How to do sites. Open access to Yandex statistics for Landing Store
Colleagues, first the very essence - I open general access to Yandex statistics for one of my sites (Landing clothing store). Who will be interested in analyzing user behavior, and get some dry statistics, you are welcome.
And now I will tell why and why all this was decided to be done.
Note: all values ​​related to visitor statistics are given as of the writing date of the article - 04.03.15.
')
Some time ago I wondered about spending a small budget and decided to try to start my own business. After working through several options I stopped at creating sites. Since the possibilities were limited, there was no talk of a full-fledged web-studio. The choice fell on the creation of templates for ready-made sites that are hosted on the site of ready-made solutions of the Bitrix Marketplace (hereinafter referred to as MP).
A designer, coder and programmer has appeared. And so (in fact, several months have passed, but this is a completely different story ...) our first ready-made solution came out - adaptive Landing clothing store “Beauty”. We successfully published it in the MP and I decided to put a counter on it from Yandex.
In fact, at first it was just curious how much a person looks at my decision at all, but everything turned out to be much more interesting.
Analyzing the statistics of visits, it turned out that the hypotheses and assumptions, in accordance with which the tasks for the development of sites were set, are not always accurate, and sometimes diametrically opposed, which ultimately directly affected the sale of the solution, or rather the lack thereof.
So, what finally managed to learn from the analysis of visits to your site:
Setting the task for the designer to create the site, I didn’t go deep into the details and said - just make a beautiful site. The same was with the layout - it was required that it was just adaptive and that's it. At that moment it seemed to me that if the layout is correctly displayed on the phone, it means that it is correctly displayed on all other screens. This turned out to be one of the delusions. Of course, they also beat the criteria for successful completion of HTML and CSS validators (links to sources at the end of the article), but in fact this also had little impact on the success of the site. And such a superficial approach ultimately bore fruit, negative naturally.
I even could not think that users with a screen resolution of 1920 and more can be so much! Therefore, at this resolution, the finished layout was not even tested. Through special resources, only mobile layout was tested - it turned out to be a big mistake. Everything happened exactly the opposite.
The number of visits to users with a resolution of 1920 and more turned out to be the most! 20 users out of 80. At this resolution, the design of the site was not even tested - this is the first blunder. It turned out that on such a screen the site looks rather poor:
Testing was carried out using the service quirktools.com/screenfly
The site cap is too stretched, and the workspace is too compressed.
Total: out of 25% of visits with a high resolution, 25% of failures. And the higher the resolution, the higher the percentage of failures:
• 1920 × 1080 - 23.5%
• 1920 × 1200 - 33.3%
While visits from mobile devices - only 17%.
As a result, a lot of effort was left to adapt the site to mobile versions, but there were no mobile clients.
One of the requirements for the layout was the correct display in all browsers. It seems that this is nothing special, but everything should be in moderation. We spent a lot of time adapting the site to IE 8 and below. And what happened in fact? And the fact that with IE in general, no one comes - 0%. 43% of the visit is on Chrome, 15% on Safari, 14% on Opera, the rest on trifles.
Conclusion: it is not necessary to spend large resources (temporary and material) to support stupid browsers. It is better to invest in other parts.
I think this item does not need explanations.
The design of our site is very simple. Animation in one form or another is completely absent - there are no animated buttons, menu items, icons, pictures, etc. This, in my opinion, turned out to be another omission. I do not in any way want to talk about turning the site into a New Year tree, but still it would be better to highlight some elements.
After analyzing the map of clicks, it turned out that visitors, for some reason, always strive to click on the pictures, even if it is clear that they are not animated anyway.
Benefits:
Order Card:
It would be better to provide active and non-active states for these elements, so that when you hover the pointer, they will at least somehow change. Then it would be both more interesting and easier for human perception because glance at least for something, but would catch. And it so happened that we were driving a finger over a piece of newspaper.
As is known, the site logo is usually made to link to the main page. But I thought - what is it for landing? After all, this is a one-page, and the menu is not floating, so clicking on the logo will lead to nothing. Nevertheless, the logo all the same click. Even despite the fact that when you hover over it, the mouse pointer does not change. And this is understandable - everyone is accustomed to the fact that the logo is a link, and then there is some kind of gap patterns.
Conclusion: it is always better to make a logo with a link than a link, even if it does not lead anywhere. This is more familiar to the user.
Summarizing what was currently possible to learn from the statistics of Yandex, I would like to highlight the main mistakes that in my purely personal opinion should not be done when developing websites:
1. The most important thing is testing. The more testing tools you do not use, the more bounce you will get;
2. Do not deprive widescreen monitors . No need to proceed from the theory that the resolution of 1366x768 - is 90% of all visitors. In practice, it turned out that the largest percentage of visits was with a resolution of 1920 and above;
3. Development costs for stupid browsers are not justified . Technologies go forward, and the smallest percentage use it;
4. Competently evaluate the labor costs for adaptive for mobile screens . This, of course, is necessary, but with a limited budget, it would be more appropriate to focus on other more significant aspects and not to focus on the ideal layout for the phone. Best the enemy of the good. Let it be simple at first, and when the profit from the project goes, then you can work on it.
5. Do not turn the site into a piece of newspaper . If the button is pressed, then when pointing it should somehow declare it, for example, not an obsessive background change. If there is an icon with a small text block - select it a little, it draws attention of the user, for sure he will be pleased.
6. Logo always make a link .
PS: all the hypotheses, conclusions, conclusions given in the article are a purely subjective assessment and in no case do not claim to be undeniable truths.
I will be glad to hear your assessments and / or observations that will appear when you view site statistics. Of course, the MP audience is specific, but nevertheless, some observations can be traced here.
1. HTML validator: validator.w3.org
2. CSS validator: jigsaw.w3.org/css-validator
3. Service for testing adaptive layout: quirktools.com/screenfly
And now I will tell why and why all this was decided to be done.
Note: all values ​​related to visitor statistics are given as of the writing date of the article - 04.03.15.
')
Some time ago I wondered about spending a small budget and decided to try to start my own business. After working through several options I stopped at creating sites. Since the possibilities were limited, there was no talk of a full-fledged web-studio. The choice fell on the creation of templates for ready-made sites that are hosted on the site of ready-made solutions of the Bitrix Marketplace (hereinafter referred to as MP).
A designer, coder and programmer has appeared. And so (in fact, several months have passed, but this is a completely different story ...) our first ready-made solution came out - adaptive Landing clothing store “Beauty”. We successfully published it in the MP and I decided to put a counter on it from Yandex.
In fact, at first it was just curious how much a person looks at my decision at all, but everything turned out to be much more interesting.
Analyzing the statistics of visits, it turned out that the hypotheses and assumptions, in accordance with which the tasks for the development of sites were set, are not always accurate, and sometimes diametrically opposed, which ultimately directly affected the sale of the solution, or rather the lack thereof.
So, what finally managed to learn from the analysis of visits to your site:
1. Users with widescreen monitors are more than you think.
Setting the task for the designer to create the site, I didn’t go deep into the details and said - just make a beautiful site. The same was with the layout - it was required that it was just adaptive and that's it. At that moment it seemed to me that if the layout is correctly displayed on the phone, it means that it is correctly displayed on all other screens. This turned out to be one of the delusions. Of course, they also beat the criteria for successful completion of HTML and CSS validators (links to sources at the end of the article), but in fact this also had little impact on the success of the site. And such a superficial approach ultimately bore fruit, negative naturally.
I even could not think that users with a screen resolution of 1920 and more can be so much! Therefore, at this resolution, the finished layout was not even tested. Through special resources, only mobile layout was tested - it turned out to be a big mistake. Everything happened exactly the opposite.
The number of visits to users with a resolution of 1920 and more turned out to be the most! 20 users out of 80. At this resolution, the design of the site was not even tested - this is the first blunder. It turned out that on such a screen the site looks rather poor:
Testing was carried out using the service quirktools.com/screenfly
The site cap is too stretched, and the workspace is too compressed.
Total: out of 25% of visits with a high resolution, 25% of failures. And the higher the resolution, the higher the percentage of failures:
• 1920 × 1080 - 23.5%
• 1920 × 1200 - 33.3%
While visits from mobile devices - only 17%.
As a result, a lot of effort was left to adapt the site to mobile versions, but there were no mobile clients.
2. Stupid browsers go into non-existence - this is a fact
One of the requirements for the layout was the correct display in all browsers. It seems that this is nothing special, but everything should be in moderation. We spent a lot of time adapting the site to IE 8 and below. And what happened in fact? And the fact that with IE in general, no one comes - 0%. 43% of the visit is on Chrome, 15% on Safari, 14% on Opera, the rest on trifles.
Conclusion: it is not necessary to spend large resources (temporary and material) to support stupid browsers. It is better to invest in other parts.
3. Cookies are enabled on 100%.
I think this item does not need explanations.
4. Users always want to click on pictures.
The design of our site is very simple. Animation in one form or another is completely absent - there are no animated buttons, menu items, icons, pictures, etc. This, in my opinion, turned out to be another omission. I do not in any way want to talk about turning the site into a New Year tree, but still it would be better to highlight some elements.
After analyzing the map of clicks, it turned out that visitors, for some reason, always strive to click on the pictures, even if it is clear that they are not animated anyway.
Benefits:
Order Card:
It would be better to provide active and non-active states for these elements, so that when you hover the pointer, they will at least somehow change. Then it would be both more interesting and easier for human perception because glance at least for something, but would catch. And it so happened that we were driving a finger over a piece of newspaper.
5. The placebo effect
As is known, the site logo is usually made to link to the main page. But I thought - what is it for landing? After all, this is a one-page, and the menu is not floating, so clicking on the logo will lead to nothing. Nevertheless, the logo all the same click. Even despite the fact that when you hover over it, the mouse pointer does not change. And this is understandable - everyone is accustomed to the fact that the logo is a link, and then there is some kind of gap patterns.
Conclusion: it is always better to make a logo with a link than a link, even if it does not lead anywhere. This is more familiar to the user.
Summarizing what was currently possible to learn from the statistics of Yandex, I would like to highlight the main mistakes that in my purely personal opinion should not be done when developing websites:
1. The most important thing is testing. The more testing tools you do not use, the more bounce you will get;
2. Do not deprive widescreen monitors . No need to proceed from the theory that the resolution of 1366x768 - is 90% of all visitors. In practice, it turned out that the largest percentage of visits was with a resolution of 1920 and above;
3. Development costs for stupid browsers are not justified . Technologies go forward, and the smallest percentage use it;
4. Competently evaluate the labor costs for adaptive for mobile screens . This, of course, is necessary, but with a limited budget, it would be more appropriate to focus on other more significant aspects and not to focus on the ideal layout for the phone. Best the enemy of the good. Let it be simple at first, and when the profit from the project goes, then you can work on it.
5. Do not turn the site into a piece of newspaper . If the button is pressed, then when pointing it should somehow declare it, for example, not an obsessive background change. If there is an icon with a small text block - select it a little, it draws attention of the user, for sure he will be pleased.
6. Logo always make a link .
PS: all the hypotheses, conclusions, conclusions given in the article are a purely subjective assessment and in no case do not claim to be undeniable truths.
I will be glad to hear your assessments and / or observations that will appear when you view site statistics. Of course, the MP audience is specific, but nevertheless, some observations can be traced here.
Links to the tools cited in the article
1. HTML validator: validator.w3.org
2. CSS validator: jigsaw.w3.org/css-validator
3. Service for testing adaptive layout: quirktools.com/screenfly
Source: https://habr.com/ru/post/286372/
All Articles