Zexler's blog

Web analytics without numbers. How to see the product by user’s eyes

10 0 comments

Analytics doesn't mean only collecting data, it means understanding what is happening.

1. A lot of data is required.

2. People working on products understand and know their users well.

The above statements are mistakes, and the reverse is true. Today we will talk about how the reverse statement for the first mistake can help to correct the unpleasant truth emerging from the fallacy of number two.

The first mistake.

Imagine an endless pool with black and white balls. Your task is to find out which part of the balls is white. Initially, you are in a state of complete uncertainty. How much data do you need to form the answer to the question posed?

If you take 100 random balls and calculate the proportion of whites, then you will know the answer with an accuracy of ± 9.8%. Another 100 balls will increase the accuracy of your answer to ± 6.9%. Another 200 balls - up to ± 4.9%. Another 600 balls - up to ± 3,1%. Another 9,000 balls - up to ± 1%.

Have noticed that the first 100 balls gave us much more knowledge than the next 100? And the balls from 100 to 200 clarified the picture of reality more than balls from 1000 to 10,000?

Let's remember this and move on to the second mistake.

The second mistake.

The second misconception is that we think we understand the products we are working on. In fact, we do not understand.

We are always looking for a way to reduce this misunderstanding. For this, we try new tools and approaches, communicate with users, conduct experiments, and do analytics.

The goal of all these activities is simple - to increase the understanding of your product and users, to find opportunities for growth and development, to increase the share of correctly taken decisions.

Looking at the product through the eyes of users.

What was all this long introduction? I want to tell you about one method that will let you know about your product a lot. This is a very simple way, it is as follows.


Take a few of your users and watch how they use the product.

No, you do not need to conduct interviews with these users, you do not need to consider the metrics that characterize their behavior in the product. It is necessary to literally take a specific user and monitor how he interacts with the product throughout the period of its use.

How to do it? If you have a competently tuned analyst in the product, then it's very simple:

  • Select the group of users you want to study (you will need different users to get answers to different questions).
  • Unload for each user from this group the sequence of all events with their parameters from the moment of using and until the current moment.
  • At the output for each user, you must have a sequence of events in your product. For convenience, break this sequence into separate sessions.
  • Now you have everything you need to look at the product through the eyes of a user. Take the sequence of events and start to look at it, taking notes along the way.

You will see how the user goes through training, what he does first after that, what difficulties he encounters along the way, at which point his first session ends, did he understand by this time the essence of your product, after what time he comes back, Why he comes back and much more.

And now back to the first mistake. You just need to study this way a small number of users (50-100) to form an understanding of:

  • How and for what do people use your product?
  • What problems do they face?
  • Where is the actual use different from the one you designed?
  • Which users remain and which ones are left? What is the difference?
  • Real examples of use

For some reason, many people think that the analyst is about numbers, about mathematics. In my opinion, the analyst is about understanding the processes that are taking place. Figures and metrics can help with this and solve several problems well, but do not confine oneself to them. For several tasks, other approaches, including the analysis of sessions, will be much more useful. PS Disassemble the session of 50 of your users. If you do not find anything interesting, well, okay. You will lose only a couple of hours of your time. But I'm sure you will

Share this article: