A good bullet point for your bucket list is to write something that will change a lot of lives. There are a couple ways you could theoretically pull this off: you could come up with an amazing new idea, or at least a compelling sales pitch for an old one. But a good way to get a tactical advantage is to choose your audience well: write for a reader who is on the cusp of greatness, and nudge them in the right direction. (Other people play this game: one reason Steve Jobs’ Stanford commencement address inspired so many people to build companies is that it’s a good speech; another reason is that the initial audience was a bunch of people who had just graduated from Stanford.)

Reid Hoffman has achieved this with Blitzscaling. If it sells reasonably well and makes just a handful of people change their minds and wind up building a massive company specifically because they read this book. Normally, VCs try to catalyze the creation of generational wealth by investing in high-performing companies — if he’s willing to accelerate a company’s growth in exchange for $18 per copy rather than points of equity, it’s a huge wealth transfer.

But let’s back up. What is “blitzscaling”? Hoffman defines it as when companies “prioritize speed over efficiency in an environment of uncertainty.” And this is important. When we look at large, recognizable companies, we tend to see them doing roughly the same thing, but a little bit bigger and a little bit better each year. Sell 2% more widgets for 3% more money, lower your cost-of-widgets-sold by a couple tens of basis points a year, buy back some stock, and you, too, can achieve above-average compounded growth rates for centuries.

Exciting to some people, boring to many more; this is a book about how a company goes from an uncertain prospect to an unshakable institution.

“Blitzscaling” sounds temptingly similar to a buzzword. You can imagine an MBA using it in a normal conversation. But it’s not. The thing about buzzwords is that they try to draw a sharp dividing line in what’s really a continuum: empowerment, decentralization, holocracy — these are all synonyms, but they were pitched as big new ideas in the 90s, 2000s, and 2010s, respectively. They all refer to a bias towards letting employees make decisions rather than wait for managers to give them orders, but that’s something that has to happen on a continuum. “Holocracy” means your marketing person unilaterally decides to change some ad copy, not that your summer intern negotiates an acquisition.

Blitzscaling is more like a phase change — a higher-energy state where the old rules don’t apply. There are particular circumstances where it works, and in those circumstances it’s necessary. That’s the core insight from the book: when you reach a point where your company can scale fast, you’ve also reached a point where a competitor could potentially do it, too, and if they outraise you and outspend you, they’ll win even if their product is worse.

Hoffman draws a useful 2x2 matrix of different kinds of scaling: slow/fast and certain/uncertain. Slow scaling is easy to understand and easy to model, but the difference between “fastscaling” and “blitzscaling” is worth dwelling on. “Fastscaling” is his term for growth in the face of certainty. The best prototypical example is probably franchise businesses in the last half of the twentieth century. Once the US had interstate highways, credit cards, and broadcast TV, you could be reasonably sure that any business that worked well in one city would work well in another. Whether you’re selling burgers, coffee, Scandinavian furniture, or clothes, once your model works in Tucson it’ll work in Tallahassee and Trenton, too.[1] But if your business appeals to 0.01% of the population of global Internet users, there’s no good way to know whether you’ve found a set of early adopters or merely a nice niche. No way, that is, other than to spend money on marketing and closely watch your incremental ROI.

What makes a company blitzscalable? The main thing Hoffman cites is network effects, especially two-sided network effects. Once you know a product gets more valuable as more people use it, the sole victory condition is ubiquity. For a two-sided network, this effect is even stronger: if your business connects employers with job-seekers, or helps people get dates, or connects suppliers with buyers, your economics with respect to one side are better to the extent that you have bigger scale on the other side, so the winners win by an even bigger margin. There are other categories of companies that can blitzscale — I’m personally fond of the set of companies that define new property rights.

But network effects are probably the best explanation: any sufficiently transformative business will reshape whatever’s adjacent to it in the supply chain, which is a de facto network effect even if it’s not thought of as such. Did Carnegie Steel have network effects? Steel led to a massive boost to industrial productivity generally, more or less allowed the US to build out its railroad system — which, for freight at least, remains the best in the world — and eventually revolutionized construction. The capacity and economies of scale from the US railroad buildout created enough cheap steel that it presented a de facto subsidy to anyone who wanted to build structures that didn’t fall apart. This stimulated new sources of demand, so by the time the railroad buildout was done, demand for steel in construction and machinery could absorb the supply. Oil in the Rockefeller era is similar: sufficiently abundant and ubiquitous oil made the automobile possible, and the possibility of cars led to a large investment in road infrastructure in the 20s and 30s, and then again in the 50s. Carnegie and Rockefeller aren’t considered case studies in network effects, but if you squint you can see it.

The Case Studies

It’s brave to write a book about technology companies while citing contemporary examples. I mean, it’s brave to write any business book doing this — ever look at the average returns from investing in the Good to Great portfolio? (Not so great). But given the instability of the tech sector, and the power-law distribution of returns, it’s entirely likely that some of the companies cited as good examples will flame out spectacularly within a year or two. (It’s also possible that others will do extraordinarily well — the book is a diversified portfolio of individually bold bets.)

A couple of the older war stories really stood out to me:

• Airbnb got cloned in Europe, and the clone hired a lot of salespeople and got a quick lead. They could have given up on Europe, but if you’re in the leisure travel industry and you’ve lost Europe, you’ve lost, period. So they acquired a smaller competitor to the clone, and scaled up even faster. Airbnb is a particularly interesting case as a two-sided network, because in theory it can evolve into a sort of 1.5-sided network: every night stayed by an Airbnb user is an empty bed at that user’s home; if you’re both a guest and a host you can travel for free.
• LinkedIn competed head-to-head with Plaxo in business networking. What LinkedIn figured out was that Plaxo’s product (the address book) was really a feature; the product was a profile.
• Hoffman cites a fellow VC who used to run growth marketing for Twitter, who realized that the main determinant of user loyalty was frequent logins (the cutoff: usage for seven days in one month) and the main determinant of that was following at least thirty accounts. What’s exciting about this is that Chamath Palihapitiya had exactly the same answer, with slight differences in the values of each variable.
• Ride-sharing companies subsidize both sides in a new market, because the flip-side of increasing returns to scale is really poor returns at the beginning. This is just math, but sometimes hard to internalize. If you look at a timeline of Airbnb, it’s incredible how long they kept at it before their product achieved any kind of sustained growth.

Hard Questions

The difference between a good book and a life-changing book is that a life-changing book doesn’t just tell you what you could do to improve; it tells you you’re doing something wrong, right now. It puts a pebble in your shoe and irritates you to no end. Starting Strength reminds you that you’re getting weaker every day unless you’re either doing squats or you’re too sore to walk. Twelve Rules for Life tells you that you’re fighting a war with entropy, and entropy advances whenever you stand still. Dawn to Decadence shows you that you have to really, really step up you’re reading if you hope to be that erudite by the time you’re 93.

And Blitzscaling tells you that if you’ve achieved product-market fit, and you’re growing — you’re in a fight to the death with all of your competitors. If your industry is growing fast and you’re growing almost-as-fast, congrats: you’ve peaked. You’ll be less relevant in a year, and even less relevant after that. Anyone who is interested in the industry and ambitious about changing it will pass on working with you, and your best options are either a) you and your faster-growing competitors are wrong about the business, and it supports several winners (with crappier economics), or b) you’re a strategic acquisition for someone one step up or down the supply chain from your competition. In either case, you’re letting your competitors take the lead in defining your business: if there are problems with scaling, they’ll naturally run into them first; if not, the threat they present determines who wants to buy you.[2]

That’s hard to swallow. Sometimes you’ll read about a hot sector that comes out of nowhere, with three or four companies all competing for the big prize. They all seem to reach Series A scale in the space of a few months, and sometimes you’ll see round sizes like this: $3m, $5m, $20m. Guess who wins? It doesn’t really matter which company had a better product at that point; network effects dominate, and it’s really hard to raise a round if your pitch is “It’s been a whole quarter since we last raised, and now our business faces a new existential threat.”

But it’s true, and the sooner you recognize that, the better you’re equipped to respond.

I recommend Blitzscaling; I’ve omitted some of the good points from this review. (If the distinction between ethical pirates and sociopaths saves you from one bad hire, that’s well worth the price of the book; the line about vanity metrics is also an instant antidote to a bad shibboleth.) And while there are some parts that resonated with me, there were some misses as well: The chapter on social responsibility seemed both too expansive about companies’ obligations and too blasé about the apps-and-kids problem — yes, previous generations complained about MTV and books, but the addictiveness feedback loop is orders of magnitude faster now.

I suspect that different parts of the book will appeal to different people. I’m a vigorously-exploit-economies-of-scale-with-high-gross-margins kinda guy, so that’s what really resonated with me, but there are other approaches, too. Blitzscaling is required reading if you remotely suspect that you’re on to something big.