Despite Lean Startup and Agile success, many startups still fail to create products ensuring their survival
Eric Ries started writing about The Lean Startup in September 2008, three years later he published his famous book “The Lean Startup” and it became rapidly popular among the startup eco-system because of its promise:
The promise of the lean startup is that instead of building our companies according to myths, we can guide them with facts and the knowledge required to use those facts well. Or put another way, that we won’t waste our time building products or services that nobody wants. Eric Ries — The promise of Lean Startup
A decade later, we can say that Lean Startup kept its promise of reducing waste: time-to-market and cost of new products have been reduced. We don’t see people working 3 years on a product without showing them to the public anymore.
While it now costs a lot less money and time to start a business, the Lean Startup didn’t have any impact on the survival rate of new business:
Lean Startup has had no impact on the early-stage entrepreneur community according to US small-business data: “45% of business in the US fail during first five years. The number has been stable since 1995” Study by Karen McGovern-Hill
Lean startup’s misleading idea that quick iterations based on customer feedback create great products has 3 problems
The lean startup approach to product development has two steps :
- Create a minimal viable product from scratch
- “Ferocious customer-centric rapid iteration” to determine the value and incrementally improve the product
Following those 2 guidelines, the Lean Startup promises you’ll end up with a great product.
I used to believe this promise but my experience has made me rethink it. I now realise that Lean Startup is more a business methodology than a product one. And it is not sufficient to create a great product and falls short because of three problems:
Problem 1: Lean Startup does not tell you where to start. Innovation should start from researching what alternatives customers already use to get the job done
With the emphasis in the methodology on starting quickly to test business hypothesis, lean startupers don’t focus enough on which business hypothesis they should test. The hard thing is to find where the value is for the customer. But if you don’t actively look for it at first you will probably completely miss the point unless you are lucky.
Let’s say you just finished your thesis about creating a machine learning program able to design adapted trainings for athletes so they can improve their performance. Now you want to create a product on top of your algo.
You have the idea of creating a virtual personal coach that analyzes and provides advice to the athlete. You start to list the following business hypotheses which are unanswered questions:
- How much are users willing to pay for this service?
- Is a website the best medium to display the advice?
- What information does the user really care about?
You build the MVP. 4 months after the launch you realize it’s not popular. It’s a dead end because you never knew why it was not working: was it the price? Is it because it’s not an app? You tested a few combinations of everything, but nothing is working and you are running out of time.
Problem 2: Lean startup’s iterations on user feedback only reach local optimisations. You stay near where you started and miss the bold opinions required for an excellent product
I’ve been working with Lean Startup methodology at Theodo since 2013. It allowed me to read hundreds of user’s feedback and I don’t recall a lot that made us realize something that motivated a radical change to the product’s architecture. A lot of them were interesting to make small adjustments (maybe partially because of the anchoring bias or halo effect). For example, feedback is great to understand that some crucial information is missing or that something is not very clear to use. So iterations after iterations, you end up doing a series of local optimizations which don’t guarantee you get to a great product.
Lean Startup practitioners often put too much hope in user feedback to get great ideas. They tend to forget that user feedback is not the only valid way of getting great ideas about their products. What about books? What about mentors and specialists? What about analyzing state of the art already existing products?
After all, you didn’t learn about Lean Startup from your customers ;)
Problem 3: Lean startup does not tell you how to scale if the MVP is successful
There are many scaling challenges (people, sales, logistic, versioning, customer services, ..) that await you once your product is a success.
On the product side one of the challenge is dealing with the many clients asking for change for their specific needs. At this stage negotiations can be hard, it is difficult to say “no” to your customers. Those new features can pile up quickly until it’s too late: too many bugs, new features take too much time to add, the tech team is exhausted. At that moment, a complete rewrite that put the whole business at risk can be seen as the only solution.
Conventional wisdom says that to beat your competitors you need to one-up them. If they have four features, you need five (or 15, or 25). If they’re spending x, you need to spend xx. If they have 20, you need 30. This sort of one-upping Cold War mentality is a dead-end. Jason Fried, David Heinemeier Hansson, and Matthew Linderman in Getting Real
Another challenge is paradoxically to decide “what next?”. This is hard because you need to switch from building a list of features to having real thoughts about how to improve your product. The second one is much harder.
Unfortunately, Lean Startup doesn’t provide much help to deal with these challenges.
Lean Engineering is a product methodology designed to create great products and scale them
Lean Startup is more a business methodology than a product one. To succeed in creating great products, you need to complete it with a product methodology. Some methods like Design Thinking or Marty Cagan’s method from his book “INSPIRED: How to Create Tech Products Customers Love” have a lot of success in the community. I want to introduce you a lesser-known methodology: Lean Engineering.
It is a part of the Lean Management methodology that helped Pixar and Amazon get their amazing success. Lean Engineering is an education system to learn how to make a product that will be adopted by the public. I will share 3 benefits you will get when designing your product using this methodology.
Benefit 1: Start with a good understanding of the “job to be done”, based on market research of which alternatives customers “hire” and what their critical performance is
While most people using Lean Startup start their product “from scratch”, Lean Engineering suggests looking at which solutions people are currently using to solve the problems your product wants to solve.
For example, let’s get back to your virtual personal coach that analyses and provides advice to athletes. This time, instead of starting from scratch, you follow the Lean Engineering advice and start with looking for the value in the current products, which are in your case Strava-like apps and real sport coaches. After interviewing real users, and not future hypothetical consumers, you understand that:
- people value having a real person talking to them because it’s motivating and they need that to actually do the sport
- people like to compare their performance with their friends and other random people
- people want to feel unique
- people want to be proud of themselves: metrics showing that they progressed are popular for that
- people want to brag in subtle ways
Gathering this information will allow you to create a very different product and test different hypothesis, based on the good parts (called the heritage) customers want to keep from existing solutions. And more importantly you now have a model to evaluate the different versions of your product to understand why it is working or not.
You can also detect the legacy (specificities that don’t bring value anymore) of the previous product. For example, a legacy of real sport coaches is that they’re not available all the time, so the client has to adapt his own agenda to be coached, and can’t decide to change its session time last minute.
This will help you create a great first version of your product which you will be confident to iterate on.
Benefit 2: iterate using a product theory based on providing the right emotions to customers
Give the users the emotions they are craving
People love products that give them the right emotions and feelings. Whatsapp is a success because it makes us feel like everything happens in an instant. The sensation of speed is very satisfying. But they are losing tons of users because they became bad on the “privacy feeling”.
Building an amazing product is about understanding which emotions and sensations your customers want to feel while using it and then to maximize them. A banking app should make you feel your money is well protected.
Marty Cagan in “INSPIRED: How to Create Tech Products Customers Love” also points out that emotions are often the driver behind consumer purchases:
Once you have clearly identified and prioritized the dominant buying emotions your customers bring to your product, focus on that emotion and ask yourself where else they might be able to get that need met? Therefore, you’ll find your real competition.
Write “the concept paper” to clarify which parts of your product are critical to delivering the right emotions
Once you have understood which emotions are important for the customers in your context, it is time to write down the concept paper of your product. It answers the question: “How will your product convey the emotions and in what manner?”. It also makes the challenges and the tradeoffs explicit.
For example, the concept of Citymapper is: “Allow anyone to get many transport options to go from A to B in an instant”. Here the tradeoff is getting many pertinent transport options versus the time required to calculate them. There is also a UX challenge because if we want the user to get it in an instant it must be super simple yet to get many transport options you might need a lot of inputs. Those challenges can then be translated into measurable metrics called critical performances. One performance for CityMapper could be “I can see my transport option in less than 10s”.
The concept and the critical performances are crucial to make good decisions about the conception of the product. It also allows people to try bold decisions as they are more aware of what they are doing. User testing becomes useful: get feedback on those bold opinions instead of hoping to get them from the users.
Benefit 3: Scale beyond MVP with a learning system paced by a takt time of new versions
Scale your product thanks to paced releases
The most famous example of a product paced by a takt time is the iPhone: every year in September a new version is released. This forces the Apple team to think at least once a year about where to put the value for the next product increment. Are they frustrated because of the battery? The slowness of websites? The size of the screen?
This allows the team to visualise the challenge of the next month: “we need to increase the life of the battery by 20% while having a 30% more performant processors”. Once the challenge is known, it is easier to find technical solutions … because we know what we are looking for.
Scale your team by creating a learning environment
During the first month of creating a product, teams have to mostly focus on creating new features. But after a while they reach a point where they need more to improve than to add features. And to improve features, you need to have a deeper understanding of the needs of your customer. This is why a learning system like Lean Engineering that focuses on developing deep understanding in people is efficient in the long run. You have more and more allies to build a great product.
“Good thinking, good products” is not an idle slogan — just as “to make products, first we make people”. The deeper lesson here is that lean is not a set of organizational techniques to extract more value from operational processes, as it is often interpreted in many firms, but a set of principles to create the conditions to think more deeply and learn by constantly trying to improve things and seek a better way, small step by small step” Michael Ballé and Daryl Powell in “Good Thinking, Good Products”
Lean startup is a good toolbox to test your business hypotheses. However, it is misleading on product conception because asking users is just not enough to build a great product. You need to have a product theory and Lean Engineering is offering one with emotions at its foundation. A good first step if this article rings a bell is to read “Learning To Scale”.
I hope this will help you make great products.