The Lean Startup
A Flyby Book Club Review
This is the review for The Lean Startup by Eric Ries.
Apologies for the slight delay. I've been slowly recovering from some kind of weird space virus and had a hard time getting through the book and writing the review. I didn't want to rush it out and skimp on quality, so hence the delay.
The post is about a 15-minute shot of solid insights. There are multiple sections, so you can read them one at a time or all together. Whatever you prefer. Hopefully, it's all straightforward in my current state of fogginess and doesn't come across as an incoherent fuzzy lumpkins… gulp!
Let me know what you think using the comments below.
Speaking of length, what is the ideal length of a post for you?
There are two ways to look at the process of entrepreneurship and building your product.
As if you were building a car or a rocket.
A car has two feedback loops. The internal engine, a complex piece of machinery, must be built and perfected before the car can even start. Then, there's the other feedback loop. The one between the steering wheel and the driver. Without this second feedback loop, nobody would be there to start the engine or drive the car (let's ignore self-driving cars for now).
A rocket, on the other hand, is self-guided. It has complex machinery in the engine but also steering systems to guide it to its destination, likely some planet in outer space. A tiny miscalculation can cause it to go thousands of miles off-course and miss the destination completely.
The book's central thesis is that people treat entrepreneurship as if they are building a self-guided rocket with no one on the other end. Whereas it needs to be treated as a car, where half the job is on the product side, and the other half needs to be honed alongside the customer.
I thought this was an insane insight. The message here is that you don't need to completely perfect and finish every detail of your product before shipping it. This can take years, and you might even end up with something nobody wants (more on this in a second). It's OK to start scaling the product once the core functionality is solid (i.e., the main engine) and then start rapidly iterating from there alongside customers (i.e., user feedback). The way I think of it is treating your product as a T.V. series rather than a feature film. You spend a bunch of time perfecting the pilot and then launch it. Once it's out, you see what resonated with the audience and iterate on the next few episodes from there.
But.. speaking of a feature film, what if you are building something harder to ship in parts, and the entire product is supposed to be consumed in one sitting? E.g., A video game, novel, or precisely a feature film/movie. I would say it's still possible to pull this off. In that case, it's about running the lean methodology with your marketing strategy. That is, extract the parts of the product that can be released earlier as promotional material and then keep chugging along on the core product. For example, suppose you are building a video game. In that case, you can post teasers and behind-the-scenes, making-of videos on social media to get people more interested in your journey. This way, you build fans long before your product ships and will likely have a more successful launch. Another option would be to start a development blog to share your wisdom. I’m kind of doing that here but on the entrepreneurship side.
If you can ship in parts, however, the core idea here is the MVP or minimal viable product. There's also a comparison to film production here, where movies have a pre-production stage. For example, in a Pixar Movie, which is 3D rendered, the final render takes a lot of time, including sound, animations, rendering, voice acting, editing, etc. So, the studio goes through pre-production, where the story and characters are fleshed out with hand-drawn storyboards and temp voiceovers. This allows them to make most of their “mistakes” earlier in pre-production before they get to the more time-consuming production stages. Big video game productions work like this, too.
An MVP for a software product is similar. It's the initial set of all the core features but lacks the sophistication and polish of the final version. This is then tested with live users, and feedback is incorporated as the product is iterated upon and improved.
In short, it's the practice of Thinking Big but starting small and filling in the gaps as you progress alongside your users. It's building the engine and working with the driver to perfect the car.
The book's author was working on a startup that combined 3D Avatar and virtual world technology from video games with the spicy social effect of IM chats. This is exactly what we learned in the last book club, Blue Ocean Strategy. You take two completely different ideas from different industries and combine them to form a new category. This book discusses taking that strategy to the market and what happened.
They decided to treat the IM integration as a separate add-on so users could integrate it with their own IM client. Based on their research, this was the more sensible thing to do. Their team spent months building out the MVP and polishing the app's first version. When they launched it, there were crickets. Turns out, no one was interested.
They brought in focus groups to see what was going on. They thought the avatars were fun, but none of the testers were interested in connecting the app to their own IM client, as that would be weird, and they didn't perceive the app as 'being cool.' They were happy to play with random strangers, however. This is a lot like how modern multiplayer games work today, but the team was running under the assumption that no one would want to download another IM client. They thought most people would tolerate a max of 1 or 2, but their target market, mainly teenagers, all had 7-8 IM clients. So, much of the dev time and code for the IM stuff was wasted. This is what the book is trying to avoid: waste. In fact, this whole methodology is based on the concepts of lean manufacturing, which is to identify and eliminate waste and operate more efficiently.
While this all makes sense, and I was on board so far, I didn’t quite agree with the author's other core premise that, as entrepreneurs, "we don't know what the customers want."
Let's zoom out and try and analyze what I think happened here.
As we've seen in previous posts many times, there are two ways to go about taking your product to market. One is serving the market at scale, and the other is scratching your own itch. The author here is talking about the first path, where you are separate from the customer and are building something with the market in mind. In this case, you are indeed running on assumptions and can run into a case where the audience simply doesn't care about your product because they have a different idea about how things should be. This is what happened in the book, but it’s also not the only way to do things.
The approach I use is scratching my own itch. That means I'm primarily building a product for myself first and then sharing it with the world by taking it to market. In this case, you know exactly what the customer wants because you are the customer. External input is still valued and taken into consideration, but the vision is firm and doesn’t need to be market-tested. This only works if you genuinely want to use the product and have built it for yourself out of some necessity, not if you are trying to project yourself into the market mind, which can lead you to have flawed assumptions about what the user wants.
An example of a chat app similar to the one discussed in this book, which was born from a scratch-your-own-itch mindset, was Slack. The team was also initially building a product for a target market and decided to develop their own messaging system on the side to communicate effectively across regions. In this secondary case, they were the customers themselves and could fine-tune the chat app to their needs and knew exactly how to shape it. Their core product developed for the market (a video game) flopped. So, instead, they took the tool they made for themselves to market and turned it into a billion-dollar success. This is the power of building something you actually want to use yourself. We also discussed these two approaches in Book Club #2 - How to Build Your Calling.
So, in the case of building a product for external customers, a lot of this methodology of rapid testing and iteration is quite useful, but most of it can be skipped if you are scratching your own itch, as you'll naturally end up doing it without much thought. Using the car analogy, you can think of it as building yourself a car where you are also the driver. That's scratching your own itch. You are so interested in driving this car that you end up creating the best damn car possible. You just have to switch gears every now and then from car engineer to car driver. This is the craftsmen's or the artist's way of doing things. Of course, others would also likely find your creation valuable once you take it to market. It's just a matter of finding them. And if you find other passionate drivers, you can take their input aboard if it aligns with your vision.
Versus, if you were building a car for the audience and were lukewarm about driving it yourself, then you are trying to serve the market at scale, and the lean startup methodology would come in handy. This would be akin to the scientific approach of testing, experimenting, and seeing what sticks and what doesn't. In this case, the methods in the book are like giving you a powerful flashlight in the dark so you know where to look. But the way I see it, scratching your own itch is like turning the lights on. So, if I had to choose, I would say building something for yourself first makes much more sense. Maybe this is easier to pull off if you are flying solo; I'm not sure. But to me, it's the only way to fly.
A lot of the book goes into great detail about how to approach this rapid testing and validated learning methodology. I skimmed over a lot of it as it's not how I'm currently doing things. But if you are trying to build something for external customers who are not you, check it out. There’s also a section on when to pivot and when to persevere; have a look if you are interested.
The gist is testing out what works and what doesn't as quickly as possible with your target customers. If you've ever run digital ads, you'll be familiar with A/B testing different ad sets and getting the performance metrics for each set. Then you can decide which sets works better and put more money into them. This is basically the same concept but on the product side. You A/B test different features with customers and deploy several iterations to learn what works as fast as possible. Under this methodology, the main goal is validated learning.
The key to validated learning is measuring using actionable metrics instead of vanity metrics. Vanity metrics are those that look good on the surface but don't really add much to your core business and bottom line. Actionable metrics are the ones that actually move the needle and can improve your business substantially.
Let's take an example of a newsletter. A vanity metric would be the Subscriber Count. If one newsletter boasts 10,000 subscribers, you would assume it's more valuable than another with 2,000, right? Not necessarily. The actionable metric here is Open Rates. That is, how many people actually read the emails.
If newsletter A has 10,000 subs but an open rate of 9%. That's effectively 900 people who are reading the newsletter. Versus, if the one with 2,000% has an open rate of 50%, then it has effectively 1000 people reading the newsletter. Newsletter B is more valuable than A.
Subscriber counts are the vanity metric, and open rates are the actionable metrics. Further actionable metrics would be read-through/completion rates, forwards, or the number of other publications referencing your articles. This would be a better target to optimize for. It's how you build a sustainable business rather than a good-looking shell.
In this case, some ways to do that would be to improve the content, make the newsletter more interesting, and send a "re-engage" email every so often. Another option would be to prune out people who haven't been reading the emails for a long time. While this would hurt the vanity metric of subscriber count, it would increase the open rates and lead to a more engaged audience. This, in turn, would generate more revenue per customer if you chose to monetize the product at some point.
It's the same for an app, with total downloads being the vanity metric compared to active users being actionable. Or if you have a revenue-generating product, then revenue per customer or paid subscription rates would also be actionable.
Can you think of some vanity vs actionable metrics for your own product? How can you move from measuring the former to optimizing for the latter?
As a side note, it was challenging for me to get through many of these sections, which discussed ideas about not being the customer yourself. Maybe it was just because of the space flu, but I was tempted to just drop this book down and put it away permanently.
This is what I used to do before starting the book club. If a book didn't grip me at the start, I would move on to the next. Many people who read a lot of books recommend this approach. Turns out this would have been a huge mistake. Since I had to write this post, I powered through and finished it, and I'm super happy I did. It had some solid insights throughout that even apply to my own stuff. Also, I would have completely missed the growth section, which is up next, had I done this. This was an awesome chapter, in my opinion, and opened up some very fruitful areas of research for me.
Lesson learned. The strategy I'm applying now is to skim the boring chapters and always finish a book if I pick it up. I've found there's always something to learn, even in a book you didn't like at the start or in areas which might be boring.
Let's move on to my favorite section of the book. Growth!
Or, more accurately. Sustainable Growth. The book defines this as when:
"New customers come from the actions of past customers."
For this, the book provides the three engines of growth.
These are sticky, viral, and paid. Each of these can be considered independent engines that can be set in motion to drive sustainable product growth. The term outside of this book for this concept is Flywheel.
The first, and in my opinion, the most valuable, is the sticky engine. This is where you focus on adding long-term value and user retention. Keeping users coming back by continuously improving the product or just making something so damn good that users can't picture their life without it. This includes products that have a steep learning curve and hence make it hard to switch from (e.g., 3D art packages, databases, etc.) and also stuff where there is user-generated content and sharing involved. Most subscription services would also fall into this category. This is why subscribers continue to pay and subscribe. It requires building a product with purpose and attention to detail. This is also what sparks word-of-mouth growth. We covered this in more detail in Book Club #1 - How to Be Remarkable.
The data to measure here is churn rate vs growth rate. You want the churn rate (number of users leaving) to be less than the growth rate (number of users coming in). If there is more growth than churn, then this would signal positive compounding over time.
The second engine is the viral engine. That is, generating network effects or products that skyrocket on their own. This is different from word of mouth. Although that can definitely cause a product to go viral, the book is talking about engineered virality. Specifically, making it so the product can't help to spread itself just as a side effect of someone using it. The example in the book is Hotmail. When it first came out, growth was slow. Soon after, the team added a short and subtle message with a link at the end of each email sent using the service: "P.S. Get your free email at Hotmail." This started the fireworks, and within 6 months, they gained a million new users. Eighteen months later, with 12 million subscribers, they sold to Microsoft for $400 million. I still have a Hotmail address myself. This shows you how sticky they were as well.
Other examples not in the book are logos, which make it obvious what the brand is, e.g., Starbucks coffee cups, McDonald's or designer shopping bags, and the TikTok logo on every video generated and shared from the platform. Again, this only works if the product is sticky and "cool." Otherwise, no one is going to care what the logo represents. The name of the game is "being cool." We'll look into this in a future post. But as a side note, if you are scratching your own itch, you already know what’s cool because it's cool to you, and you are the customer.
At the heart of viral growth is the viral coefficient (K). This is a number that represents the virality of the product. Here is a reconstructed chart from the book showing this. To learn how to generate a chart like this, see my guide to data analysis using chat-gpt. This saved me a good hour or so of scratching my head with math equations and doing some python gymnastics.
As you can see in the graph, if K < 1, the product will eventually fall flat and stop growing. K = 1 would be linear growth, and anything above 1 would indicate exponential growth. Every new user who signs up will bring in a friend or more.
The final engine of growth is the paid engine. This one is quite simple: generating new customers by spending money on advertising. This is also what most new brands default to. For this to be sustainable, the ad spend must come from the revenue. This would mean existing customers are funding future growth. Otherwise, the growth isn't sustainable, and the business will eventually run out of money. For this to be the case, you want to combine this with a sticky product so that the advertising spend brings in long-term customers. This can be tracked using a customer's LTV (lifetime value). If this number exceeds the ad spend, the book suggests that the product will grow. For example, if the LTV is $10 and it costs $5 to acquire a new customer, then paid advertising is a valid approach. I'd also add that you’ll need to create captivating ads with great stories and copy. Consumers are generally quite sick of ads, so only high-quality ones stand a chance of capturing attention.
The book suggests focusing on one engine at a time. Otherwise, it will be hard to measure. Once you've mastered one engine, you can move on to another. The way to optimize and fine-tune this is, of course, by measuring this data. This would depend on the product and category in question and which analytics you have access to. From those, you can likely create your own tables to measure this stuff or find a service that provides this.
Regarding which one to focus on, I'd say the most important are stickiness and word of mouth. Then, once you know you have a good product, add some hints of virality. Finally, experimenting with paid ads. You can also use paid ads or other techniques to attract initial customers. You will need *some* kind of initial growth to kickstart the network effects, after all. But be careful about relying on this as an engine before the revenue per customer is more than the cost of advertising, as the business can eventually run out of money.
The book doesn't mention this, but organically growing your brand on social media can be a great free way to grow. This costs time and some creativity to make your content engaging. But it's likely a skill worth investing in or hiring for. The platform you choose will depend on the type of product. Video content does better on IG/TikTok. Text is better on Twitter. This is the same as a paid acquisition, but you aren't paying for it (unless you hire someone to do the work).
For me, the holy grail would be to build an extremely sticky product, something people love so much that they share it with their friends (word of mouth), but not engineering it to be viral too soon. Many products that try to go viral too fast either can't handle the huge traffic or get thrown into the category of fads (e.g., the many social media apps that cropped up and then just "disappeared" - Vine, clubhouse, threads?). So, for me, stickiness is a more critical factor, combined with word-of-mouth growth. If both these are met, then the third engine (advertising) would just be adding fuel to the fire. But without stickiness, the business wouldn't last very long.
I'll add some books on stickiness to future book club sessions.
That's it for now. I hope you enjoyed the review.
The book has more sections about innovation in large organizations, management, and team dynamics. But I didn't find those relevant or very interesting, so I skipped them. Feel free to check them out if you are interested in that stuff.
Reading a book like this, which covers pretty much everything about building a startup, gives you a pretty good idea of where your interests lie. For me, I enjoyed the building a sticky product/brand, validated learning, and growth elements the most. But, I wasn't very interested in the team building, raising capital, or "building something where I'm not a user" sections. So, I'm optimizing my life to align with the elements that interest me. You can do the same— follow your interests and delegate the rest, or set stuff up in a way where those things aren't really even a consideration. I'm doing a bit of both.
Overall, even though some parts were hard to power through, it was a great book in the end, and I highly recommend it to all entrepreneurs. Especially if you have a tech-based business. But I'm pretty sure every business has some kind of tech or data in it these days and could benefit from the insights here.
Check it out.
Please help me improve! How did you like this post?
With your anonymous feedback, I can improve the newsletter.