Executive Summary: This is a book about Ben Horowitz's war stories. Ben Horowitz has good war stories, if you care about the narrow space of Venture Backed fast growth technology startups. I'm not so sure that they generalize to the point of making a good management guide. You might be better off reading some Drucker.
First: the absolute preliminaries: Ben Horowitz co-founded LoudCloud with Marc Andreessen in 1999, with a plan to do enterprise managed services (what has now grown to the SaaS and IaaS space). They were technological pioneers in that space. They weathered the dot com bubble burst, but it was touch and go (since most of their clients were other tech firms going bankrupt). They IPOd in a post-bubble environment with a declining market. They pivoted to become Opsware, a tech services firm (i.e. further up in the backend of the managed services stack). They pivoted as a public corporation close to bankruptcy which is a nightmare I wouldn't wish on my worst enemies. They weathered NASDAQ delisting threats, mass employee revolts, ..., the whole Mad Max scenario. So he offers the prospect of excellent data from the inside. Ben also has a reputation in the valley as a no-bullshit guy, so this is valuable
Second: some initial thoughts. To the extent that Ben's stories will generalize, they only apply to fast growth venture backed tech startups that plan to IPO. It bears repeating that this is but one subset of the technology space - the subset that gets a disproportionately large amount of press coverage and fanfare. So much of the lessons are the kind that you've probably already osmosed from the Paul Graham/Peter Thiel/A360Z/AVC crowd. And if you haven't, then you should probably start with Paul Graham and Peter Thiel rather than jumping into this book, which has a narrower scope and doesn't make much effort to make clear the context. That's okay, since it's aimed at the specific crowd, but just be aware if you're not of that crowd and thinking of reading this book.
Ben tells the story from the technical CEO perspective, which is a fairly rare perspective to get, given the depth of Ben's experience: Note that Paul Graham only walked the walk through a ~50 million dollar acquisition, and never faced the challenges of being the CEO of a publicly traded corporation. Peter Thiel did, but is ultimately a non-technical founder. Besides, the PayPal mafia is fairly cagey about the war stories around PayPal given the extent of the internal fights (e.g. the ouster of Elon Musk from CEO) that they want to cover up. Much of the other available writings are from the V.C. perspective, and lack founder experience.
Thus, this book does give us an opportunity to read the thoughts of a founder CEO all the way through and beyond the IPO stage. But it is by no means a masterpiece that stands alone. It does offer an incremental view if you are primarily interested in the narrow niche of venture backed startups, and serves as a 4 star or 5 star book of anecdotes that have potentially generalizable lessons, if read critically, while fully aware of the appropriate broader context. That's a big asterisk though. So I will attempt to summarize the core lessons that I personally got out of the books, given that I am not solely interested in this space:
1. Being CEO in a fast growth venture startup requires learning at an incredible pace. The job description changes almost completely within 6 months given the growth rate. There isn't any way to get that experience except networking with founders who have previous lived experience as founder CEOs of large companies. This is why, for example, Zuckerberg had long walks with Jobs. A top notch VC firm will definitely help you get that, which is why they are so valuable. The rotating board circuit in the top Silicon Valley firms will provide this, if you can get them to sit on your board (Horowitz, for example, holds Bill Campbell's advice in particularly high esteem). You probably can't, unless Sequioa, Founders Fund, Andreessen-Horowitz, or KPCB is backing you.
2. At the end of the day, you have to make incredibly complex strategic decisions with extremely limited information, which creates crazy stressful environments.
3. To Ben's credit, he does focus on good management, but mostly doesn't add anything new to the corpus. It's mostly the usual good stuff: don't do arbitrary things, don't publicly shame people, don't let people get away with enriching themselves at the company's expense through strong arming, don't do short term things that jeopardize the long term, and if you do, it will rapidly become the culture and will remain permanently broken. Again, a decent coverage of the standard MBA literature covers this space better.
4. Being CEO is a lonely job. You cannot expect your executive team to understand all your decisions given their limited scope. You also cannot spend the time to give them the broader scope all the time, since that takes away from doing their own jobs. But if you don't, they might stop trusting you. It's Catch-22s all the way down.
5. In addition, while the CEO mentor network can help with your emotions, they can't help you with the decisions, since it will take them weeks to get up to speed on the context. You're really on your own on these decisions.
6. It's also emotionally harder when your final decision is 55/45, and you're facing opposition from underlings/your board that is 90/10 the other way, but lack the appropriate context. So you're on the fence leaning one way, but others are convinced that you're wrong. At this point, you should just learn to trust yourself, since you're the only one with the appropriate context. Your board should be more experienced than you, so they should be aware of this information asymmetry. Consequentially, they'll usually just rubberstamp what you decide anyway. This doesn't make the decisions any easier to make.
7. You absolutely need to develop decision making processes that are insulated from your own emotional rollercoaster/insecurity. Ben likes writing things down until the words are convincing on their own. I do too. Whatever your methods are, you will absolutely need to rely on them since you're going to be alone in the decision making processes, and need some check to make sure you're operating logically with no emotional biases. So develop those early. (I personally recommend the LessWrong corpus, but you might find that too kooky. Your choice.)
8. Ben Horowitz really likes rap and hiphop. That's not my thing, but whatever, I can respect his artistic side. Anyone who criticizes him for that aspect of his personality is really missing the point.
Context that Ben Horowitz did not mention that I think would have greatly improved the book:
1. The tides are turning, and the days when Venture Capitalists could oust founder CEOs and replace them with "adults" are permanently gone. Zuckerberg, and the fact that founders (like Horowitz and Thiel) now lead top Valley VC firms has permanently changed the landscape. So a lot of the fighting that Ben had to do, and the perpetual underconfidence that he did it with, won't apply to future founder CEOs. This book adds to the growing corpus that helps make your case to not be ousted, which partially obsoletes that aspect of lessons in the book. Ironic, I know.
2. Fast growth tech startups that have a distinctive tech advantage can win despite terribly dysfunctional management. When you're in a greenfield technology space, you can do whatever you want, and still win just from the tech advantage. For example, Valve gets cited a lot for it's "no management" structure, but they sit on a giant money spigot called Steam that just prints money. If I also owned such an oil derrick, I could probably get away with mandatory sex parties every Thursday. This doesn't make the case for workplace orgies.
3. There is great talk about how Ben rescued OpsWare from near disaster through their final acquisition by HP for 1.65 billion dollars. But there is not a single mention of how OpsWare fared post-acquisition. For all I know (and given the next point, I suspect) this was a shitty deal for HP.
4. It doesn't help that Ben Horowitz heavily raided the OpsWare executive team to form the core team at his next venture (the VC firm Andreessen-Horowitz). That can't have been good for OpsWare at all.
5. Ben talks about how you really need your executive team to hit the ground running. You can cover for one executive to get them up to speed, but you can't cover for every executive. Thus, you need to hire top tier talent, often from outside. This is great. But this also forms a good explanation for why top executives often seem to be from a different planet, move in completely different circles, and get paid orders of magnitude more. But Ben doesn't cover any of these pathologies, or how it really is a necessary side-effect of the realities of the executive responsibilities.
6. From Ben's raiding of his previous form for top executive talent, I get an increasing suspicion that the top talent moves in tight-knit teams that gel well together. Your management priority should be to network and form these connections from an early stage (definitely before you found your own VC-backed fast growth startup). If you don't have this focus, you have to develop it at a late stage in the game, and for that you _really_ need top VCs, otherwise you will not be plugged in at all and have no access to these folk.