bubble). It’s worth noting that the kind of massive collapse in value that occurred in that bubble meant that some people held stock that was worth less than the taxes that were owed on it: they might have acquired some stock that was worth (say) $100, incurring a tax liability of between $15 and $37, and then found themselves with stock that was worth $10. Multiplying those numbers by thousands or even millions gives a sense of both how heady it could be, and how damaging it could be if one made poor choices.
Compared to the IBM acquisition, the Cisco acquisition was a big deal in terms of the money I ultimately realized. I was able to arrange my life so that I mostly don’t have to think about money, as long as I avoid certain exotic expenses like real estate, cars, boats, or planes.
Or, more accurately, money is usually not the limiting factor in what I might want to do: instead, I find that my time, energy, or attention is the scarce element. I find that I do more thinking about money overall, in the sense of being concerned about whether I am pursuing the right strategies and have sensible advice with respect to my investments. It’s also still the case that I didn’t make enough money to retire… or, at least, I didn’t feel like I had enough money to retire.
Exit #3: Riverbed
Post-acquisition life at Cisco was a stressful experience, full of politics and bad technical decisions, but it did have the merit of leading me to Riverbed Technology. I joined Riverbed because it was one of the companies that we were considering partnering with at Cisco, and it seemed to be clearly the preferable choice in my analysis. When Cisco bought a different company (not as good, in my opinion), I knew that my next task at Cisco would be to put lipstick on that pig. Rather than doing that, I joined Riverbed.
My reasoning at the time was that failing with Riverbed would be more fun than succeeding at Cisco; succeeding at Riverbed would be way more fun. Happily, it turned out as I might have hoped.
One key element of our success at Riverbed was my knowledge from having been at Cisco. Not that I ever disclosed Cisco confidential information, but my experience with that business unit’s culture meant that I had a high degree of confidence that they would not be very good at the things they tried to do. When Cisco people made competitive claims or technical arguments, it was easy for me to understand the flaws in their reasoning and the corners they were cutting. The first few years of competing with Cisco were effectively a larger scale version of the evaluation process between the two companies that I had previously participated in, but now the market was judging whether my arguments or the other arguments were better. Happily, the market seemed to be convinced by the Riverbed architecture. Indeed, after a few years, Cisco adopted a variant of the Riverbed architecture for its competing product (although they didn’t describe it that way, of course).
The Riverbed IPO happened a few years after I joined the company, but the time had flown by. The Riverbed IPO has the distinction of being the best-performing IPO that year.
I continued to have a good time at Riverbed for a number of years, but eventually felt that it was not working out anymore. The exciting parts of the technology seemed to be in the past, the future of the company seemed to be a bit unclear, a number of attempts at diversification failed, and I had an opportunity to go teach at MIT. So I went to go teach and write a book.
After I left, Riverbed was sold to private equity. As far as I can tell, as an outsider, it has been stumbling along in various ways ever since, including two bankruptcy filings. Overall, whatever “brand equity” I had from having been “early at Riverbed” is largely gone – it’s no longer a company that carries the shining reputation it once had.
Exit #4: Dropbox
I joined Dropbox when I realized that I wasn’t especially keen on retirement. Dropbox had a great culture, and I enjoyed my time there, but I never really found my niche. All my experience was in enterprise software and systems, whereas Dropbox was very much focused on consumers and small businesses. At the point where they hired me, they were going to pursue the enterprise market; so there was a brief time when there seemed to be great alignment between myself and the company.
Unfortunately, for reasons that I don’t really understand, that direction was abandoned. At the same time, my manager left the company. He was the CTO, with plenty of money from his previous work, and decided to stay home with his small child. I can’t say that I fault him for the decision, but it did have the unfortunate impact of leaving me somewhat adrift in the organization. In retrospect, one lesson is that I should have started to leave after he left and I wasn’t well-integrated elsewhere.
I continued to do what I could to make things better in various ways, but another lesson from those days is that it was hard for me to assemble big initiatives that could have a big impact. I’m happy to join in with other people’s initiatives and help them succeed and fix their flaws, but I’m not necessarily the best person to point at a distant dot on the horizon and marshal the forces to get to that dot.
Dropbox had a successful IPO in 2018. I was no longer employed by Dropbox at that point, but I had been allowed to keep my options and I made a little money – better than the IBM/Lotus deal, but not by much.
Exit #5: Netskope
Netskope’s successful IPO in 2025 could be faulted in pure venture capital terms, compared to my other exits: of the startups I’ve worked for, it was the longest time from my hiring to the exit, and the company required more funding rounds than any of the others.
Still, I remember a wry remark from my graduate school days about what constituted a “good” dissertation: one that had the requisite number of signatures from faculty members. Likewise, I think that any IPO is a Big Deal, and I feel very lucky to have been able to go through the process twice (or perhaps 2.5 times if we count Dropbox). In contrast to my time at Riverbed, the work at Netskope has been about fixing a deeply flawed early implementation to realize the potential of a superior architecture. It’s also been a huge contrast to Riverbed and Dropbox in terms of future potential: in those companies there was an ongoing challenge of finding the “next big thing,” after the one amazing thing the company did well. In contrast, at Netskope the challenge is to avoid being spread too thin on the sheer profusion of valuable “next things” our customers want us to do.
Lessons
So, what conclusions can be drawn from my five (ish) exits? I’d highlight the following ideas.
The team matters. At Sightpath, Riverbed, and Netskope, for different reasons, I had personal connections to key players. That is easy to misunderstand as just some kind of “old-boys network” (and, certainly, a flaw I perceive in all these tech companies is how male-dominated they have been). But for me the value came in two forms. First, there was an initial level of trust that gave me scope to find and define a role for myself, without needing to be slotted into a specific job. Second, there was an implicit shared level of excellence and integrity. Over time, the organization figured out who wasn’t competent, who wasn’t willing to work to get better, or who was sleazy, and pushed them out. Those of us who had worked together before didn’t have to spend time debating standards, or whether certain behaviors were problematic.
The work differs. Although I have been a “distributed systems guy” for my entire career, that has had entirely different intersections with the technology and market in each of these companies. At Lotus, I was a researcher and standards guy. At Sightpath/Cisco, I was an architect and standards guy. At Riverbed, I was a competitive guru (and led the largest team of my career). At Dropbox, I evaluated system performance. At Netskope, I have focused on failures, uptime, and (recently) AI.
I’m now in my early sixties. I still feel young and healthy as a person (for which I am grateful) but I recognize that I am now quite old by the standards of the tech business. Although I can still remember the days when I was consistently the youngest person in the room, I’m now typically the oldest person in the room. Accordingly, it seems unlikely that I will be pursuing any additional exits in the future… but then again, who knows?