Competing with Unicorns

× logo-full-size.png


Curriculum Vitae

Public Keys


Books I read

SHA 256 Calculator

PLs & Tools


RSS Feed




I keep a list of books I have read and what I thought about them for future reference later. These thoughts are only for me to remind myself of what I felt while reading it. I forgot so much about the other books I read, and I feel like this should help with that. I usually don't care if there are a bunch of errors in these notes. Therefore they are cheap to create for me. This piece will be a little different. I have so many thoughts about Competing with Unicorns by Jonathan Rasmusson that I felt the need to sort my thoughts more thoroughly. If you want the one-sentence summary: ''Spotify is a great company to work for, and its management style is superior to anything else.'' I'm not saying the statement is wrong, but it's so unreflected and uncritical. I will get to more details soon, but this is my first impression.

Specifically, I will touch on the following points (in no particular order):

1. Missing Critical Distance

I already summarized this point in the introduction. The book explains how Spotify works and organizes itself. The generalization to all Unicorns (or tech companies, Rasmusson uses them interchangeably p. xii) seems a bit eager. In principle, I don't have a problem with it, but this makes the whole thing more of a case study. I would bet there are as many differences between the tech companies he explicitly mentioned (Google, Apple, Amazon, Facebook, and Spotify) as there are similarities. How do we know that these identified similarities are where the competitive advantage lies? We don't, especially if the first-hand experience seems to rely on a single company from the list.

Having only one data point exaggerates the impression that Spotify's style is superior to all other management types. Even if all other tech companies manage the same way and this is a competitive advantage, I really can't trust this advice.

It also does not help that there are no counterpoints raised in the whole book. (At least I didn't see any. If there are any, I would like to know.) I won't count things like ''Some people don't like to work like this.''-like statements as counterpoints. That's an entirely pointless statement because

  1. I can say this about any way of work.
  2. ''Some'' is so unquantified that it doesn't help a discussion anyway.

The same is true for the reverse statement, by the way. If you say that most people enjoy working a certain way, I want to know:

  1. How do we know that?
  2. Why is that?

I don't want your opinions. I want tangible evidence. Given that these companies ''Learn with Data'' (p. 81), I expect to get some.

To get back to the lack of counterpoints. There are two explanations for their absence. Firstly, there might legitimately be none. I seriously doubt that. For example, do we know how this management style behaves in crisis? He doesn't mention that. We understand that authoritarian leadership outperforms transformative leadership in harsh economic conditions. (https://doi.org/10.5465/amd.2014.0132) I hypothesis that the direction displayed by tech companies is closer to a transformative style than to an authoritarian. So the question becomes, how do they respond to similar conditions. My gut feeling is that they behave even worse. This counterpoint I thought of on the spot. And to be clear, I might be wrong, and they outperform any other form of management. We don't have any data in either direction.

On the other hand, this would explain why older companies have a more traditional structure. They have been in enough crises that have a form to accommodate this. The only company on Rasmusson's list that was in trouble is Apple. But did Apple behave like a unicorn described in this book then? Given that Steve Jobs needed to come back and that he was not a consensus kind of guy, I would guess no.

I'm not trying to convince you that there are drawbacks. I do think there are because I haven't seen a perfect thing in my life. Additionally, I don't see evidence that there aren't any in organizations structured like Spotify. But I still don't know. I'm just saying it seems strange that he doesn't mention any.

2. Validity vs. Correctness

This book also mentions the significant advantage that is the ability to iterate over a product quickly. (p. 8) I agree with the point. I want to say that just because you iterate over a product doesn't mean it is okay to ship an incorrect one.

Validity means shipping the right product, so the product customers want. Correctness means it's working correctly. Most companies sacrifice correctness while testing for validity, and this is not an excuse. I like working products. And don't want a service/game/program that gets patched with an update because they are still searching for a better market fit.

There are also limits to what I want companies to test out in public validity tests. If you need product testers, then hire some. At least declare the thing a beta (or even alpha). Don't ship something as done when it is not. These are not complaints about the book perse. But I still wanted to mention them.

Also, I know this is a little out of place, but on page 80, he writes, ''They[Tech companies] don't cut corners on tech - they treat their infrastructure as a first-class citizen.''. Really? Do they? Every bug where something didn't work because they didn't have a test for it, was what? Someone thought they didn't need the test and cut a corner. What is the behemoth of the Google Suit Applications (Docs, Spreadsheet, Gmail, Drive)? Are you telling me it has to be this slow because they didn't cut corners when implementing it? I think I remember to have read that the whole thing was one enormous application, which is part of the reason it runs so slowly. I don't know where so this might be wrong. But saying they don't cut corners is insane. Engineering is an exercise in corner-cutting. We only call them trade-offs, but essentially there is no real difference.

3. How to Handle Data

An adjacent topic is how these companies handle data. Rasmusson explains that these unicorns collect a lot of data to develop the perfect product. I believe this is true. But yet again, it was utterly unreflective. Maybe it is not a morally right thing to collect all this information about people. I know it gives you significant advantages when you collect all this data. I honestly didn't expect him to mention that you should still value and enable privacy wherever possible because it is your moral responsibility. There is also the real responsibility that comes with handling other people's data. There should have been a mention that you should protect the data against misuse. We have had enough scandals where user data leaked that this should have been at least a footnote.

4. Conforming to Company Culture

Chapter nine, ''Reinforce Through Culture'' (p. 89), Rasmusson describes how different companies value different cultures, and that tech companies reinforce behavior through their culture. I want to remark that all these lists of qualities people value in their employees always has asterisk'. For example, Amazon(you can find the list at www.amazon.jobs). One item on the list is ''Insist on the Highest Standards''. This statement would imply that everyone working for Amazon insists on the highest standards. Not true. Management decided that it is enough to have contractors deliver their packages because it allows them to save money for various reasons. One of which is they don't have to pay the delivery men as much. They spy on their employees in Amazon's package centers. What highest moral standard is that? What they mean is ''insist on highest product quality''. We can debate this because they allow people to sell a lot of garbage on their platform. But at least it is closer because they don't care about humans working there. They care about the product. What I'm getting at is that these we need to interpret these lists correctly.

They are companies that try to create value, first and foremost. For the longest time, Google had a motto not to be evil (''Don't be evil''). They still did some things (like collecting all the data) that were at least questionable. These things always have to benefit the company, and they will interpret these things loosely when they need to.

5. Formating

While the last points were more book adjacent, this is really about the book. These range from nitpicks to things that I would not have accepted for print. For all of you have to keep in mind that this book is short. It's 114 pages. You can easily read it in a day. And somethings you get away with in long books become irritating. For example, my general stand on repetition is that it is good. It allows the reader to follow the discussion more easily. It gets problematic when you repeat yourself almost verbatim. He talks about Data Scientists twice, once on page 30 and once on page 86. Both sections have the same length, the exact contents, and sometimes the same sentences. They are both a page, and I often skip through the book before I start reading, and when I read page 86, I was first confused because I thought I might have read it when cutting through. Well, I didn't is was just repetitive. I could excuse this if it were a challenging topic where you need to repeat ourselves more oft not to lose the reader. But everything discussed in this book is straight forward, and the two sections are 50 pages apart. That's half the book, but still very close. If something is just a repetition of something said shortly before, leave it out.

On the topic of repetition, the book has a lot of pictures. And they also repeat many times. They could have just put less of them in there and reference them. I can turn a page back if I want to look at them again.

I think you could condense the whole thing down to maybe 40 pages without losing anything. There are a few places where you can write something down. We would need to leave them out or collect them in the book's back.

The book would also appreciate some more formatting. A sentence is not a paragraph, and it doesn't look good on a page, especially if it is just one line. (p. 54) Sentences introducing an enumeration should be on the same page as the enumeration. (p. 75) If you refer to a video, provide a link, don't tell me to ''Check them out by Googling >>Spotify Engineering Culture<<'' (p. 24). Later, he provided links. Why he didn't here is beyond me. Also, there is more than one kind of bullet list (p. 92 and 93). I haven't figured out if there was a reason. I just noticed, and it irritated me.

6. Conclusion

I have criticized this book a lot. It is not that bad. It is easy to read, and if all you are looking for is anecdotal evidence that Unicorns have a superior management style, this is the book for you. If you are looking for a more balanced comparison, this is not it. There are also some errors I haven't mentioned: typos and alike. I don't regret having read this book, but I doubt it will become a classic for management to read.