Operational excellence foundations differ by industry: Cyrus Mistry of GoGuardian

The operational excellence foundations that define a COO at an airline have almost nothing in common with the ones that define a COO at a software company. Cyrus Mistry has worked close enough to both to explain the difference, and it reframes what the title actually means.

Mistry is COO at GoGuardian, an edtech platform used by 22 million students across more than 10,000 schools, helping over 500,000 educators manage classroom devices and improve student safety. The company raised $200 million from Tiger Global at a valuation well above a billion dollars. Before that he was director of product management at Google, covering education and SMB platforms, and he holds four degrees from Johns Hopkins plus postgraduate work in natural language processing and machine learning.

Developer, then technical sales, then twenty years of product

Mistry started as a developer, and his career changed three weeks into his first job.

He presented his development project, and the reaction was not about the code. Someone noticed he could stand up and talk about it, and enjoyed doing so, and suggested he move toward technical sales, which he points out is one of the most important roles in enterprise software.

He did that for a few years, at a company whose sequence he describes as backwards: sell, then design, then build. Selling something and then finding someone to construct it pushed him toward product management, where he spent the majority of a twenty five year career. The COO title arrived only recently, as an expansion of a chief product officer role.

How the role got drawn at GoGuardian

GoGuardian's products divide into keeping students safe and helping them learn, with some doing both. Mistry's example is the teacher-facing product, where an educator can chat with students and keep them on task while also pushing a link to every screen at once.

He was hired to run product, and then engineering. As the conversation developed, the company added strategy and operations, the business intelligence team and the research organization. There was no COO, so the title followed the scope.

His timing was unusual. The Tiger Global round closed three or four weeks after he joined, and he is candid that it shaped his confidence in the decision.

The reasoning is worth borrowing. When you can ride on someone else's diligence, take it. His analogy is the reason companies love recruiting ex-Googlers: the hiring bar was high, someone else already did that work, and you can draft off it. Tiger had turned the company inside out before investing, so his read was that if Tiger loved it, he had little to worry about.

Leaving fifteen years at Google

Mistry knew GoGuardian long before joining. He was the first product manager to bet that Chromebooks would work in schools and businesses, and by the time he left he was the longest-serving product manager on Chrome OS. He had met GoGuardian's founders within their first year or two.

Three things moved him. He knew the market potential. He wanted to run an entire organization, noting that at Google, even as an executive, there are a great many senior directors and vice presidents who want to know what is happening, and that moving quickly is its own reward.

The third reason is personal and stated with unusual precision. Fifteen years at one company, eleven of them on a single product. He did the arithmetic at 44 and worked out that a quarter of his life, one in four waking days, had gone to one product.

What transfers, and what does not

Asked what is different between running a product area and being COO, Mistry starts with the overlap.

Leading a product line means being its general manager. You own the P&L, the customer satisfaction, the growth rate and the innovation, and it sits on your shoulders. A COO who also holds product and engineering does exactly the same thing at a larger scope. His conclusion is that strong generalists with real business grounding, people who can act as general managers, are well suited to the COO seat.

Then the surprise, which he says he still repeats to his own CEO and CFO.

If someone had told him before joining that 80% of the job would be people, he would not have believed it. He expected product strategy and execution. What he got was promotion requests, resignations, backfills, where to hire and why hiring velocity was off. Every single thing, in his words, is about people.

He is quick to say this is a good thing rather than a complaint. Getting the right people and the right leaders in place, keeping them engaged, doing the diversity and inclusion work properly, is the job. He simply had no idea how large a share of a COO's time it consumes.

Operations means different things in different industries

This is the section that makes the episode worth sending to people.

Mistry sketches a spectrum. At one end sit companies whose center of mass genuinely is operations. Airlines are the canonical case, complex enough to have produced an academic discipline and, he notes, actual films about the operation of an airport, before you even get to coordinating crews, weather, boarding and baggage. His marker of how remarkable it is: turning an aircraft around in twenty minutes.

He also worked at Disney for years, where roughly 80% of the company reported into the COO. Park operations, hotel operations, food and beverage across the resorts. Everything about staffing that organization, where people are needed and when, is operations.

Software is different. Mistry's careful distinction is that a cloud provider is not a software company in this sense. Google, like Amazon and Microsoft, runs an enormous physical operation, with data centers, parts inventory and just-in-time logistics. That is real operations work.

Software engineering runs on a different cadence, until a product becomes a service at scale and moves into an operations group of its own.

His conclusion is the useful one for anyone comparing job titles across industries. A COO in software is not doing the same job as a COO of park operations at Disney or an airline. The category of headaches is fundamentally different, which he notes with some gratitude.

Being careful about what you import

With Google behind him and a new CTO who ran Google Beijing and Google Pay engineering, Mistry is deliberate about what practices come along.

Some things Google does are excellent. Others he considers legacy weight, and he names one specifically: promotions, which he says are consistently among the lowest rated aspects of working there.

His diagnosis of the general problem is that people cannot tell what they need to do to be promoted, or why a colleague was promoted and they were not. GoGuardian's response is a project to clarify the career ladders so the answer becomes checkable rather than subjective, and to remove bias, including the version where a manager simply likes someone.

He is even-handed about Google's attempted fix. Promotion committees deliberately excluded your manager and made a third party determination from your packet. The problem was context. Without it, a different bias entered.

His illustration is memorable. Someone from Search changes the shade of blue on an ad and generates a very large number within days. Someone else spends seven years on a product, launches it proudly, and generates a fraction of that. The Search reviewer reads the second person as having added little value. His verdict is that separate promotion committees have real pros and cons, and that how you handle promotions and career movement deserves careful design either way.

Scaling engineering by scaling management

Mistry's view on growing an engineering organization is about the middle layer.

You cannot scale with one strong leader at the top because not everyone can report to them. What matters is having strong people in the middle, and keeping the bar high, since A players hire B players and B players hire C players.

He is realistic about the market pressure working against that. Two decades after everyone concluded computer science was the field to be in, the supply and demand imbalance has still not resolved, because every job and every company now needs technology. The easy response, when you cannot find good people, is to lower the bar and get someone into the seat. His position is to hold the bar and pay for genuinely top talent instead, where you can.

The related priority is one more companies should adopt. Looking at what technology talent costs, his first instinct is not recruiting, it is making sure his existing team is paid commensurately so they do not leave, in a market where everyone is being approached and everyone knows what everyone earns. What keeps him up at night is regretted attrition rather than open roles.

Deploying the money

On the $200 million, Mistry splits it into organic and inorganic.

Organic is growing the product and engineering teams, and his enthusiasm is about the leverage available. GoGuardian achieved deep penetration in K-12 with a remarkably small engineering team, so adding people could take them from shipping a handful of features to shipping many times that.

On acquisitions, his framework is a checklist of reasons: people, resources including capital equipment, customers, revenue that opens a new area, and intellectual property that accelerates what you are building.

His standard is that a deal should tick more than one box. Buying a company purely as a fast way to hire is not necessarily bad, since you acquire a specific skillset rather than just headcount. But acquiring software you want, with strong people, plus customers and revenue, is a four-way win, and that is what justifies being selective.

The cruise, and two million Chromebooks

Michael's standing question produced the best crisis story the show has collected.

Mistry admits up front that he does not really take vacations, because he brings his laptop, which makes it working from somewhere else. He is explicit that this is not a model he wants for his team, and that he is working on disconnecting properly.

He had won a cruise to the Bahamas, bought the internet plan, and promised his wife he would stay off the laptop. On day two of seven, checking his phone, an email arrived saying there was a problem with Chromebooks.

A push had broken Wi-Fi on a very large number of devices. His first thought was to ship a fix, and the answer was that they could not, because the affected devices could not reach the network to receive it. He recalls the number as somewhere in the low millions.

These were K-12 students. He spent nights in the ship's library on video calls while his wife slept, working out how to communicate with schools and how to even identify which devices were affected, since the broken ones could not report back.

The resolution came from the engineering team and was elegantly indirect. If a device could be manually connected to any guest network, it would check for an update, receive the fix, and restore itself to the correct network. The complication was that many schools did not have a guest network, so they had to create one or use a teacher's phone hotspot.

His summary is the interesting part: many schools worked it out independently and moved quickly, which turned the incident into a communications project more than a technical one.

The 5 things I took away from this conversation

1. Compare COO roles within an industry, not across them. An airline COO, a theme park COO and a software COO share a title and little else. Mistry's spectrum is the clearest explanation I have heard of why benchmarking your role against the wrong industry produces bad conclusions about what you should be doing.

2. Draft off someone else's diligence. Tiger Global had already investigated GoGuardian exhaustively before Mistry joined, and he treated that as evidence rather than doing it all again. The same logic applies to hiring from companies with famously high bars. It is a rational shortcut, provided you know whose work you are trusting.

3. Expect the job to be mostly people, and be glad about it. Mistry thought he was signing up for product strategy and found 80% of his time going to promotions, departures, backfills and hiring velocity. Anyone stepping into a COO role for the first time should plan their calendar around that reality rather than being surprised by it.

4. Retention is cheaper than recruiting in a hot market. His first move when looking at compensation data was making sure current employees were paid correctly, not chasing new candidates. In a market where everyone knows what everyone earns, protecting the people who already know your product is the higher-return activity.

5. Promotion clarity is a bias control, not an HR nicety. When people cannot articulate what earns a promotion, the decision defaults to who a manager likes. Google's committee fix removed the manager and imported a different bias by removing context. Either way, the design deserves real attention rather than inheritance.

FAQ

What are the operational excellence foundations in a software company? Less physical and logistical than in operations-heavy industries. Mistry contrasts airlines and theme parks, where staffing, scheduling and physical throughput dominate, with software, where the work is closer to general management of products and teams. He notes cloud infrastructure is the exception, since data centers involve genuine physical operations.

What does a chief operating officer do in a company like GoGuardian? At GoGuardian the COO owns product and engineering, plus strategy and operations, business intelligence and research. Mistry describes the underlying job as being the general manager, owning the P&L, customer satisfaction, growth and innovation, with the majority of actual time going to people decisions.

Does product management prepare you to be a COO? Mistry argues yes, because leading a product line means acting as its general manager and carrying full ownership of its outcomes. What it does not prepare you for is the share of time spent on hiring, promotions, departures and team structure, which he estimates at around 80% of his week.

How should a company handle promotions to reduce bias? By making the ladders and criteria explicit enough that whether someone qualifies is checkable rather than subjective. Mistry notes that Google's promotion committee approach removed the manager's bias but introduced another, because reviewers lacked context about the difficulty and impact of work in other parts of the company.

What justifies an acquisition? Mistry lists people, resources, customers, revenue that opens a new market, and intellectual property that accelerates the roadmap. His test is that a deal should satisfy more than one of those. Acquiring purely for hiring speed can still be reasonable if it delivers a specific skillset, but the strongest deals combine several.

Also mentioned

• GoGuardian and its student safety and classroom management products
• Tiger Global, whose $200 million round closed weeks after Mistry joined
• Chrome OS and Chromebooks, where Mistry was the longest-serving product manager and led the push into education
• Disney, where he saw an organization with roughly 80% of the company reporting into operations
• Industrial and operations engineering at the University of Michigan, Mistry's example of operations as an academic discipline
• Cyrus Mistry on LinkedIn

Listen to the full episode

Cyrus Mistry on Between Two COO's

Between Two COO's is hosted by Michael Koenig. Subscribe on Apple Podcasts, Spotify, or wherever you listen.