The most common idea for platform optimization of user experience is to determine the user's access path and obtain an effective path from it. Website map is a commonly used behavior analysis tool to understand the user's behavior trajectory in the website and assist users to visit the website and other services. This article will introduce and discuss some simple uses of website maps, hoping to help you.

preface

The most common idea of platform experience optimization is to determine the user's access path and find the most effective path from the access process. From the perspective of the whole website, website map is a commonly used behavior analysis tool.

The application of website map is essentially an exploration. Through the map (if you don't have a website map, are you lost in the APP, To further consider, if we concatenate the three groups of data that customers visit each time, package them into a node cell, and then concatenate the cells generated by customers' visits in our company along the timeline, a complete life cycle of customers in our company will be formed.

The exploration of the life cycle will be described in detail in the following article. In this article, we will first introduce and discuss some simple uses of the website map. If there are shortcomings, please correct them~

1、 Access fluency of users

The most intuitive idea for users after entering the website is to find the target transaction and complete it. In this process, users will click one page after another to complete the jump. The two main factors that affect users' jump are: jump duration (jump performance) and clarity of transaction information. These factors are resistance. From the data point of view, conversion rate is a major indicator to calculate whether there is resistance in the access process.

In a normal access process, what causes the conversion rate to decrease?

From the perspective of data, we can indirectly judge by where the customer went:

1) After visiting Page C on the above page, the customer returned to other pages, and then completed the transaction on other pages, or stayed. We can preliminarily judge that the customer's resistance this time was due to the failure to find the desired transaction. As shown below:

This situation is what we often call the wrong way. Most of the problems are caused by unclear expression on the diversion page and ambiguous button presentation. The customer can't understand the transaction corresponding to the button, so try to look at it one by one. This is a time consuming and patient consumption for the customer.

2) Some customers arrive at page C, then return to the previous page, and then return. As shown in the figure:

For customers, this is often because the information they make decisions on page C stays on page A or page B, leading them to return to page AB to check the corresponding message, and then return to page C to decide whether to continue to page D. In this case, the calculation often needs to combine the conversion rate of PV and UV. By comparison, it can be found that there are certain differences between PV and UV conversion rates of C-D pages. PV conversion rates are significantly smaller, while UV conversion rates are in a reasonable conversion range.

3) After arriving at page C, the customer jumps away without stopping, as shown in the figure:

Because there are few data left, it is difficult to determine the reason for this situation. Let's guess several obvious reasons: the page jump is relatively slow, the user's intention is not to clinch a deal, the current page's product attraction is not strong enough, and so on.

2、 Identify lost users

When you visit this app, do you want to find a function but can't find it on either side? I believe there is. Especially when I first used a software, the author wanted to find a function of "QQ Email Reminder" when using WeChat. After a long time of searching from left to right, I finally found the function one by one in "Settings" - "General" - "Auxiliary Functions":

From the above description, do you have any ideas?

What would a customer do if he could not determine the functional location?

1) Search the corresponding function directly;

2) Walk through similar function menus according to your own cognition;

For the operation in WeChat just now, the author has also traversed the "Account Management", "General" and other buttons. It's not easy to design such a simple interface on WeChat. What about some other apps? I don't think we can reach the destination in three steps, can we?

Let's summarize the customer's lost behavior through the above case:

1) Search back and forth between multiple similar pages;

2) Enter the search bar to search;

The intuitive findings in accessing data are:

1) The access link is long;

2) Repeatedly jump to the same page;

3) Some pages stay longer;

If we calculate the Session_pv (total number of pages)/session_uv (number of duplicate pages removed) We will find that this value will be higher. Let's call this value Labyrinthine index 。

If there are 100 customers and 100 visits, 100 ratios will be generated.

If the mode and median of this ratio are higher than the average value, it means that most customers have a high maze index, and there is a certain phenomenon of getting lost during customer visits.

Based on this assumption, we further traverse the path of high repetition, explore one by one, and discover the secret behind repetition.

For example:

We find that in the process of customer access, the repetition rate of "Account Management" and "General" menus is high (PV/UV), and the probability of two menus appearing at the same time (association rules) is also high. Will we draw some inferential conclusions when we take this access path (pattern recognition) separately?

For example:

1) The name setting is not clear, and the customer can't determine its corresponding function range when seeing it, so it needs to click in to determine;

2) The information is not complete enough, and the customer needs to operate half in "account management" and the other half in "general";

After reading this, will you have some new ideas?

Welcome to leave a message~

Well, let's stop here. In the website, there is no end to the analysis of the path. After all, this is the place to visually present the customer's actions, such as: how to determine the user's intention? Under what circumstances will customers form orders? Can users' access paths be added when predicting customer transactions? Is there any secret trace in the access path of customers under Bayesian conditions?

That's all for today. If you are interested, please come and chat~

Columnist

Ye Shuijing, WeChat official account: livandata, everyone is a columnist of product managers. Internet veterans in the financial industry focus on data driven, integrate algorithms and data into product design and operation strategies, and build financial growth methodology.

This article is originally published by everyone who is a product manager. Reproduction is prohibited without the permission of the author.

The picture is from Unsplash, based on CC0 protocol.

This article only represents the author himself. Everyone is a product manager. The platform only provides information storage space services.