Lean startup: 7 steps to quickly create a quality product

On Habré, a lean startup theme has been raised repeatedly. Someone argued, someone agreed.

Nevertheless, this approach works, and at any level, as Eric Rice claimed. Here for example the practice of applying lean startup techniques from Braden Kovitsa from Google Ventures Design Studio .



So:

1. “Clickable” layout

Most founders think they need to create a 100% natural interface, the final version they want to present to the gist of users, before entering the market. Wrong! Practice shows that even a template site with “clickable” basic elements allows you to collect a large amount of useful feedback from users within a certain period of time. Examples are HomeAway , AVOS , and Duo Security .



At first, I thought that such prototypes would not cause much interest among users, but over time I realized that users “pecked” on such a version of your product - using their experience, you can understand exactly what you should change! From here a lesson - start a prototype in work at once, and then finish it "on the move".

')

2. User Interviews

Instead of working locked up, collect data to help you generate new ideas for your product. In particular, discuss with potential users the problem your startup is intended to solve. Yes, maybe you have heard all this a hundred times, but do not neglect communicating with the client: it is, for example, rinsing the mouth after eating - everyone knows that it is useful, but not everyone has a habit.



It’s very easy to hang on the details: how to find people to talk to? What to talk to them about? etc. Special agencies have been involved in search questions for focus groups for decades; therefore, there are many methods to do this quickly and accurately. For starters, you can create a small questionnaire - the so-called "screener" - it will help you select the "right" people. Then create a conversation script to always direct it in the right direction. If you want to learn more about the successful tactics of the focus group, you should read our detailed manual .



Now that you have no choice, read this article - and go ahead to the “fields”!

3. "Soothers"

To understand whether the function you are running will be required, you can run some part of it. A similar method was applied by the CustomMade project, which allowed people to purchase handmade items. By launching only half of the service (in fact - one button), the project received a large number of clicks from visitors to access it. Only after that the company “finished” the service to the final look. As a result, the data were obtained without a thorough initial study, but they helped to improve the service and get 3 times more responses than expected. Details of this method can be found in this video .

4. Analysis of competitors

When a team develops a new product, each puts forward various hypotheses on the organization of the process. It is quite easy to evaluate someone else's product and identify valuable features with their further use. The main thing here is not to be mistaken, otherwise you can copy a large number of functions completely unnecessary to the user.



As a rule, we perceive the products of competitors as free prototypes. We look at the demand among potential customers, the fact that they love or do not accept, ignore or hate. Based on this data, we can find better solutions for design, marketing and sales.

5. Mini-polls

Surveys are a pretty nice way to get data sitting in a comfortable chair in your own office. However, developing such a survey is not an easy task! Communicating with sociologists, I am terrified at the mere thought of how many easy ways there are to accidentally spoil a questionnaire and make it useless! Despite this, there are a number of proven methods that work exactly.



We adapt the questionnaire to the behavior of customers that we want to study. For example, if we want to understand why our user responds well to one pricing policy, and badly to another, we try to motivate them to answer questions immediately after purchase, and not after a couple of days, when a customer reads our thank-you email.



We mainly rely on “open-ended questions” that help us to understand customer's opinion exactly. After reading the 100 short original and author's answers, you will get much more information than from the statistics, which tells about 32% of respondents who chose option B.



Here are some more successful tactics used in mini-polls.

6. Prototype with real data

The “clickable” layout is good only for the first stage of launching the service. Even more information about user preferences will help you find out the layout with real data. You will probably want to skip this step and launch the final version of the product. Well, or at least "working prototype." Know: this is all wrong! The development of the PRESENT product takes much longer than you think. Do you want to do a minimal amount of "bumps" on the head after the product launch? The layout with real data is the shortest and smartest way.



RetailMeNot is a service that created such a prototype in 2 days. He was "full of holes" and did not have many functions in comparison with the final version of the product, but this was enough for valuable feedback. The benefit of the prototype was invaluable - half of the possible functions were absolutely not needed by the client!



The project repeated this trick 3 more times, gaining new opportunities to improve the design, convenience and clickability.



Few startups resort to creating a prototype with real data. Maybe this is why only a few grow into large successful companies!

7. "On a visit to the client"

Go where your customers are and see how they use your product. Maybe it sounds like a truism, but many believe that they know their client based on the numerous presentations of colleagues at meetings, telephone conversations and statistical reports ... Perhaps, but to truly understand your client, you have to feel how and where he works, how and where he spends his leisure time, in other words, to understand WHAT he lives!



Foundation Medicine is an example of a startup who closely contacted the user: in order to improve their reports, the founders went to the oncology center and studied the interactions of the end user - doctors - with their product. So, for example, they found out that the time for the design of the report was wasted - the doctors still received them only by fax, which means the font was shallow, and the meaning conveyed by color disappeared. The problem was solved very easily, but it would not have been discovered without an “investigation” on the spot.



Lean startup is, first of all, the study of the properties of the product in the fastest and most inexpensive way. Earlier in the text were listed seven of them, but, for sure, there is still a great many.



Did you have a similar experience? Can you share and complement this seven?