We have available a Lean Coach Certification MOOC Training Course. Join now!
- 1 What is Lean Startup?
- 2 Build
- 3 Measure
- 4 Learn
- 5 Conclusion
What is Lean Startup?
The origins of this concept go back to the 80s, with the philosophy “lean manufacturing “that applied Toyota for the production of automobiles. This was a production system that reduced to the minimum expression all those expenses that did not They served to create value for the final consumer. One of the key decisions and more known was to distribute the inventory in small quantities throughout the chain
production, these were known as “kanban“, this avoided storing all the inventory in a centralized location, this made it possible to greatly speed up the tasks of most basic manufacturing.
Based on “lean manufacturing“, Eric Ries developed Lean Startup. Ries had failed in some business projects before devising Lean Startup and the conclusion that he got from the failures was simple, he believed that the products did not adapted to consumer demand because they develop the product before have results, so they went blind to the demands of customers.
After a series of failures in various businesses, he created a startup called IMVU, with this business there were several circumstances that supposed the first Learning about the Lean mentality of Eric Ries. He himself explains it in his book and in numerous conferences and interviews, in this project he learned the importance of continuous innovation, the hypothesis approach and the value of the validation of the
themselves through real experiments and learning, from the launch of Minimum Viable Products (MVP) and, ultimately, the development of the scientific method applied to the creation of products, the cycle create-measure-learn, concepts that will go explaining during this guide when appropriate.
Thus, Lean Startup focuses on continuously increasing the value of the product for the end user from the first phase of development with the minimum cost possible, avoiding waste as he calls it.
For this reason, Ries explains how to consider at all times the Customer’s opinion during the process of product development through improvement continuous and experimentation. This allows you to create a product without large investments initials and reducing production costs, but with great value for the end user.
Finally, it is necessary to clarify that, although it is true that the initial approach of Ries on his method was focused on Startups, he explains the evolution of the concept to something more global that would serve for any type of company and product that you want to launch, because for Ries a startup is “that human institution designed to create a product under conditions of extreme uncertainty “so it’s a product development methodology valid for any type of company, always it will follow the same pattern, the create-measure-learn cycle.
Creating basically means transforming an idea into something tangible, in this case implies, raise an initial hypothesis about the product and a functional version at low cost of said product to compare its performance in the market.
So, before creating, we must make hypotheses, that is, to develop the First, the hypothesis must be proposed and, based on this hypothesis, design a Minimum Viable Product (MVP) to measure its actual performance.
So, the first step will always be to propose a hypothesis to be able to develop the MVP that allows to initiate the circuit of continuous innovation. But, from where does this hypothesis come from? Well, very simple, it is composed with internal elements and external.
In order to build a hypothesis, an approach must be made to solve problems, It’s not worth just believing that an idea is valid because someone thought it was good, the business world is full of failures that started like this.
The Lean Startup method tries to improve that primitive creation approach of products with scientific bases, that is, following the performance patterns of the scientific analysis method. Each planning must be well-founded on a falsifiable fact, always raising a problem and its possible (and hypothetical) solution.
A hypothesis always starts from the same place, the intuition of someone at the moment given, that is, of the visionary, the person who sees a possibility in the market to introduce a new product or service. It is the entrepreneur or company (as an entity) who, seeing a possibility in the market develops the strategy in
based on the vision, which will then become a product.
But the hypothesis, continuing with the scientific approach, should not carried out in a hasty or subjective manner, one should not succumb to the temptation that seems to have every entrepreneur to launch the product and “see what happens”, it is It is necessary to carry out a second step that consists of presenting the hypothesis to the public, passing from a subjective to an objective hypothesis.
To remove all subjectivity from the hypothesis, an external investigation is carried out asking the public about the hypothetical product and thus knowing the opinions of the potential market for the same product.
And it is at this moment that a valid hypothesis is formed, after having been validated by the public the hypothesis can start to become something tangible The phase of development of the minimum viable product that will be presented by the results for a more effective learning.
Minimum Viable Product (MVP)
The minimum viable product or MVP (Minimum Viable Product), this would be the most basic functional version of the product, or the product in its minimal expression. That is, the product that allows with an effort and minimum cost get valuable results and conclusions.
The objective of the MVP is obvious, it must serve to validate the initial conjectures and validate the key processes of product development. This serves to get a point starting and start to undertake the project with the security of the data that validate the hypotheses.
When Ries explains what a Minimum Viable Product is, it clearly says that the product should never be bad but simple and functional, it is about making a product useful for the consumer, whose basic functions are fully functional, but without “Unnecessary flourishes” to obtain results.
So, a MVP is something similar to the “classic” beta version of video games, because the product is available to be marketed if it is considered necessary, that, with the MVP you can start to reap the first income of your business while evaluating results and optimizing the product, although, as is logical, the costs will be higher as the real benefit of the MVP is learning.
So, a MVP must be created following different guidelines (although similar in many cases) to the creation of the majority of products that we found in the market. These are:
- First, identify the profile of your potential client. That is, get your product fits and fits the characteristics of the target audience, This information must be obtained during the investigation of the hypothesis.
- Second, check if your hypothetical product and your ideal client fit together, that is does asking directly to your current or potential clients and others agents related to your ideal market. Meet the public to know if customer and product “follow the same line”.
- Third, analyze the competition. This does not only mean knowing who They are to try their products in the first person and to know perfectly the products that rival yours.
The next step when the product is designed and launched to the market is the analysis of results, that is, measuring the interaction and responses of customers with the product. Know what the client really thinks about the MVP.
In this part of the process you should consider not only what the results are we want to obtain and how to get them, but also what are the elements which will help us perform these tasks such as split test or A / B tests, know the differences between vain and actionable indicators, set the metrics correct, the value of experimenting in the first person (“on the spot”) or creating lots small among others.
Achievement of objectives and metrics
To measure the performance or achievement of the objectives, some metrics and thus validate the hypothesis. Surely you will already know the definition of one of this concept and the purpose of them, even so, it is important to remember that for the Lean Startup method is more important if possible.
If defining an objective at the time of making the hypothesis was important, it is equally important to know how to properly measure your performance. Well, if the meter is wrong, the result of the hypothesis would be distorted. So, then, you I will show some concepts that will help you reduce the margin of error of the
measurements, because they are advice or actions of great value for this phase and the outcome end of the hypothesis.
Split test or A/B test
The split test or A / B test consists of creating two identical versions of an element of the product with a single variation and analyze which of the two variants obtain better results according to the objective and defined metric. These tests or experiments serve to know the behavior of the consumer with a specific element and power optimize small components of the product with great security of success and agility.
It is important to know that these experiments are usually done with only two very similar variants with a single differential element, because that element will be the object of analysis. If there are more variations it will be difficult to discern what is the factor that determined the achievement of the objective, so they are often used very sporadic experiments with more than two options or with two very disparate.
Now, surely all the above does not come again and you will already have read on more than one occasion, and although I do not find you anything new I consider It is important enough to know this concept and go a little deeper into it. For Lean Startup, experimentation is essential.
If you did not know this concept, let me show you an example to have an idea clear of what a split test is about. These experiments are usually done in mobile applications and websites, especially in the e-commerce sector, where almost imperative for usability and design. For example, following the case of an ecommerce, who wants to improve their sales conversion inside, do tests on the position and color of the purchase button, the situation of the “car”, the data required to the payment, the steps or pages necessary to make the purchase, the situation of others elements to favor cross-selling, texts and descriptions as well as images of the product or the images of the product are just some of the elements (that in definitive could be any element of the site) on which you can make a experiment. Thus, in a short time you can automatically launch the option with better results and again perform another test, this allows to optimize the product until unsuspected limits.
But maybe you’re wondering why to do A / B testing and not another kind of experiments, why are they so important in this methodology? Well, for these three basic reasons:
- First, they allow for reliable results of consumer behavior and behavior patterns. These experiments are not infallible, but they have demonstrated great reliability, they also offer clear and obvious results, which tt greatly facilitates decision making.
- Second, it allows continuous optimization of the product. It is a method that offers results quickly so changes and improvements can be constants.
- Finally, they are relatively inexpensive. They are not usually costs because there is a large number of very cheap or free digital tools to make them (in case of digital products). In the case of physical products, the cost can be vary depending on each situation.
Vain Indicators vs. Actionable Indicators
Another element that you must bear in mind when making measurements is a error that happens frequently, unfortunately, in most companies and usually be encouraged in many cases by the person who is in charge of the project is confusing vain data with positive or actionable data for the business or product, that is, to confuse false or unrealistic indicators with realistic indicators that they provide more pronounced results. And it is that the vain data are usually translated in unfavorable results for the strategy.
But, maybe you do not know this concept, the vain indicators are that data that by themselves have no value, but give the impression of being really relevant and are often the focus of attention when the strategy is poorly planned. Me I explain, a vain indicator would be, for example, the volume of visits in a site. By
if only this data does not imply any sale and, if it increases by 50%, it does not necessarily entails an increase in sales or income (maybe those visit they do not mean any increase in sales so we would not only be thinking that the hypothesis is correct, but that, in addition, we would be reducing the half the average income per visit, we would be going in full direction opposite).
Now, the actionable indicators are those data or metrics that do offer Realistic results and are significant for the development of the product. Following with the previous example, an actionable data would be the percentage of visits that become sale or the volume of revenue per visit, that is, the actionable data are
contrastables then use two elements combined (for example, visits and income, visits and number of sales, etc.).
Here lies the value of the actionable data and the danger, on the other hand, of the vain indicators. This is no longer an exclusive lesson of this method as it should be Take into consideration in any situation that requires it. But if it’s something considerably needed in Lean Startup because the higher the efficiency
of the analysis will also be the lesson learned and the optimization of the product or decision to pivot, that is, the lower the costs and the greater the benefits.
Finally, as a conclusion to this section, the importance should be stressed to use actionable indicators and know how to recognize them, because we are not only talking fifteen of doing a bad analysis but, as you will see later, choosing a wrong indicator or measuring with wrong parameters can be a huge turning point, a
perhaps insurmountable error that, on the other hand, if the indicators are selected well, they can make modifications much earlier and thus reduce time and costs.
Small lots (or one-piece flow)
In this case, this lesson is inherited from Taiichi Ohno’s Lean Manufacturing, because thanks to the Japanese manufacturers we know that, although it may seem, perform production or task performance in small batches, but in a way continued is better than executing isolated tasks in a massive way and joining the tasks to the final.
To understand this concept there is nothing better than this magnificent example that explain James Womack and Daniel Jones in his book Lean Thinking, which I recommend if you are interested in this topic The example explains how an adult person is being helped by her young daughters to fill envelopes with letter of publicity to send to customers. While the girls begin folding all the letters for later insert them into the envelopes and thus perform the process automatically, style mass production of Ford, the father performs it individually, first doubles one letter, introduces it in the envelope and continues with the whole process one by one, contrary to the girls, he uses a process more similar to Toyota’s just-in-time.
Then the strange girls tell their father that he is wrong, that way to do it is wrong and goes against all logic, because otherwise the letters more quickly and is a more efficient process.
So, the father proposes to do the test and see which method is better, which get better results in the same time. And they set out to perform the test that, by surprise of the girls, it has turned out to be more effective following the father’s method.
But, how is that possible? (And no, the cause is not the efficiency of little girls) As it turns out, despite what may seem, the efficiency of a process is greater if everything is done together than if it is done in parts, that is, creating lots of “Actions” is more effective than performing the actions individually, although, perhaps the speed of performing an action is often greater, they should also be Consider the times of piece transport, change from one action to another. These times are not considered, but they are also important, plus they are a cost greater than the preparation of lots.
But, as if that were not enough, there is another reason why the creation of lots is more efficient, since it considerably reduces the time it will take to discover a failure, an error or a failure in the chain of product development. Well, as we saw In the example of the father and the daughters, what would happen if the envelope is considerably smaller to the size at which the cards are folded? In that case, the daughters would have lost time considerable while the father would have detected the error in seconds so there would be
had more time to “optimize” the size of the letter and, it would not have damaged so many letters (so the extra cost would be reduced as well).
For all these reasons, the development in batches is so important. Reduce costs and cost overruns, both economic and in working time, key elements in the development of Lean projects and clearly important also in many other types of business or product.
Test for yourself
Last but not least (but quite the opposite), try by yourself or test the product on the spot. This Ries puts it as an example of good practices that all project managers should consider, regardless of That it may seem unnecessary, a waste of time or something vain, is not. Ries in his book he names a lot of examples in which to try a product for oneself has meant a substantial change in the development of a product.
Now, the real value of “proving by yourself” the product, that is, get into the mud and know perfectly the strengths and weaknesses, it is something very Logical that, by shame or laziness, does not usually occur on many occasions. Well. who has time to lose trying to interact with the product? This those responsible for developing it will do it … Well, it’s not like that!
Ries explains that, the moment you interact with your product, you change your perception about it, you realize elements that you have not seen and, just in case outside, change your focus on the hypothesis because you’re on the side of development to the side of customers or consumers. This change of focus is very positive because the perception rotates 360 degrees.
Finally, testing for oneself does not only imply testing the product, see what functionalities it has and “play” with it. It is also good in this phase to interact with other consumers after doing the test, interviewing clients
after knowing your product from your point of view generates greater empathy with the public and allows to know other aspects of it.
Learning is the third and final step in the optimization of the product, is the moment to obtain the conclusions about the hypothesis and its results. It is the final step that will lead to persevere in that hypothesis or, on the contrary, to pivot to a new strategy. It is a turning point in the development of the product because it is the moment that really generates value throughout the cycle. In this phase, as mentioned previously, the crucial decision must be made about the hypothesis, persevere in it or pivot to a new one, as this has been proven wrong.
This is usually the easy and comfortable decision, persevere means maintaining the hypothesis as correct and you make the determination to continue optimizing the product following the same guidelines and parameters because the hypothesis is considered correct.
It is often said that perseverance is the comfortable option because nobody likes it recognize and accept the error, instead, think that the results obtained are good It’s simpler, to come to the conclusion that you should change is usually an arduous process.
For this reason, as you will see below, you should pay close attention to the indicators used to make the decisions and with the selected metrics, because if The pillars of learning (the collection of information) are based on knowing how to do correct questions (hypothesis) and know how to measure their result (metrics and indicators).
How it was explained in the part “vain indicators vs. actionable “, when the mistake of using vain indicators falls on the implications not only affect the immediate results and internal perception about the hypothesis, also implies taking a wrong decision, that is, it may seem that the results are favorable and decide to persevere with the hypothesis when in fact the correct decision would be to pivot of strategy.
When the results are not as expected, the hypothesis is proved wrong and it is not enough to optimize the product because the problem is deeper, the approach strategic is equivocal, the solution is the pivot. Pivoting, as Ries explains in his book, it’s just another way to call for change. The strategic pivot is necessary because it avoids unnecessary waste of resources with a hypothesis doomed to failure.
The pivot is something natural, it happens more times than one could imagine Therefore, most failures are usually caused by not making the decision to modify the strategic approach at the right time, whether due to disability, lack of data or any of the reasons stated in this guide. Therefore, one should not fear “failure” of having to pivot a strategy then, the real failure will end up being the fact itself of not having been able to make this change of course.
Catalog of pivots
So, as the final culmination of this guide to apply this method would only be missing know the possibilities that exist and are defined by Ries in his book to pivot the strategy. Once you are clear that the hypothesis should vary you should know the different existing variants to continue improving the project. I do not believe necessary, but even so I’ll do it, clarify that the types of pivot defined that you I will expose below are just a reference, are the different paths to guide you in this way, but can the characteristics of a specific project should be governed by a combination of these or other pivot forms. That said, the pivot types defined by Ries in his book are:
Zoom pointer (zoom-in)
The approach pivot occurs when a characteristic, within the Many that a product may have, it becomes the product itself. This situation usually occurs when users feel more affinity, it seems useful or they are simply attracted to a single feature among all the functionalities of the product.
Remote pivot (zoom-out)
In this case it is the opposite situation to the “Zoom-in”, when it seems that a product with a unique feature or functionality is sufficient proves not to be, so the entire product becomes part of an even greater product being this first product only one function of the big one.
Consumer segment pivot
This happens when a product solves a problem or meets a need to a segment of consumers that was not raised from the beginning, that is, the segment that hypothetically was considered as potential consumers is not becomes a real consumer as there is another segment, not considered at the beginning, for which product is useful for some reason. In this case the pivot occurs towards a new consumer segment in a normally natural way.
Consumer need pivot
This situation happens when a need is detected in the consumer it can be solved with the product but it is not being covered. In this point is when the vision of the product developers should realize that “gap” in
the market and offer a solution to the market’s need.
This refers to a change in application on a platform or vice versa. That is, for some reason, often related to the growth of the project. The company needs to make a change of vehicle. This is a very common situation
especially in startups where the change can occur several times.
This type of pivot occurs when a company finds a solution technology that allows you to better achieve your goals or be more efficient in some aspect and considers a change towards that technology that offers you better possibilities adapted to your need.
This pivot is produced by the discovery that a main solution can be supplied through another medium or channel with greater efficiency. That is, the use of a channel different from the one currently being used would save resources or could be more efficient with the same cost on some of the key aspects of
business, so it would mean a substantial improvement in the product.
Growth engine pivot
Ries in his book defines three types of growth, viral growth, sticky and remunerated growth. Evidently there are other classifications of growth, what is clear is that the growth pivot refers to the change in the growth strategy. This modification in the growth engine also usually causes a change in business capture (explained below).
Value capture pivot
As Ries explains in his book, the capture of value is what is usually known as monetization or income model. As I suppose you can imagine, the pivot value capture consists of modifying the way to monetize the product, it is say, change the income strategy. This kind of pivot must be meditated a lot the change in the way to get income implies very deep changes in consumer behavior, in purchasing psychology, that may have a enormous scope in the rest of the business, the product and, above all, the strategy of marketing.
Business architecture pivot.
This type of pivot starts from the business models proposed by Geoffrey Moore in which he distinguishes two main architectures, those that are more widespread.
These models are; the complex system, which consists of having high margins and low volume, widely used in B2B. While the volume of operations model, which, on the contrary, it uses low margins and high volume, used for the public in general (B2C).
So, the business architecture pivot is to change a model to another (or to any other outside of these two here defined). That is, move from a low margin model and “exclusive” sale to reduce margins to sell “Massively”.
Finally, we would only briefly recall the process discussed here to try to clarify the main ideas of the process proposed by Eric Ries in his Startup method. The process is cyclical, based on three phases: first, the creation of product based on the hypotheses raised, second, measurement of results by means of measurement tools and techniques combined with the planning of metrics and indicators that help assess the process and, third, learn and draw conclusions about the previous stage to be able to decide between persevering in the hypothesis to optimize the product or pivot strategy.
By definition this process of creating-measuring-learning is cyclical because the market it is changing and always evolves so that a product will never be static, simply be better or worse adapted to the market, you will need a process of more extensive or lesser improvement, but by definition it will never be finished (and less so the times in which the technology is renewed every few months and the consumers have a lower sensitivity to change than before).
So, the most important lesson that emerges from the Lean method is that innovation is continuous, the evolution of the market does not stop and, if a company does not change at the same pace or anticipate going to fail, it will become a failure, more if it is a startup or recently created project which has neither the
experience or the sustenance of the market as a guide to design their strategies. For these reasons, it is vital to have a clear and well-defined strategic vision, that the entire team be well coordinated and row in the same direction, know how to make sound decisions at the exact moment and, above all, think like an entrepreneur because, according to the own definition of Ries in the first pages of his book, the entrepreneurs are everywhere, are all those people who work under the definition that he poses of startup (“human institution designed to create a low product conditions of extreme uncertainty “) because the Lean Startup method can work in companies of any size, sector or activity.
So … What are you waiting for to start applying the Lean method?