Traffic sources in Google Analytics: who is more important? Analyzing traffic sources

After creating a site, we immediately become interested in who is visiting it and how users find it in the first place. You can, of course, ask users how they came to you :), but the easiest way is to install analytics systems. I use both Metrica and Google Analytics, but I prefer the latter. And in this article I will tell you how they behave and interact with each other different sources traffic.

When you log into your Google Analytics account, you see something like this:

Organic Search is traffic from search engines(Google, Yandex, etc.).
Direct– direct visits to the site.
Referral is the traffic you get from links on other sites.
Social– traffic from social networks.
Other– traffic the source of which Analytics could not determine.
Email– visits from your mailings.
Campaigns– this is a separate type of traffic via links with utm tagging.

It would seem that everything is quite simple: a user found you on Google and came to your website - this is an organic visit; you were mentioned on the forum, and users who want to learn more came to you - this is referral traffic; someone shared your link on Twitter and people came to you through it - that's social traffic.

But think about it, you often find something interesting on the Internet and save it in your bookmarks to watch later. In the first case we get an organic visit, in the second - a direct one. How will this be shown in Analytics?

Hierarchy of traffic sources, or who is more important.

First, try to answer the question: which traffic source gives the least amount of traffic? useful information for you? In my opinion, this is direct traffic, and therefore it is in the hierarchy of sources that it looks like the “weakest link”; it is precisely this that is “eaten up” by all other channels.

Examples of recording traffic sources in Google Analytics

Organic -> Direct = Organic

Misha reads a lot and follows all the new trends in real estate. He found one on Google interesting article(organic), but there was no time to read it, so Misha simply left it in the browser until better times. They arrived a week later, Misha got to the article, which was already open in his browser (direct). However, this young man will still be considered a user coming from a search engine.

Direct -> Referral -> Organic = Organic

Sasha came to the site via a link sent to him by a friend (direct); a week later, while reading some blog that links to your site, Sasha clicks on the link and again ends up with you (referral). After another 2 weeks, Sasha developed a very important question, and to solve it, he Googled it, as everyone usually does in modern world. From the search, Sasha clicked again on your site. In Analytics, this visit will appear as an organic one that “overshadowed” the referral visit.

After Sasha was convinced for the third time that your site was useful, he saved it to his bookmarks, and then returned to you from them several more times. Attention, question: what source of the visit will be recorded in these cases?

Referral -> Direct -> Direct = Referral

Masha came to the culinary site from a forum (referral), where she was looking for how to make mulled wine. She liked the recipe and saved the link to it in her recipe file. After that, there were several more holidays when Masha needed to prepare mulled wine, and she came to the culinary site, copying the link from her file. But in Google Analytics, Masha was still listed as a user who came from another site (referral). Then Masha remembered the recipe and stopped visiting that culinary site, but she often looks for other recipes. But that's a completely different story.

It is important to know

All previous conclusions on the “seniority” of traffic sources were made under the following conditions:

The user visits the site from the same computer and browser.

The user always has Javascipt enabled in the browser.

The user does not clear cookies.

Between the first and last visit no more than 6 months pass in the chain*.

6 months– this is how long by default information about any campaign is saved in Google Analytics. Universal Analytics allows you to change this period up to 24 months.

This is how you track users who come to you every day. Such a system once again proves a simple principle of life: not everything is always as it seems; The user who comes via the link is not always registered as direct. I hope I was able to help you understand how things really work with the interaction of traffic sources.

How do visitors behave on your site and do they come back to you? How does an advertising campaign affect other traffic acquisition channels?

Link accounts

In order to see the full picture of your effectiveness advertising campaign you will need to link your account Google AdWords with an account Google Analytics.

This will give you the opportunity to see not only more data in the AdWords interface, but also track how the advertising campaign affects other channels, and see the payback of your advertising. And also the connection between AdWords and Analytics will allow you to set up more subtle settings through the Analytics account.

So that you can track in detail which keyword brought you orders and which page was the most effective, you should also tag the links. There are plenty of resources online for manually tagging links if you don't use the automatic tagging feature.

For example, Google itself offers a URL builder, a handy resource that helps you add required parameters to track campaigns, from campaign source to keyword.

Define your goals

First of all, before you track the effectiveness of an advertising campaign, you need to determine what you will track? Subscribe, order, watch a video, click the “Order” button, etc. as well as the correctness of the configured goals.

The correctness of setting goals (as well as setting the analytics code itself on the site) can be seen in Real time report. In addition, you can see who is on this moment be on your website, from which channels it came this traffic what content are your visitors currently viewing and if we're talking about about the correct setting of goals, then you can track correct setting in the Conversions section.

This report shows any technical problems on the website or with a real-time advertising campaign.

Channels and Traffic Sources

In Channels and Traffic Sources you can see from which channels your visitors come. And also how long they stay on your pages, i.e. How interested they are in your site's content. Which channels generate the most conversions (or let’s say, achieving goals).

So, what does each of these channels mean and include?

Referal – transitions from sites where links leading to your site are located.

If you see enough large percentage traffic, then you should pay attention, this may be referral spam. It does not affect the operation of the advertising campaign or the operation of the site, but it unpleasantly spoils the statistics. In the screenshot, pay attention to the Average session duration - it is equal to 0. That is. there was no activity on the site.

Organic Search – traffic from search network. In this channel you can see for which queries you are found in organic results.

Direct – when visitors go to the site by entering a link on the site in the browser line or via a bookmark. Traffic from this channel is also called “branded”. And over time the Display Network operates, the percentage of traffic by this channel should increase. This means that visitors are already familiar with your brand and enter your brand name directly into search bar browser.

Paid Search– this is paid search traffic, which is what we are currently considering.

To understand more detailed information, you can go directly to each channel.

The picture shows which campaign attracted a certain number of clicks and orders.

What to look for?

Please note the discrepancy in the number of clicks and sessions. The thing is that in Google Analytics, clicks– deciphers the number of transitions to your site from search engines, and sessions– this is the period of time during which the visitor interacts with your site. Term Session includes the concepts of both the number of screens viewed and transactions on the site.

The bounce rate is also quite important. This indicator is related to the effectiveness of pages on the site and the interest of visitors in the content on these pages. The higher the bounce rate, the more likely the site or page has a problem, such as: the page content is inconsistent key query or advertisement, not found on the site necessary information or clicking occurs.

If you have a one-page website, then the failure rate can reach up to 90%. The viewing depth of a single page is checked by the percentage of scrolling (for this you need to use a special page scrolling script and set it to ).

For regular blog sites or online stores, the failure rate is no more than 40-50%

The small percentage of refusals should also be alarming. Some of the reasons could be: an incorrectly installed analytics counter, or the same analytics counter is located in both the GTM and on the site at the same time, or the navigation on the site is bad and people simply “wander” in search of the information they need.

Conclusions need to be made both for the site in general and for each individual page. Does the landing page match the key query that leads the visitor to the landing page. Does the name of the product or product group contain the key query itself?

Those. the keyword should be mentioned in the ad and the information in the ad should be mentioned on the landing page.

But it's possible that campaigns that didn't result in direct transactions helped drive conversions through other channels. This can be seen in cross-channel funnels, how users completed their conversions, which we will look at in more detail in the next article.

In the Traffic Sources section - AdWords section - Keywords tab, pay attention to which words bring you more income and whether their cost per click pays off.

I would like to draw your attention to the fact that with the help of Google Analytics you can see how many actions a visitor had to take in order to make a conversion.

Those. In the picture you can see that 26 conversions were completed in one interaction, the rest of the conversions were completed in two or more interactions.

We can say that the navigation on the site is done perfectly, but this is a one-page site, in the case of stores, you need to pay more attention to what difficulties your visitors may encounter while getting to their “cherished goal”.

I want to warn you right away that many people make the mistake of conducting analytics only on traffic sources in standard reports. Paid Search and Display often don't look as impressive as Direct or Organic in standard reports. But in order to understand all the details, I advise you to do a more detailed analysis.

In the report Multi-channel conversions You can look at the Time to Conversion report to see when conversions usually occur.

You can also see when your visitors usually make a purchasing decision. The Multichannel Conversions – Time to Conversions report can help you with this.

By using this report You can understand after what time to launch remarketing.

Pay attention to the Time of Day report in the AdWords section

Pay attention to the amount of traffic on certain days of the week and which days bring in the most purchases.

Tying Google accounts Analytics and AdWords, you can quickly and conveniently track the performance of advertising campaigns and optimize promotion based on data. We have already written about Analytics and AdWords. Today we have a selection of Google Analytics reports to evaluate the effectiveness of paid traffic.

Group of reports “Traffic Sources”

This group allows you to compare channels, sources, ad campaigns, and keywords using standard metrics such as bounce rate or average session duration, as well as conversion metrics.

Campaign report

To build a report, open the “Traffic Sources” report group, then “All Traffic” and select the standard “Source\Channel” report.

Select the source of interest, for example google/cpc, and click on it.

Set the Campaign option. To do this, in the “Main parameter” line, click on the “Other” item, then “Traffic sources” - “Campaign”.

Here you can evaluate the amount of traffic that a source and channel provides and its quality based on bounce rates, viewing depth, and session duration. This will help in optimizing work with channels and budget allocation.

Using this report, you can analyze the profitability of paid sources. To build a report at the source level, open the “Traffic Sources” report group, then “Campaigns” - “Cost Analysis”.

In the table, in the “Source or channel” column, select the required source, for example google / cpc and click on it. Data for campaigns belonging to this traffic source will open.

You can see which campaigns are performing better and which need to be improved. You can immediately analyze the effectiveness of keywords in the campaign of interest by selecting it in the column and clicking on it. Based on this data, adjust your bids based on the profitability of keywords or completely disable those that are not generating revenue.

Time to conversion

To find out how long it takes on average for a user to make a purchase decision, use the Time to Conversion report. To view statistics, select the “Conversions” report group, then “Multi-channel” - “Time to conversion”.

The report shows how many conversions and with what value occur on different days after the user first encounters the site. This data will help you understand the purchase cycle and, when setting up retargeting, determine how long to show ads to the user.

Multi-channel sequences

Typically, before placing an order, the user visits the site several times and from different sources. For example, users might first click on an ad and then arrive at the site with free search. And if this is a common situation, then we can assume that conversions are made mainly by organic users, and advertising is not working effectively. But after turning off advertising there will be an immediate decrease in sales.

Therefore, it is important to track the entire path potential clients: where they usually come to the site and what routes they take to return to it before converting. It will also help you set up cross-marketing using different channels of user interaction and correctly set up attribution models in AdWords. This can be done using the Multi-Channel Funnels report group. We’ll tell you more about these reports; we also have a large and detailed one on this topic.

Shows all channels that the user interacts with before making a conversion. To view the report, find the “Conversions” item in the lower right sidebar, then select “Multi-Channel” - “Top Paths to Conversion.”

So, often the user comes first with Email, social networks and search, both paid and free.

A conversion is considered associated for source 1 if the user first went to the site from it, but completed the conversion when returning to it from source 2. In the report, you can see the number of associated conversions and conversions by the last significant transition for each source.

To build a report, under Conversions, select Assisted Conversions.

By default, the report is built by channel groups. It can be drilled down to a specific channel, source or campaign. For example, you need to obtain data on campaigns on the social network VKontakte. To do this, select the “Source or channel” parameter and click on vk_ppcworld/social.

The more associated conversions, the more often the source is used as an auxiliary one. Please review this report before disabling any channels or campaigns. They may not bring direct conversions, but they are important stage on the way to completing an order.

Mobile Devices Report

Allows you to analyze the amount of traffic from mobile devices and determine what share of total traffic it compiles and also compares the conversion rates on them. We have articles from eLama and Molinos on why it's important to work with these reports.

In the “Audience” section, find the “Technology” tab. In it, select “Mobile devices”. The “Review” item will show the ratio of desktop, smartphone and tablet indicators.

The “Devices” item displays indicators for specific models smartphones or tablets.

If you notice that your site receives a lot of traffic from certain devices with a high bounce rate, the pages may not be optimized for that device. It may also be that all traffic from mobile devices has a high percentage of failures. In this case, the site is most likely not optimized for mobile users or are not displayed correctly. The following report from our collection will help you check the site loading speed.

Page loading time

The report is needed to understand how long it takes to open your site, find pages that take a long time to load, and fix the problem. It's especially important to track data for advertisers who receive the majority of their traffic from mobile devices. Users viewing the site on smartphones close the page if it does not load within a few seconds.

To view loading speed data, open the list of “Behavior” reports, then “Site Load Speed” - “Page Load Time”. As in previous reports, add the Paid Traffic segment.

Add the “Mobile Device Traffic” segment to your report to find out how quickly pages load for mobile users. By analogy, you can look at statistics on users of desktops and tablets.

In the same list you can find tips on increasing your download speed. Path to the report: “Behavior” - “Site loading speed” - “Loading acceleration”. Segment - “Paid traffic”. To the right of each page you'll see PageSpeed suggestions. Click on them and you will see recommendations on how to optimize your site and increase its loading speed.

Before we understand what custom channel groups are in Google Analytics, let’s look again at two basic terms in web analytics - source And channel.

Source– the last resource that the user visited before moving on to your content. For example, through the Yandex search engine or by advertisement paid advertising in Google AdWords. Channel– source type. For example, organic search (organic), paid advertising (cpc), transition from another site (referral), etc.

Google Analytics comes with 9 standard channels by default:

Direct- direct traffic (transition to the site through direct entry of the domain into address bar browser or through bookmarks and saved pages);
Organic Search- search traffic from organic results;
Paid Search– search traffic from paid channels(contextual advertising);
Social- traffic from social networks;
Email- traffic from email newsletters;
Affiliates- traffic from partner sites;
Referral- traffic of transitions from sites;
Display– media/banner advertising;
Other Advertising– traffic that the system could not recognize (other in reports).

All of these channels make up a channel group, which in Analytics is called the default channel group. « Default Channel Grouping» . You can see what it looks like at the presentation level in the section “Channel Settings – Channel Group”.

Default Channel Grouping

By clicking on the group name, inside you will see all the available rules and labels that will be displayed in reports. The color symbol is used in multi-channel funnel reports.

In Google Analytics, standard channel groups are subject to rules defined by the system.

System defined channel

You can edit default groups, add new channels, remove existing channels, and change rule definitions. Edit by yourself system definition(System Defined Channel) is not possible, but you can override the parameters of its channels.

Google Analytics allows you to create channel groups in two ways:

creating your own channel group (at the user level);

Custom channel groups at user level

One user can have up to 100 of their own channel groups. When you create a new one, you can apply it to historical data.

Presentation Layer Channel Settings

You can create a maximum of 50 channel groups in a view. When you are given , creating a new channel group at the presentation layer is not always available. In this case, you can use the first method and create your own channel group, and then use it in multi-channel funnel reports and in "Traffic Sources - Overview" or “Traffic sources – All traffic – Channels.”

Any changes to the default channel group:

irreversibly change traffic classification in Google Analytics;
visible to all users of the view;
do not apply to historical traffic data.

The function is available in the channel settings "Brand name management". It allows you to analyze the effectiveness branded And unbranded paid advertising, as well as various keywords to get conversions.

As an example, consider:

creating your own group of channels, which will include data on mailings from the Getresponse service and advertising in Yandex.Direct and Google AdWords;
copying the default channel group. We will translate all source names into Russian to make it easier to navigate the reports;
creating channels using brand name management.

Creating your own channel group

Let's go to the right presentation and section "Own channel groups (BETA)» . Click +NEW GROUP OF CHANNELS.

We enter the name and define our channels by rules.

Defining your own rules

In our example, these are simple parameters "Source or Channel" and type exact match or contains(as in the screenshot above) for Getresponse email newsletters that have their own .

Selecting the displayed color for reports on multi-channel funnels and in the graph View we see final result. Similarly, we create two other channels for Yandex.Direct and Google AdWords. When defining a rule we can add additional condition OR/AND.

Additional conditions displayed channel color

By moving rules, you can customize the order in which they are applied. Rules are processed in the order they are listed; Each channel group must contain at least one label. All channels not included in the first rule (i.e., the first label) are included in the second, if one is specified. Channels that are not included in the second rule (i.e., the second label) are included in the third, if one is specified. All channels remaining after applying all the rules are included in the label Other (Other) .

Let's save our group. As a result, we received a custom group of channels with the ability to change, copy, delete, promote and import into the gallery Google solutions Analytics. When you send your own channel group, only the configuration is published. The confidentiality of your data will not be compromised.

Actions with channel groups

Promotion is an opportunity to provide access to own group channels at the presentation level, that is, if you open access, a copy of this group of channels will be created in the section "Channel Settings".

Accessing a channel group at the presentation level

Now we can view data on the created channels in reports. Let's go to “Traffic Sources – All Traffic – Channels” and select our custom channel group as the main parameter.

"Traffic sources - All traffic - Channel"

In reports "Associated Conversions" And "Main conversion paths" section "Conversions" initially a group of channels is used for multi-channel sequences default. To see data for your own channel group, simply change the main parameter.

“Top Conversion Paths” report

Copy default channel group

Now let's look at an example of copying an existing channel group « Default Channel Grouping» with renaming of the main sources. To do this, at the presentation level, go to “Channel settings – Channel group”.

Copying a channel group

Enter the name (I have this Channel groupRU) and when selecting a channel, determined by the system, Google Analytics itself shows us matches in Russian. All that remains is to simply manually rename 9 fields. You can change the displayed colors if you wish.

Russification of a group of channels

"Traffic Sources - Overview" report

Brand name management

By using branded And unbranded channels, you can more accurately analyze the effectiveness of advertising campaigns for high or low value CTR, cost per click, and conversions that occurred from the last interaction or associated conversions.

To create such a group, go to the presentation level in “Channel Settings – Manage Brand Names”.

Channel Settings - Manage Brand Names

include brand names with spelling errors in the list;

For example, in addition to google, include the names gogole and googel.

don't add various options capitalize as this function is not case sensitive.

For example, Google titles and google are treated the same.

Do not add plural forms because this function uses the collation contains. It means that branded Every request that contains the brand name anywhere in it will be recognized.

For example, the name adword corresponds to the terms adwords and google adword.

In field Indicate brand names enter one of the spelling options on each line, and then add them to the active list using the button Add brand names.

Adding brand names

After saving we will be asked to create channels. Click Yes, set up now.

Setting up channels for brands

We are notified that the channels branded And unbranded advertisements are created automatically.

Channels created automatically

Branded Paid Search

and the rest ( Generic Paid Search):

Generic Paid Search

Note: new channels Branded Paid Search And Generic Paid Search created in the default channel group Default Channel Grouping. And Google warns us from above that any change to the default channel group for presentation leads to an irreversible change in the classification model for new traffic. The statistics are preserved. To change how data appears in reports without affecting the data itself, create new group channels.

Google Alert

Therefore it is better to create separate group channels and add two new traffic there (branded and all the rest). After 24 hours in Google reports Analytics, you will see the division of traffic into channels with new names.

Vk.com -

In this article, briefly, about basic concepts and principles of operation of Yandex.Metrica and Google Analytics. And also about the differences between these two systems. I pointed to simple examples How to quickly find out basic data about the site’s operation.

What this article doesn't provide is recommendations for setting up analytics, or technical issues its functioning. First of all, I wanted to give people who are far from web analytics a minimal necessary base to understand basic terms. After reading the article, you will be able to clearly check the contractor’s analytical reports and quickly check the basic parameters of the site’s operation yourself.

Basic Concepts

Let's start with the basic analytics parameters. In Yandex Metrics or Google Analytics reports, you will regularly encounter concepts such as “visits”, “visitors”, “views”, “bounces”, “view depth”, “goal achievements”, “conversions”. Let's figure out what all this means.

But we will start with the concept of traffic.

Traffic- this is the flow of visitors to your site.

It is a flow, like the flow of a river. When they say it's coming high traffic“or about “traffic growth” - it means that a lot of visitors come to the site, or their number has increased noticeably. If they say that there is low-quality traffic, it means that visitors do not perform the necessary targeted actions, or the percentage of erroneous transitions to the site (bounce rate) is greatly overestimated.

Views (pageviews, hit) is called opening or updating (F5) any one page of the site.

A person came to your site - 1 view. Spins the mouse wheel and scrolls this page up and down - still 1 view. Pressed F5 on the keyboard or “Refresh” in the browser - this will already be 2 views. I clicked on the menu and went to another page of your site - that’s 3 views. Clicked on external link and went to another site - still 3 views.

In order for the analytics system to know about the occurrence of an event, a site element (button, link, form) must using JavaScript-code to send a message to the counter about the occurrence of such an event. That is, for each event, the necessary “triggers” must be written in the site code, which will fire when the visitor performs the desired action.

Visit(session, session) begins with a visit to your site and includes all views and events on your site during a certain period of time. This is the main accounting unit to which the analytics system links most of the operational data about what is happening on the site: where the visitor came to the site, what pages he looked at, what he clicked on - all this is tied to the visit.

As long as a visitor interacts with your site's content in some way, all views and events are recorded in one visit. The visit ends when no new events are received from the visitor within 30 minutes. You can change the amount of inactivity before ending your session. This is necessary in cases where the site has long articles or videos that take more than 30 minutes to view. With a standard timeout of 30 minutes, if a visitor closes the site but returns to it after 20 minutes, all new views and actions will be added to the previous visit. The exception is transitions from advertising. With each new transition from an advertisement - regardless of how much time has passed since the last activity of this visitor on the site - a new visit will be recorded.

In each browser, when you visit the site, it is saved individual identifier in the form of a cookie file. Using this file, analytics systems recognize the visitor and distinguish one visitor from another.

Visitor- this is not some kind certain person, and the browser. In Google Analytics, a visitor is called a User.

If the same person accessed the site from the same computer, but used different browsers- analytics systems will consider it as two different visitors. The same picture will happen if a person visited the site with different devices, for example, first from a computer and then from a tablet. Therefore, it is important to understand thatthe number of visitors in analytics reports is always more quantity real people who visited the site.

Important traffic indicators

How Yandex and Google count refusals

Almost all Metrics and Analytics reports contain an important indicator -bounce rate(bounce rate) in Analytics or simplyfailuresin Metrica. This is one of the main criteria for assessing the quality of traffic coming to the site. Although they are called almost the same, these indicators are calculated completely differently:

Yandex.Metrica considers a refusal a visit that lasted less than 15 seconds and during which time only one view occurred. The person went to the site, realized that this was not what he was looking for, and left the site.
Analytics considers visiting only one page as a refusal, without performing any other actions. The time spent on the page does not matter. Do you feel the difference? If you have a one-page site, then Analytics will consider the lion's share of visitors to be bounced.

This feature of Analytics cannot be changed. But you can create an event using JavaScript, which will poll the visitor’s browser after 15 seconds and send the event to Analytics. In this case, the bounce rate will work more correctly.

What is viewing depth

The second important indicator of traffic quality isviewing depth. It shows the number of page views in one visit. In summary reports, the viewing depth will show the ratio of the number of visits to the number of views.

If your site consists of only one page, then it would be logical that the viewing depth would always be equal to 1. But in practice, you will most likely see values of 1.1 – 1.3. Don't be surprised - this means that one of the visitors clicked on refresh (F5) of the page.

By the way, with a one-page site, the viewing depth value tending to 2 may indirectly indicate some problems on the site. For some reason, visitors often refresh your website page. On multi-page sites, this indicator, together with the residence time, can be used to judge the interest of visitors in the content of your site, i.e., the quality of traffic.

Conversion,Literally, it is a transition from one state to another. In the context of analytics, conversion is the execution of some targeted actions. Conversions are usually divided into two types:

Macro conversions- this is the fulfillment of goals that affect your income. For an online store, this could be placing a purchase or paying for it. For a landing page, this could be sending contact information, or a request for a service. For news portal it could be viewing a certain number of pages.
Micro conversionsindicate the visitor’s interest and involvement in the site materials. For an online store, this could be downloading a price list, viewing a page with delivery information, or viewing a specific group of products. For a landing page, this could be scrolling to the end of the page or watching a video on the site. For a news portal, this could be scrolling to the bottom of the page or subscribing to news.

When we build some reports, we want to understand how some page of the site works. And get some information about visitors and their actions. What we want to know about is usually calledparameter. A parameter is any non-numeric data. For example, in the screenshot below, incoming traffic channels are selected as the main parameter.

The data that we are interested in learning about the selected parameter is calledmetric(in Yandex Metrica) orindicator(in Google Analytics). Metrics contain numeric values. In the screenshot below, “Conversions”, “Goal Achievements”, “Targeted Visits”, “Targeted Visitors” are selected as metrics, and the parameter is countries.

How analytics systems collect data

In order for Yandex Metrica or Google Analytics to “see” your visitors, you need to enter the Metrica or Analytics counter code into the HTML code of the site. You can enter the counters of both systems; they will not interfere with each other’s operation. The counter code is generated by the analytics system after registration and basic settings. Registration and further work with analytics in Yandex and Google are free.

Data collection

When a visitor lands on your site, the browser downloads the site code and displays the page. Together with the site code, the analytics counter loads and performs its work. It records information about the browser ID from cookies, as well as the address of the site from which the visitor came to the page, and the type of transition: a transition from search results, or from an advertisement, or from social networks, or a direct visit to the site (the visitor entered the site address manually or called it from “Favorites”). It is calledtransition source And source type (channel).

In reports by type of referral source, you can see where more visitors come to you from - from social networks, from advertising, or from search. And based on this data, decide which direction of website promotion needs to be improved - promotion in search engines or contextual advertising, or maybe it’s worth developing groups on social networks.

The counter also collects information about all the visitor’s actions on the site and counts the time spent on each page. And Yandex Metrica can even record mouse movements and scrolling with the wheel and then show a video of the visitor’s actions on the site in Webvisor. This is very useful when you need to analyze the usability of a website. In addition, counters collect technical information about the visitor's browser, operating system, screen resolution, etc. This information can be useful when analyzing any problems on the site.

The counter links all data to unique identifier visitor (i.e., his browser) and sends it to the analytics system servers. Next, the system collects chains of events from one visitor and forms visits from them. A visit in Metrica or a session in Analytics is a basic accounting unit.

Data processing and reporting

The next major difference between the Yandex and Google analytics systems is the structure of the account and the processing of data received from counters.

A Google Analytics account has a hierarchical tree structure and can combine several resources with different counters. Within each resource there can be several so-called “views”, i.e., options for customizing and filtering data from the site.

Once you create an Analytics account, you will have one property and one All Data view. It is highly undesirable to touch this representation. If it is necessary to make any changes to basic settings representation, it is better to create a new one and make all the changes to it.

Important! All changes in settings, data filtering, creating new goals or changing current ones only affect the new data received. Data received before the change will not be recalculated.

Before saving, when the Analytics counter transmits data from the site to the system servers, the data is divided according to the number of views available for this resource. And they are processed in accordance with the filters and settings of each of the views. And only after that they are saved on Google servers. That is, all filtered data is simply discarded. And by changing the view settings, return the data to old look it is forbidden. That is why it is always necessary to leave the “All Data” view unchanged. This is insurance in case it turns out that the production view is not configured correctly.

There are no views in Yandex Metrica. Setting up the counter, granting access to other users, setting goals and filtering data - affect the entire counter. When processing data coming from the meter, filters and processors are also first applied, and then the data is stored on servers. Changes made immediately affect all newly received data.

How to analyze information

If you have both analytics counters installed, regularAlways conduct analysis using only one system. Google and Yandex use different information processing algorithms. And the indicator values will always differ between them. Take it for granted. Therefore, all your contractors, developers, designers, sales people and marketers, and you yourself, must analyze the work using one analytics system. The other can be used to solve isolated irregular issues when the capabilities of the first system are not enough.

My second recommendation is thispay attention to trends, not the accuracy of numbers.There are situations when the metric values for the same parameter in different reports are slightly different. But there is nothing strange about this, since each report is calculated using different algorithms. And for us, to analyze and build hypotheses, what is important is not the absolute accuracy of the numbers, but the dynamics, trends in changes in key indicators and, ultimately, correct understanding situation and its causes.

Therefore, it is much more important for us to set up all the necessary goals from the very beginning in order to be able to take measurements according to a variety of criteria.

Working with reports

In order to view reports in Yandex Metrica, you need to go to the Metrica website by the address and provide your login details. A list of counters for this account will open.

By clicking on the desired counter, you will be taken to the main Metrica interface. By default, the Summary report opens.

Menu
Counter name and number
Select the desired time period
Buttons for adding widgets to the Summary
Blocks of the widgets themselves
Setting up widgets
The second setting option is the ability to drag a block to another location by grabbing a small square with the mouse

Explore the widgets in the Summary and Widget Library and arrange them in the order you want. And you can, literally in a few seconds, receive from the Summary the basic information you need about the operation of the site.

To work with Google Analyticsfollow this link and enter your username and password. You are immediately taken to home page Analytics of the main resource and default views.

Name of the resource and current view
Menu with report blocks
"Administrator" button to enter setup mode
Brief summaries from different report blocks
Link to go to full version this report
Suggestions for improving analytics settings
Access to email notification settings and help

Another difference between Metrica and Analytics is that Google first processes the received data and brings it to a certain type. Because of this, there may be a delay from 2 hours to a day between what is happening on the site and the display of data in the report, depending on the report and the volume of data. In Yandex Metrica, data is recorded in the database as it is, “raw”, so to speak, and their processing occurs at the time the report is generated. The difference between what is happening on the site and the report data usually does not exceed 10 minutes.

When working with small volumes data in Metrica, the processing delay is minimal and invisible to the analyst. When the volume of data begins to noticeably affect the speed of report generation, Metrica turns on the sampling mode, processing not all the data, but 1/10. The report appears new element"Accuracy" control. After generating the report, the accuracy can be either increased or decreased.

Main reports

In Google Analytics, to monitor the current situation on the site and check changes made, there is a separate block of reports called “Real-time reports" Using these reports, you can monitor the main parameters of current site traffic, instantly check how recent design or advertising changes have affected traffic, etc. But, the final data in the main reports may differ slightly from the real-time reports.

These reports include:

an overview report with brief summaries from other reports;
“Locations” - the geography of transitions to your website, the most active regions in the last half hour;
“Traffic sources” - dynamics of changes in traffic from different channels and types of traffic (paid, free, search, advertising, social networks, e-mail newsletters, direct visits);
“Content” - the number of visitors to the most popular pages in the last 30 minutes;
“Events” - all triggers on the page in the last 30 minutes;
“Conversions” - goals achieved in the last half hour.

Site traffic.In Yandex Metrica, in the left menu, select “Reports” - “ Standard reports" - "Attendance".

This is the main report for tracking website traffic dynamics. By changing the date range for the data sample, you can understand on what days and at what time of day the site has peaks in traffic. On what days or hours are there the least number of visitors, how does the dynamics of visits change throughout the year, how the overall attendance was affected by any changes on the site or in the advertising policy. You can also filter the data by selecting one of the goals and analyzing the traffic of only that group of people who have achieved the selected goal on the site.

In Google Analytics, to view site traffic over a period of time, you need to select “Audiences” - “Overview” (1) in the left menu and select the desired time range (2). By changing the details of the schedule (3), you can see changes in the dynamics of attendance by hour, day, week or month. Below graph is shown general Statistics(4) activity and quality of traffic for the selected period of time - the total number of visitors for the selected period, the percentage of new and returning users, the average number of sessions per visitor, the average number of pages viewed per session, the average time spent on the site and bounce rate.

By default, the schedule is built for site visitors. But you can choose another basic indicator (5), for example, “Sessions”. And add an additional indicator (6) to the chart, for example, “Average duration of sessions.” This will give you the opportunity to analyze the data in more depth. At first we saw that on December 9 there was a decline in traffic on the site and, probably, it was a bad day for the site,

Adding the session duration indicator shows that the average duration of sessions increased noticeably on this day. On December 9, fewer visitors came to the site, but they were more interested in the site's materials than visitors on other days. That is, the quality of traffic on this day was better than on other days.

Report by channel/traffic sources

Allows you to find out where more visitors come to you from, or from which sources more applications/purchases are made. In Yandex Metrica, this report is called from the general menu “Reports” - “Standard reports” - “Sources” - “Sources, Summary” (1). By changing the report time period (2) and the level of detail (3), you can study in detail the peaks and valleys of traffic for each channel.

In the table below the graph you can see detailed data for each channel.

By clicking on “Select a goal” (1), you can select the goal of interest and leave data only for those visits in which the goal was achieved. This way, it will be clear which channel brings in more orders or sales. Using the “Metrics” button (2) you can add traffic indicators that interest you.

In Google Analytics, for this task, the “Channels” report from the “Traffic Sources” group is used.

"All traffic" (1). When you select a date range (2), you can immediately compare the indicators with the previous period (3).

In the table, the data is sorted by default by the “Default group channel” parameter (1). For a more detailed analysis, you can basic parameter switch to “Source or channel” (2) and view data for each traffic source separately. To obtain data for a specific goal, you need to select it in the “Conversions” list (3).

Channels in Google Analytics

Organic Search - organic search, transition to a site from search results in Yandex, Google, Rambler or other search engines.
Referral - referral traffic, transitions from other sites following links to your resource.
Direct - direct visits when the visitor entered the site address in the browser or called it up from Favorites.
Social - traffic from social networks.
Paid Search - paid traffic from search engines, contextual advertising.
Display - paid media traffic.

View search queries. In Google Analytics, to view the search queries that lead to your site, you must first link the analytics system to the SEO service Search Console. Yandex and Google encrypt almost all user requests to maintain information confidentiality. Therefore, you won’t be able to view these queries directly in Analytics. If the connection to the console is configured, the request report will be located at the address “Traffic Sources” - “Search Console” - “Requests” (1). The main parameter is a list of search queries (2). It is sorted by the number of transitions to the site (3) from each request. The CTR indicator (4) shows the “click-through rate” of a request - this is the ratio of the number of clicks to the number of impressions, multiplied by 100%. The lower this indicator, the worse the query performs. That is, users often ask this question in a search engine, but rarely go to your site from this request. You need to consider whether this request is important for your business, and whether the material on the site on this issue needs to be further strengthened. Another important indicator is the average position (5). It tells you where your site appears for that query.

In Yandex Metrica this report is located along the path: “Reports” - “Standard reports” - “Sources” - “ Search queries" In this report standard set metrics - visits, visitors, bounce rate, viewing depth and time on the site. Therefore, it will not be possible to perform such an analysis of the effectiveness of queries as in Analytics. To do this you will have to use a separate service.Yandex.Webmaster .

View popular site pages.This report will help you understand which materials or products on the site attract more attention from visitors, and which ones need to be improved or, perhaps, removed from the site altogether.

In Yandex Metrica, this report is located in “Reports” - “Standard reports” - “Content” - “Popular” (1). You can see the popularity of each page, and also increase the depth of analysis by adding using “Grouping” (2) Additional information. For example, referral source information or demographic information.

Google Analytics offers more information to analyze popular pages. Open the report “Behavior” - “Site Content” - “All Pages” (1). In addition to the number of views on each page (2), we can estimate the average time spent on each page (3). And also, two important indicators will help us better understand visitor engagement and behavior. The “Entrances” indicator (4) will tell you how many visitors started their visit from this page. And the indicator “Percentage of exits” (5) will indicate how many sessions ended with viewing this particular page. That is, the user looked at this page and left the site. For a more detailed analysis, you can add an additional parameter (6), for example, traffic source. And see which pages people go to most often from blogs and which ones from searches.

Let's consider reports on advertising campaigns. In order for these reports to work, automatic link marking must be enabled in Yandex Direct and Google Adwords. And in Google it is necessary to additionally link the advertising system with the main working view in Analytics. At the same time, it will be possible to view reports on Direct campaigns in Yandex Metrica, and Adwords campaigns in Analytics.

In Google Analytics, open the report “Traffic Sources” - “AdWords” - “Campaigns” (1). In this report you can see the basic data for each campaign - the number of transitions, the spent budget for the selected period of time, average price per campaign click, campaign bounce rate, average view depth (pages/session), and number of completions of the selected goal. Based on these data, it is already possible to draw conclusions about the effectiveness of each campaign - how much money the campaign spent and achieved the desired goals. To view information on the groups of a particular campaign, you can click on the name of the campaign in the table (2). Information on the effectiveness of each group in this campaign will open. If you click on the group name, you will see a list of keywords for this campaign that were clicked on from advertisements. If it is more convenient for you to see all campaigns and groups on one sheet, you can enable the display of groups with an additional parameter (3).

“Reports” - “Standard reports” - “Sources” - “Direct, summary” (1). Here, the main criteria for assessing the effectiveness of campaigns are the bounce rate, viewing depth and time on the site. By analyzing these indicators, you can evaluate the effectiveness of each campaign. To view data on specific announcements And keywords You can expand each line using the “+” sign to the left of the name (2). Extra options for analysis are added using the “Groups” button (3). Additional indicators - click on the “Metrics” button (4). You can view the achievements of the desired goal for each campaign using the “Select goal” list (5).

You can view information on expenses by campaign using the “Direct, expenses” report in the same group of reports. The amount spent for each campaign for the selected period of time will be added to the standard indicators.

However, I recommend using information about the amounts spent - both in Analytics and in Metrics - only for a qualitative assessment of the performance of campaigns. And keep the exact balance of finances according to Yandex Direct and Google AdWords. The fact is that analytics systems and advertising platforms count transitions to the site using different algorithms. The advertising platform cares about the quality and effectiveness of advertising and looks at user actions after visiting the site. If she considers some transition suspicious and regards it as a “click” advertising budget competitors, the transition will not be counted and the money for it will be returned to the advertising campaign account. And the analytics system will simply count all visits to the site.

What to look for in reports

Any analysis makes sense only when a sufficient amount of statistical data has been collected. The more data, the more accurate the analysis. If you have just launched a website and you have 30-50 visitors, this is too small a number to make any decisions about the effectiveness of the site, advertising, or the quality of traffic. After running an advertisement or making some change on the site, you need to wait at least 300-500 visits before analyzing the results of this change.

Key indicators are the number of visitors and visits (sessions), viewing depth, bounce rate, session duration, achievement of macro goals. Analyze separately the visitors who achieved the macro goal and the entire traffic flow. Compare performance over different periods of time.

To understand whether your metrics are good or not, it is best to compare different traffic sources.

To begin with, you can take organic search traffic as a benchmark and compare how traffic from other sources differs from organic traffic. You should not take direct traffic as a standard, because these are visitors who are already familiar with your site and come to solve some specific issues. The time spent on the site and the depth of viewing should grow over time to some maximum, and the bounce rate should decrease to its minimum - it still won’t be possible to reduce bounces to zero.

We got acquainted with the basic concepts of web analytics, learned how to work with basic reports and obtain data on various parameters site and its visitors.

I wish everyone good traffic and high conversion!