How to Win in a Winner-Take-All World

This is a summary of a moderately good 🌿 branch book. I intend this post to be a comprehensive summary, but you should buy the book if you're interested in the career case studies behind Irwin’s ideas. Read more about book classifications here.

Neil Irwin’s 2019 book How to Win in a Winner-Take-All World is a career strategy book that is best seen as a collection of tactics in the face of macroeconomic change.

In all honesty, I hesitate in calling this a ???? branch book. Irwin’s list of tactics isn't cohesive — he isn’t making one point with a number of sub-arguments. Instead, the chapters have the most tenuous of connections, and sling us from baseball to Microsoft to GE to management consulting to law.

I think the easiest way to make thematic sense of How to Win is to consider each tactic that Irwin presents as a response to some underlying structural shift in the economy. I’ll organise my summary accordingly: first, I’ll describe Irwin’s take on some underlying macro-economic change. Then I’ll present his tactical suggestions for dealing with that change. Finally, I’ll critique it, introduce supporting material for it, or perhaps even applaud the tactic; the idea is to present Irwin’s ideas the same way one might thread a necklace of pearls — without care for specific order. You may read this summary in one sitting, or you may jump around, it's really up to you.

Two last bits of administrivia before we continue.

First, many of the ideas Irwin discusses dovetails nicely with our previous discussions on career moats — and I’ll reference those ideas as we go along. If you’re new to this blog, you might want to skim this overview post for proper context.

Second, I can’t help but think of a 2016-era quote from Charlie Munger that goes “microeconomics is what you do, macroeconomics is what you put up with.” This applies to jobs as much as it does investing.

Irwin’s book may begin with ‘macroeconomics’, but he always eventually ties what is true in the large to what is practically useful in the small. The end goal of How to Win is to ‘help ambitious people navigate a career in this changing economy, much as a sailor must understand winds and currents.’

I don’t think this book is the definitive take, but Irwin’s ideas are worth considering ... perhaps because there are so few people writing about this topic in the first place.

Table of Contents

1. The goal is to be a Pareto Optimal Glue Person
2. How to become Pareto Optimal?
3. Treat Career Moves Like a Lattice, not a Ladder
4. Work is like a Sports Team, not a Family
5. When is Freelancing Lucrative?
6. Using Big Data to Improve Performance
7. Management Ability as Competitive Advantage
8. Navigating the Winner-Take-All World
9. Responding to Software Eating the World

1. The Goal is to be a Pareto Optimal Glue Person

The first idea that Irwin explores is the biggest takeaway of How to Win — that the ‘most’ lucrative goal in today’s economy is to be something he calls a ‘Pareto Optimal glue person’ for your industry.

The macro: Companies today are bigger and more complex than in the past. This is partly due to globalisation (the supply chain for most businesses cross national boundaries, timezones, and jurisdictions), and partly due to technological progress (it takes way more technicians to make King Kong today than it did in 1933).

Complex, global organisations need a special class of worker to function properly. Irwin calls this class of worker a ‘glue person’ — and the prototypical example he gives to illustrate his point is Marco Revelant of Weta Digital, ‘the man who groomed King Kong’. Revelant had a background as a 3D artist, but was ultimately responsible for creating Barbershop — the Academy Award winning software that allowed artists in Weta Digital to ‘groom’ the fur of hundreds of computer generated apes in Rise of the Planet of the Apes. Revelant did not have a background in software engineering. And yet he was integral to the creation of Barbershop:

“He doesn’t actually look at the code,” explained Pablo Selva, the head of software engineering at Weta Digital. “But he understands how something like the elastic rod dynamics on hair work. He’ll understand the logic behind it even if he may not know how to implement it himself. It’s not a skill many artists have. If all of the artists were able to do that, it would be much easier to talk to them.”

Revelant is but one example (glue person, film industry) in the book. There are many others.

The tactical recommendation: Irwin asserts that ‘becoming a glue person’ for your specific industry is one of the best career bets you can take. Changing industries result in changing firms. Changing firms result in career opportunities. Irwin believes becoming a glue person is an opportunistic adaptation to a changing environment.

But ‘glue people’ as a term gets a bad rep. Irwin acknowledges that the phrase can mean ‘useless bureaucracy’. How do you differentiate good glue people from bad glue people?

Irwin eventually settles on the following definition: a good glue person is ‘Pareto optimal’. This is a fancy way of saying that good glue people ‘cannot possibly have greater ability in one area relevant to their jobs without sacrificing something in another’. Revelant can’t spend more time boning up on software engineering without giving up his ability to work and think like an artist.

Useless bureaucrats, on the other hand, are not Pareto optimal. They are the definition of a bad glue person.

I think Irwin’s notion of ‘Pareto optimality’ is an interesting take on having a ‘rare and valuable combination of skills’. We’ve covered Cal Newport’s argument that there are two paths to building a valuable career: a) you can be the best in your field at some skill, or b) you can acquire a rare and valuable combination of skills. Irwin focuses exclusively on the latter path, and makes little mention of the former.

2. How to become Pareto Optimal?

If becoming a Pareto optimal glue person is the goal, how do you actually become one? In simpler terms, how do you determine the exact combination of skills that are valuable, and how do you go about pursuing those skills?

The macro: In a changing industry, certain firms are more competitive than others. Those firms will respond more rapidly to disruption. Jobs at Tesla, for instance, are more indicative of the future of the automobile industry than jobs at GM. Volvo, which announced in 2019 that they were going to switch all cars to electric and hybrid drivetrains within a single year, is more likely to experience a shift in labour requirements.

The tactical recommendation: Changes in labour requirements will show up in job descriptions. And they’ll show up first in more innovative companies, before spreading to the rest of the industry.

To understand the opportunities available to you, Irwin argues that you should periodically read JDs from firms that are considered at the leading edge of your field. The skills they ask for will give you a clue as to the evolution of skill demand in your industry.

Irwin tells the story of Burning Glass Technologies, a software firm that had the bright idea of analysing job descriptions to figure out skill demand. Starting in 2007, BGT began scraping every job listing in the country. They discovered several useful patterns in the data.

For instance, the length of time a JD remains up gives BGT a possible signal as to the difficulty of hiring for that role. BGT then investigates such aberrations, in order to identify the skills that make the new roles so difficult to fill. An episode from 2014 demonstrates this rather clearly:

Sigelman noticed a curious pattern with a surprising category of jobs. Many schools and camps in the United States are associated with a church or other religious institution and employ a “director of religious life” to guide the spiritual experience of their enrollees. Usually, these jobs are pretty easy to fill; countless graduates of divinity schools, or just devout people who have the right personality, tend to apply. But according to Burning Glass’s data, suddenly those jobs were starting to take much longer to fill on average, for no obvious reason.

“I remember first looking at the numbers and saying, ‘Guys, we are clearly doing something wrong in our data, because there is no shortage of religious talent in the United States of America,’” Sigelman said. When they dove deeper, what they found was not so much a general shortage of people looking to work in religious life. Rather, camps were increasingly looking for a new skill set among their directors of religious life. “Most of these jobs were clearing very fast, and there was no holdup,” Sigelman said. “But there was this corner of them where they were looking for religious life directors who had social media skills and database marketing skills—people who could really market to young folks. But there weren’t too many people with that combination of skills, and that was driving the overall average to make it look like there was a shortage.” The problem, in other words, wasn’t that there was a shortage of people interested in religious life but that there was a shortage of those people who also knew how to do database marketing.”

There are other useful patterns in the data. For instance, when companies post a JD, they tend to emphasise the qualifications that they are most struggling to find.

Irwin gives the example of a pharmaceutical company hiring an in-house lawyer — in this situation, the company might not bother specifying some of the skills and qualifications that a lawyer is expected to have. Instead, such a JD would emphasise knowledge of medical science — expertise that is relatively rare in lawyers but particularly useful for lawyers in the pharmaceutical world.

BGT’s data also shows that jobs that demand combinations of skills tend to skew higher in terms of salary. The average advertised salary for an engineering manager is $95,000 a year, but goes up to $120,000 if the JD also specifies strategic planning experience. And on it goes — Irwin asserts that this is a direct consequence of supply-demand dynamics: the rarer the combination of skills; the higher the salary.

The major takeaway is the following:

In every sector, there are some companies that are known for being on the cutting edge of where the industry and its talent needs are going, and others that are slower to adjust. Even if you work at one of the laggard companies, it is beneficial to monitor job postings by the cutting-edge companies. The skills they seek are likely to offer a window into the future of the job in question. Your own employer may not realize it needs X skill yet, but if its highly innovative competitor is trying to hire a lot of people with X skill, seek out every opportunity to learn more about X.

Irwin closes with a concrete example of Tesla vs GM from BGT’s data: 54% of Tesla mechanical engineer postings sought experience with CATIA 3D design software, and 19% sought experience with 3D modelling and design more generally. These requirements were less prevalent in the GM listings.

BGT believes if you’re a mechanical engineer in the auto industry, and if you believe the rest of the industry will look more like Tesla over the next decade than it will GM, then it would be a better bet to gain experience with 3D modelling techniques today.

3. Treat Career Moves Like a Lattice, not a Ladder

Irwin notes that C-level execs aren’t exempt from this shift towards combinations of skills as opposed to specialisation. He cites Custodio et all in a 2012 paper, finding that ‘generalist’ CEOs earned a premium of 19% in annual pay — or almost an additional $1 million a year in extra compensation. I’ll note that this is a single study, so I wouldn’t read too much into it; I found the macro changes that led to this shift to be far more interesting.

The macro: In the aftermath of the 2008 financial crisis, large companies cut layers of middle management, with the aim of reducing costs. These jobs have largely vanished for good — with consequences for those who are intent on climbing the corporate ladder.

(Related: I’ve summarised economist Tyler Cowen’s views in Average Is Over elsewhere on this blog; Cowen believes that the Great Recession has resulted in a permanent contraction of the labour market. He attributes this to the twin influences of automation and outsourcing. Irwin's ideas are likely an instance of this broader economic trend.)

Irwin cites research from the human resources group at Gartner, that’s found that the post-2008 reduction in middle-management have caused promotions to become rarer and larger — where before a manager could expect to be promoted from running a team of five to a team of ten after two years, a rising manager would now have to wait five years and go from five subordinates to thirty.

This meant two things: first, workers in such companies were getting more frustrated as they went without a promotion for longer, and when managers finally made the jump, they were more likely to fail due to a lack of relevant expertise. Brian Kropp, the head of the human resources group at Gartner, said that he and his colleagues began advising companies to offer career advancement in the form of varied experiences, since they weren’t able to offer advancement in terms of linear promotions.

In the book, Irwin quotes Kropp:

“As an employee,” Kropp said, “you have to think about ‘What do I need to get on my résumé in order to be promoted at some point in the future?’ rather than ‘What task do I need to achieve this quarter to get promoted now?’ You can’t be thinking about your job anymore as just ‘How do I have a career in this company?’ You have to be thinking about, and this sounds incredibly mercenary, but you have to be thinking about ‘How do I get promoted at any company? And what are the skill sets that are going to help me be successful no matter where my career leads?’”

The tactical recommendation: The overarching point that Irwin makes in this chapter is to argue that the notion of a career ladder is dead. He’s in fairly good company here; Sheryl Sandberg, for instance, asserts that careers today are less like a ladder and more like a ‘jungle gym’:

”As you start your post-HBS career, look for opportunities, look for growth, look for impact, look for mission.  Move sideways, move down, move on, move off.  Build your skills, not your resume.  Evaluate what you can do, not the title they’re going to give you.  Do real work.  Take a sales quota, a line role, an ops job.  Don’t plan too much, and don’t expect a direct climb. If I had mapped out my career when I was sitting where you are, I would have missed my career.”

Irwin calls career advancement in today’s world ‘navigating the career lattice’ — and asserts that you can make a jump up for pay, or a jump across, for experience — but rarely both at the same time. He then offers three recommendations based on this underlying change:

1. Don’t evaluate job offers based on promotion or pay alone. Instead, evaluate in terms of skills and experiences that would put you in a good position to have even more interesting options down the road, no matter how the economic winds shift.
2. Train yourself to have a ‘three-year itch’ in every job In the first year of any new job, you’ll be learning your way around. In the second year, you’ll be making change happen. And then finally, in the third year, you’ll come into your own, which means you’re likely to start feeling comfortable. This comfort means you won’t be learning as much; which in turn means that you should probably find opportunities for learning elsewhere (or through some other means in the same organisation).
3. You won’t know where the hops are leading you while you’re doing it. Many of the case studies in How to Win are of people hopping from opportunity to opportunity. The common denominator in each of these stories are that the hops only make sense when you’re looking back, with the benefit of hindsight. In practice, when you’re making the jumps, it’s really difficult to see where all the combination of skills will eventually get you. This gets at the heart of Sandberg’s argument — that one shouldn’t have a concrete plan while navigating one’s career. Irwin, in turn, urges you to embrace the uncertainty. What matters, he says, is that you make the hops and gain the experience; you can always form a coherent narrative of your career later, with the benefit of hindsight.

4. Work is like a Sports Team, not a Family

If work today consists of a series of hops, how do you think about loyalty to your employer? This chapter is essentially the story of the Netflix culture deck (otherwise known as ‘the most important document ever to come out of Sillicon Valley’). Irwin uses it as the central story around which to explore the changing relationship between employer and employee.

The Macro: Companies are no longer loyal to their employees. Perhaps they never were (or perhaps they were only ever loyal to a small subset of their employees); the reality is that increased competition from globalisation, coupled with an increased willingness to do layoffs make for a highly transactional relationship between employee and employer.

You should be highly suspicious of assertions that ‘working at X is like being in a family’ — the economic realities no longer support this narrative. It was for this reason that Patty McCord created the Netflix culture deck with the following message front and centre:

We’re a team, not a family.

We’re like a pro sports team, not a kid’s recreational team.

Loyalty is good as a stabiliser … but unlimited loyalty to a shrinking firm, or to an ineffective employee, is not what we’re about.

McCord believed this formulation was an exercise in honesty. She thought it made explicit a reality that was implicit at many modern companies.

The tactical recommendation: I think it’s clear to most of us that company loyalty is a thing of the past. I expect to be laid-off at least once in my career; you should, too. Irwin suggests taking on the framing that Netflix uses: think of each job you take like you would an athlete signing on to a professional sports team. That relationship exists for a limited duration, which will end when the team’s needs and the employee’s abilities are no longer a match.

The tricky, practical question to all this is ‘what are the limits of loyalty?’ Irwin suggests hewing to the principles of honesty and reciprocity. If an employer is more invested and shows longer-term loyalty, reciprocate by giving really early notice, and go out of your way to train replacements. Irwin also suggests applying a higher standard for external opportunities, so you are more reluctant to leave higher-loyalty employers.

If you were hired for a specific project, however, it’s only fair to see that project to completion. Doing this is consistent with the transactional nature of the relationship with your employer.

To some degree, this is common sense. Most industries have dense networks, and reputation matters over the long term. The fact that employer-employee loyalty is dead doesn’t mean that you aren’t held to some common understanding around reciprocity.

Don’t create reputational harm for yourself by violating these new norms.

5. When is Freelancing Lucrative?

In 2017, Irwin wrote a hard-hitting piece in the New York Times on the societal impact of outsourcing. The piece, titled To Understand Rising Inequality, Consider the Janitors at Two Top Companies, Then and Now covered two janitors: Gail Evans at Eastman Kodak in the 1980s, and Marta Ramos at Apple today.

Ms Evans took advantage of Kodak’s internal training programs, and rose from janitor to CTO of Kodak over the course of a decade. Ms Ramos, on the other hand, was an employee of an outsourced cleaning company. She had no such opportunities.

The NYT piece makes a broader point about inequality, but Irwin’s purpose in referencing the piece in  How to Win is different — he argues that freelancing is a valid option for today’s ambitious worker … but only if the contract work isn’t regarded as a cost centre for the company in question.

The macro: Companies are more willing to outsource non-essential functions of the firm today. The traditional business narrative around this is that companies outsource if the task is a ‘cost centre’ — that is, an activity that isn’t a core competency of the business. Apple is willing to outsource janitorial services because it doesn’t gain a competitive advantage from cleaning its own offices, but it keeps design and engineering functions in-house because those are core competencies for the company.

In practice, however, Irwin observes that companies have become more willing to outsource semi-core competencies to contract workers. This violates the traditional business narrative, and creates lucrative opportunities for eagle-eyed workers. Irwin turns to academic Matthew Bidwell for an explanation of this effect:

• Contract work can be a form of ‘internal rules arbitrage’. Often, large companies have much stricter rules around hiring employees than hiring contract workers/vendors. This makes outsourcing a lot more attractive to address labour shortages, because it is far easier to bring on a new contract worker for a project as opposed to a new employee.
• Contracting outside firms is sometimes used to ‘force’ internal business units to be clearer about what they need. Because the organisation is paying for the service, business units can’t keep changing their minds the way they can if they were using internal development resources.
• Internal politics define much of the nuances over these contract worker arrangements. Bidwell explains that the goals of senior managers in a large corporation are usually in conflict with that of project managers — the former want lower costs, the latter want successful execution of a project. In projects where the work was seen as straightforward, senior management would often push to outsource the work — without regard for the communication and logistical costs to the project manager. On the flip side, in projects where the work was seen as more complex, project managers would often push for external guidance, with senior management balking at the cost.

Irwin also notes that firms at the cutting edge of some technical field often bring in specialised experts for specific, tricky technical problems. The intuition here is that it might be more worthwhile to use a short-term contract with an experienced CFO who is between jobs than a similar contract with a team of young MBA consultant types doing generic analysis for six weeks.

Irwin profiles two examples of this trend: the first, GE, uses an internal system called GeniusLink that matches managers with a global pool of freelance and contract talent. GeniusLink pays freelance engineers for solving specific technical problems; the example Irwin cites is a young Indonesian engineer named M Arie Kurniawan, who was paid $7000 for a method to reduce the weight of an engine bracket by 84% — but I’m not sure to what degree GeniusLink is real and useful, and not just corporate ‘we are innovative’ marketing fluff.

The second example Irwin uses is that of Michael Solomon, an agent for programmers.

When I met with Solomon in early 2018, for example, blockchain technology and cryptocurrencies were in high demand, so the small number of engineers who knew blockchain backward and forward could land lucrative contracts with any company considering an initial coin offering or other initiatives.

Solomon describes the thought process for a company thinking of hiring one of his clients versus a conventional employee: “With startups, they may think paying $200 an hour to a freelance developer seems really expensive, so they’re going to hire someone full-time, and they have a $200,000 a year budget to hire them, which works out to $100 an hour. But how long have you been looking? If it’s eight months, how much product have you not built in those eight months? And when you find that person for $200,000, how long will they stay? How much equity are you giving them? How much rent are you paying for their workspace? How does paid vacation factor in?”

I buy this calculus.

The big takeaway here is that contract work shouldn’t be seen as all negative. If the work fits into one of the situations Irwin lists above, taking on a freelance job might be a workable idea.

6. Using Big Data To Improve Performance

This is one of the weaker chapters in How to Win. For whatever reason, this chapter was also the bit of the book that was excerpted in the New York Times.

The macro: We’ve seen the rise of statistical techniques in baseball and basketball. Irwin points out that most of our work today is conducted digitally — leaving a ‘digital exhaust’ of documents, tasks, calendar events, and emails that may be mined for performance enhancing insights. He is hopeful that such mining might help us discover new ways to improve our performance at our jobs.

Tactical recommendations: I think the major takeaway here is to look out for such big data/people op insights, as they get released and/or covered by the popular press. Irwin presents a list of findings from Microsoft’s attempt to mine their productivity data:

• Managers should watch the hours they’re working. For every hour a manager spends doing email or meetings after normal working hours, it adds up to twenty minutes per direct report as well, and that cascades through the organisation.
• Do one-on-ones. Regular department-wide meetings do not correlate with managerial success. One-on-one meetings, on the other hand, do. (I’ve written about the mechanics of this effect elsewhere, and I’m not sure a data-driven approach is more convincing than what can be garnered through practice alone).
• Always work on your network within your organisation Managers with more extensive networks within the organisation tend to have subordinates that stick around longer. The root causes for this correlation aren’t clear.

This chapter is weak because what little insight it offers is blindingly obvious to a semi-competent practitioner.

The best data-driven insights are counter-intuitive — which is sort of the point, because if they were intuitive, we wouldn’t need statistical techniques to discover them. I’m not sure if this is my bias speaking: Irwin’s list of findings from Microsoft weren’t particularly interesting or insightful to me. I think Google’s discovery of psychological safety as an important team attribute was more life-changing — at the very least, it wasn’t intuitive to me as practitioner, and it made a big difference in how I thought about my day-to-day interactions with my team.

7. Management Ability as Competitive Advantage

Next up in ‘blindingly obvious findings’, Irwin suggests that good management is as much of a competitive advantage in business as having a technological edge. The smart worker picks good managers.

The macro: Irwin observes that the manager’s role is to increase the output of an organisation. (This is the central idea in Andy Grove’s High Output Management, from which much of Silicon Valley’s management style is descended from).

Irwin then observes that good management is significantly more powerful in the information age. In typical non-fiction format, he tells a number of stories to make his point (and pad the page count of this chapter), but I’ll just focus on two.

The first is about Susan Salgado at Union Square Hospitality Group. Union Square Hospitality is the company behind successful restaurants like Union Square Cafe, Gramercy Tavern, and Shake Shack. When Salgado entered the picture, however, USH had only a handful of restaurants, and Shake Shack’s enormous empire lay in the company’s future. USH founder Danny Meyer was struggling with scaling his operations; Salgado, then an MBA student at Lehigh University, offered to embed herself in USH’s operations in exchange for writing it up in her dissertation.

Salgado discovered that USH relied on what she called a ‘three-legged stool’ to thrive: they had a good hiring process, they had great systems to run the restaurants, and they maintained a good working environment. Remove one of the three ‘legs’, and USH’s operational advantage would go away.

Of the three, the third leg — the culture —  was the biggest obstacle when it came to scaling the company’s business. Meyer created a good working environment … but only when he was at that particular restaurant. Meyer recounts:

“Every time I went back to one restaurant, it had gotten off center. It was fine while I was there, and then when I would go to the other one, I would have to correct things that had gone wrong. The things that bothered me that slipped were how people were treating each other, the approach they were taking to guests.”

After her MBA, Salgado offered to work full time for Meyer at USH. Her solution to this problem was to formalise all of Danny’s practices as a series of management training classes. There, they simulated difficult situations that managers would have to deal with on a day-to-day basis: how to reprimand an employee who kept making mistakes, how to pre-emptively notice a customer growing frustrated, and how to mediate the inevitable disputes between front-of-house staff and back-of-house kitchen crew. Formalising Danny’s intuitive practices helped USH scale their operations to an empire of incredibly successful restaurants.

Irwin tells this story to illustrate how good management — as per USH’s ’three-legged stool’ approach — provides competitive advantages, even in a cut-throat business like the F&B industry. Could such an approach matter in other companies?

In 2003, John Van Reenen of LSE and John Dowdy of McKinsey had the idea of testing the influence of good management on overall business competitiveness. The mainstream opinion — then and now — was that Michael Porter-style big-picture strategy was more important than the boring nuts-and-bolts of management.

Van Reenen and Dowdy organised a study where they tracked the implementation of management best practices against the corporations’s overall financial results. Their hypothesis turned out to be correct:

Sure enough, the firms that rate highest in their management practices have higher return on capital and are less likely to go out of business. Management practices seem to be strongest in some of the countries known for high competitiveness and high incomes, like the United States, Germany, Japan, and Canada—and weaker in places known to struggle on both frontiers, like Italy, Portugal, India, and Brazil. Interestingly, some of the countries with lower average levels of management quality have much stronger results at multinational firms’ facilities within their borders, suggesting companies can export good management techniques.

Irwin concludes:

Good management acts like a technology in and of itself. It is something that can be adopted and propagated across companies much as a new technique for smelting steel or testing machine parts might be. When best management practices are in place, the productivity of people up and down the organization is higher; each individual becomes more valuable than they would be in a more poorly managed firm. Higher productivity is an essential precondition to higher compensation.

The tactical recommendation: Management is important. If you’re in an organisation with good middle managers, you’re likely to accrue benefits that others in less well-managed companies wouldn’t. Sadly, Irwin doesn’t give practical advice on detecting good management; he merely argues that working for a good manager is hugely important to one’s career.

In other news, the sun rises in the East. I found Salgado’s story the most compelling one in this chapter (I’m a huge fan of Shake Shack) … but I wish Irwin spent more time examining tricks we may use to figure out if an org is well-managed.

8. Navigating the Winner-Take-All World

Irwin asserts that many of the challenges we face today come as a result of a what he calls a ’winner-takes-all’ world.

The macro change here is that we’re seeing increasing consolidation across many, many industries. Irwin lists a number of factors that lead to the rise of super-corporations in many industries:

• Intangible inputs have become more predominant — Information industries like software and movies have always had a winner-take-all element, because they don’t experience the curse of diminishing returns (you can’t produce more widgets for free; but the cost of copying and distributing a movie or a piece of software is near zero). As ‘software eats the world’ , Irwin thinks we’re seeing these effects spread to other industries — even ones that aren’t historically associated with software.
• Network effects — Some products are more valuable the more people use them. Think Microsoft Word, or Facebook. Irwin asserts the ‘software eats the world’ thesis means more and more network effects popping up in our economy — in places that we don’t expect.
• Market power and contagious consolidation — When one industry consolidates, for whatever reason, there’s a greater incentive for those it does business with to consolidate as well, out of fear for being steamrollered by a more powerful customer or supplier.
• Crony capitalism and regulatory barriers — This is basically regulatory capture. Irwin believes we’ve seen the rise of both in America. And, finally:
• Antitrust authorities’s tolerance of consolidation — Irwin believes that the US Government has been historically very tolerant of consolidation. He points to Delta buying Northwest, United buying Continental, Facebook buying Instagram and Whatsapp, and so on.

We can quibble over the underlying factors, but I’m willing to accept Irwin’s assertion at face value. The more important question to ask is: what do you do in response to this widespread consolidation?

The tactical recommendation: Irwin argues that there are three types of firms in this age of consolidation. . Each of these types offer an ambitious worker different opportunities, and therefore different broad patterns for career advancement. The firm types are:

• Winner companies — these are companies that have won the consolidation wars.
• Ascendant companies — these are startups that are hoping to grow large enough to displace the winners.
• Afterthought companies — these are the losers in the consolidation war; legacy companies that are struggling to survive.

At a winner company (think: Microsoft circa 2001), you enjoy the safety of a healthy paycheck and the clarity of a linear career ladder. The risks here are that you become silo-ed in a narrow area, and you lose the motivation to push boundaries and take initiative; Irwin profiles a successful executive at Microsoft named Nick Caldwell, who had to be pushed by a mentor to take initiative once he reached a certain level in the Microsoft hierarchy. This push led to a sequence of events that eventually caused Caldwell to ascend to general manager; Caldwell left in 2016 to become VP of engineering at Reddit, eventually rising to become Chief Product Officer.

Caldwell, on careers:

“I’d run into two classes of people,” he said. “One would be the iterate-and-optimize kind of person. They’re fine working on Office 11, 12, 13, 14, and 15. That’s their bread and butter. But for people like me, it was more like, ‘How do you do something novel in a way that fits into this older product?’ And it’s a small number of people who are excited about that sort of thing, but if you can find them, and get those people together in the right way, that’s where all the true innovation comes from.” Even if one of your bets within such a company wins, you don’t own upside from it; you are trading the potential of a big win for the safety net of a steady salary whether your product makes it or not.

At an ascendant company (think: Zillow circa 2005), cash compensation is low, and stock compensation high risk.  The benefits of working at a startup are the lack of boundaries — and the opportunity of growing into a role that you wouldn’t otherwise be ready for.

Irwin profiles Amy Bohutinsky’s career at Zillow: she started out as a PR person for the 30-person startup, but was shoved into product discussions and marketing operations from the very beginning. Irwin writes:

The people who most thrive in startup environments are comfortable with ambiguity, with a lack of clear hierarchy or clear departmental lines. The payoff is that this very ambiguity allows you to get deeper, wider experience with crucial strategic questions than you would in a more established company. (…) “Being involved with product strategy is something I could only have experienced in a startup at the ideation stage,” said Bohutinsky. “We were just a group of people in a room without a lot of walls or lines, and that set me up for a future of realizing I had a voice in what we created that wouldn’t have been heard otherwise.”

In 2015, when Zillow acquired rival real estate site Trulia, Bohutinsky was appointed COO to oversee the merger — an urgent, bet-the-company operational imperative. Why her? When asked, Bohutinsky reflected: “I think it came from years of me showing that if you give me something I don’t know a lot about, I’ll figure out how to tackle it and how to build the teams to go do it, and figure out what I don’t know.”

At an afterthought company, the opportunities are to ascend at a company in crisis. Irwin profiles Mark Mason — a young accountant at Deloitte who stumbled into the role of corporate turnaround artist. His first stint was at a real-estate development company, who recruited him as CFO out of Deloitte … only to find itself in a real-estate-driven recession. Mason’s job was to engineer a turnaround.

Mason learnt a great deal in a short time about high-stakes negotiations and dealing with a company in crisis. His team pulled off the save — barely — and after the turnaround, Mason returned to the safe embrace of Deloitte “taking a step back to do that, both in compensation and stature.”

What his experience did give him, though, was a reputation of an accountant who had also led the restructuring of a trouble company. This reputation eventually led to an offer to be CFO of Fidelity Federal Bank, a 5 billion institution that was in financial trouble, and then eventually CEO of HomeStreet bank (who was also in crisis).

I think Irwin picks these three profiles for a reason: these are examples of three people who have successfully made the jump to the C-suite. The paths they took are different — Caldwell had to bide his time and climb the ladder before taking initiative; Bohutinsky found growth in a rocketship, and Mason advanced in his career by coming in to save distressed companies. As far as patterns go, the three types of opportunities do seem broadly correct — and a direct consequence of massive consolidation.

9. Responding to Software Eating the World

The final idea in Irwin’s book deals with one of my primary preoccupations: what do you do when software is eating the world?

Macro change: Here’s the crux of the argument. In 2011, Marc Andreesen wrote a piece in the Wall Street Journal titled ‘Why Software Is Eating The World’. The essay Andreesen wrote has been hugely influential; I urge you to read the full piece if you’re not convinced by it.

The gist of the argument is that, increasingly, almost every business today is in some aspects a software business. You must deal with this trend if you are to succeed.

The tactical recommendations: Irwin argues that every worker has two fundamental options when faced with software displacement:

1. You focus your career on non-automatable skills. Irwin believes that the best jobs here are the ones that involve improvising and reacting to a fluid environment; anything that is rote-driven is at risk of being automated away. (Ironically, this means that a gardener is more resilient to automation than an accounting clerk).
2. You reorient your career to help with the automation. That is, you become the automator, rather than the automatee. On this note, Irwin notes that you don’t have to be a software programmer. You merely have to be good at bridging domain experience with technological expertise — being a glue person, in other words.

Irwin profiles Jarlath Mellett — who, like Marco Relevant of Weta Digital at the beginning of this book — works with computer programmers without being technically proficient himself. Instead, Mallet has a background in traditional fashion. He designs clothes with pencil and paper, even as he works alongside programmers and materials engineers making cutting-edge clothing at Ministry of Supply. Mellett’s contribution? He’s ‘Pareto optimal’ — that is, he is able to speak the language of the engineers, and is able to communicate at the right level of abstraction.

Irwin reflects that Mellett resembles many of his (ridiculously successful) peers in the news business. Irwin is a journalist, not a fashion designer. But he notes that the best journos have been the ones who are quickest to adapt to the digital world we live in. This difference is attitudinal: like Mellett, they are willing to rethink the entire process of how work is done.

Irwin concludes:

In sum, as more industries become, at their core, software businesses, it doesn’t mean that everyone should be a technologist. It does mean that everyone needs to be focused on making the most of those technologies.

Conclusion

Neil Irwin has written a moderately good book on career planning in a world of consolidation and technological change. There are a few ideas I find valuable here: the most compelling one being the articulation of the ‘Pareto-optimal glue person’, and suggestions on how to pursue it.

If there’s any complaint I have with the book, it’s that Irwin hews too closely to the ‘story-lesson-story’ template of popular non-fiction. Irwin’s case studies are verbose, sprinkled with little details meant to make characters more interesting or the narratives more palatable; unfortunately, the idea payoff doesn’t often justify the storytelling. (I mean, ‘management is important; pick good mangers’ — really?)

I find some of Irwin’s ideas worth exploring further, and I intend to pursue them on this blog. But I think this book is only passably good — stick to my summary for his ideas, or buy an ebook copy, but just skim the stories if you do.

Three stars out of five.

Originally published 04 August 2019, last updated 22 July 2022.