Useful Tips

How to conduct a market analysis: methods, steps, forecasting

How to conduct user research? What types of studies exist and what insights for your product can be found?

Try to imagine that you are mistaken. And now, try to imagine that you are seriously mistaken. Presented? I'll tell you the truth, it happens.

Who are you making products for? Probably for the user? If so, then let's nod at the screen. But, no matter how paradoxical this may sound, you also need to test products on it (on the user). Obviously? Yes, not always. One of the most dangerous traps awaits each of us right around the corner. We think that we know, we know “what they need”, we invest in forces, time, resources, and at the end we get failed ideas, a hypothesis, products. No need to do this, let's do a study.

Why do research

In fact, for two main purposes: to get ideas for product development and find out how much they are in demand by the user.

The most valuable data are those that allow you to draw conclusions and on their basis to qualitatively change something in your product. Qualitatively, in the sense of doing something that will bring profit to the user and business.

You need to conduct research if you do not want to fail with your product. Without research, you will be like that hedgehog in the fog. It seems that I think something, but I don’t know what will fly or not.

Two types of research

Research is divided into two types: qualitative and quantitative. Each of them answers their own questions.

Qualitative research allows you to pull out “pains” and turn them into hypotheses. You communicate with users, are interested in situations in which they use your product, ask about how they do it and gain ideas.

Quantitative research is the next step. You have scored ideas, and now you need to understand how many people they are important. Here is this refinement related to the search for products in the catalog, how many users will improve their lives?

If you depict the process in a diagram, you get something like this:

These are these users: ask them about what is important to them, try for their sake, and they also lie to you.

Normal users and demand from them something more, there is no sense, psychology is to blame. What is important now when you ask may not be important tomorrow. Your main task is not only to conduct a quantitative survey that will show how many people this feature is important, but also to check this feature on users. Is it really going to be used?

Similar situations are often encountered because users cannot “feel” the functionality before it is developed. That is, their idea of ​​“need” is cloudy. One thinks it will work this way, the other in a different way. The difference in the imagination of people adds to you, as an interface designer, “slight uncertainty”. You think I’ll do it like that, because it will be clear, and then, once, it’s not clear.

Plus, the power of context. When a user imagines new functionality, he also imagines the context in which he will use it. But no one can say for sure how it will behave in the future in a particular situation, so again we can fly over with expectations. And then what to do?

As soon as possible give the user a "touch". The sooner he can test the functionality the better. In this case, prototypes, incomplete versions of what the user can "touch" right now, are perfect. Yes, not with such an ideal design, but not as comfortable as we would like, but already something tangible and tangible that meets basic needs.

Prototypes allow you to observe how real users interact with the functionality, what blockers are and what expectations from the interface you need to justify. They can be made of paper (low fidelity), and can consist of full-fledged functionality, up to closed versions of the application (test assemblies), landing page and other things.

I wrote more about prototypes here, you can read and choose the most suitable option for yourself. The key idea is that you do not saw the functionality until it passes the real test from users. Thus, save a lot of resources and insure yourself against failure.

Pill of truth

I would like to add that in research everything very much depends on the question itself and on who poses it. When you directly ask your friend how this idea is for him, then most likely he will say that he’s quite a good idea or will soften the answer so as not to offend you.

The blog has already published material on how to properly ask the user about grocery features. Read it, if you have not done so already, upgrade your skill of questions and it will be much easier to conduct research. The whole point is that the truth is confirmed by actions and only by observing the actions of users can we verify the hypothesis. And not just asking him about it on the forehead.

Another example: but with a prototype

We conducted several interviews with users and pulled out that many mentioned the search function in our clothing catalog on the site. A hypothesis was born: “Do users really need a search?”

If we just conduct a survey on the site and ask “Need or not a search” with two possible answers, we will get that 50% said “Yes”. So you need to develop. We will develop, release the search to the entire audience, take the data and see that in fact 2% use the search. Here you have the spent resource.

What to do? You can run an interactive prototype with a search and design as close as possible to reality. Drive it out on real users, take off the metrics and see how many people actually will use it.

Again, run the a / b test on part of the audience. Display the search icon in the menu and show it, for example, 2 weeks to 30% of your audience, and by clicking on it, put a regular pop-up with an input line. The user will enter a query, press “enter”, the popup will update and you tell him that you are currently developing a search and his (user) participation greatly helps you to do a quality search.

Hypothesis Testing Tools:

  • Google Tag Manger.
  • Functionality for a / b tests.
  • Various pop-ups, pop-ups.
  • Separate landing pages and more.

Look for more hypothesis testing tools in the same material that I recommended earlier.

Try to observe the user in the context of his tasks, this is how you get real data.

As a rule, all studies are aimed at solving some problems. You program yourself that you will find some problems and figure out how to solve them. Of course, it all depends on the goals of the product, but it is useful to concentrate not only on problems, so as not to miss insights. Try to look at the user as a whole, his habits, preferences and behavior, watch, maybe something will come up with a super idea.

Conducting research is not only useful, but also very interesting. You communicate with users, throw ideas, create various prototypes and watch how real people use them. Make adjustments and create what is truly beneficial. There is nothing cooler than seeing how many people use your product and they like it. Do research to make cool products and remember there isn’t much research!

Another strong movie from Nike. About women who are not afraid to express their emotions.