Hello IdeaFox friends!

Today I will tell you about what can harm young sites in terms of promotion. I strongly advise those who have relatively little traffic to read this article, since we are talking about refspammers who can have a bad impact on the development of your blogs.

Let's get started.

I constantly see spikes in referrals from junk sites + suspicious activity on my websites. Moreover, these are not transitions of living people, but the so-called “refspam” that is generated by villains.

This problem looks like this in Google Analytics statistics:

That is, literally for 5-10 minutes there is a sharp increase in “attendance” from which there is no benefit other than irritation.

Let me emphasize once again that these are not real people, but fake transitions that actually do not exist. This infection is called “referral spam”.

I have suspicions that such transitions may indirectly influence the behavioral factors of the site. That is, search engines may imagine that this is a cheat of behavioral sites, and this may be followed by a filter from Yandex or Google.

Especially if site traffic is small (100-200 people per day), since such activity greatly distorts behavioral factors and looks like gross cheating in statistics behavioral factors.

Why is this being done?

It is clear that if the webmaster sees a large number of transitions, then out of curiosity he will go to see who is giving him such powerful traffic.

Secondly, such transitions can generate fake user “activity”

Look here. One of my new sites has quite common behavioral factors:

1. Approximately 1.7-1.9 page views per person;
2. Bounce rate is approximately 7%;
3. Time on site is approximately 2 minutes 30 seconds.

1. 4.5 page views per user (twice as much);
2. At the same time, the failure rate increased to 9.5%;
3. The time spent on the site remained approximately the same.

There is a deterioration in behavioral factors. And if the attendance was not under 1500 people/day, then the behavioral factors would collapse.

That is, 2-3 “visitors” came from such garbage sites and began to increase the number of page views. Approximately 1000 views per IP address (!!!). And in general, such activity is similar to a light DDOS than to PF cheating.

Again, I suspect that this kind of fake activity could have a negative impact on behavioral factors (BFs) if it goes on for long enough. Yes, such traffic does not significantly affect sites with high traffic. But for young sites with little traffic it can be dangerous.

How to exclude such bots from Yandex.Metrica?

First, let’s cut out suspicious activity at the Yandex.Metrica level.

Go to the counter settings and set the settings as in the screenshot above.

1. Filter type: “Keep traffic only”;
2. Field " Page URL”;
3. Condition “Site and mirrors”;
4. Filtering robots “Filter robots by strict rules and behavior”;
5. You can check the “Do not count my attendance” checkbox. But I never install it in order to generate more traffic =) (I’m kidding, of course).

After I made these settings in Yandex.Metrica, abnormal jumps in behavioral factors were no longer visible, although there was garbage traffic. It was clearly visible from the data Google Analytics.

How to block such activity at the site level?

There are ways to remove Google statistics Analytics similar activity, which I will talk about in one of the following articles (subscribe to blog updates). It is quite complex and worthy of a separate large article.

But even if you remove suspicious activity from Metrica and Google Analytics reports, this does not mean that it will disappear. You understand that if your blog or website receives several tens of thousands of views every day, then sooner or later this may cause problems.

But there is a much simpler method that allows you to proactively ban refspam and an abnormally large number of page views

We will use the WordFence Security security plugin, which allows you to create such protection in literally 5 minutes.

I talked about how to set up WordFence Security in this article:

True, this article is quite old, and the plugin itself has changed significantly over the past two years. But in paid course By WordPress security there are more latest information according to its setup:

How to ban IPs from which abnormal activity occurs?

For example, you see that an abnormally high number of requests are coming from one IP address or the number of pages viewed is off the charts. Such IP addresses can be seen in the server or hosting logs.

But I suspect that the vast majority of bloggers will faint when they hear the phrase “look at the server logs” =)

But in WordFence Security this is clearly visible. For example, here’s some handsome guy hammering a site with 80,000 requests:

How to block such an IP address?

WordFence Security makes this very easy. Let's assume you know that there is suspicious activity coming from a certain IP address. For example, the bot hung on the site for 5 hours and looked at 100,000 pages during this time...

Go to the settings of this wonderful security plugin and go to the “Blocked IPs” menu item

We drive in the scoundrel’s IP address and block him forever.

Everything is fine, but you need to regularly look at reports (or logs), and then manually ban the IP address. Of course, dreamy bloggers won't do this =)

How to automatically ban such IPs?

Before we go further, let me emphasize that you have WordFence Security must be configured correctly. Particularly important correct setting WFS, if iThemes Security or another security plugin is running in parallel.

And once again I refer you to this article, where I talked about the basic settings:

So here it is. For such purposes, WordFence Security has a wonderful tool that does this dirty work in automatic mode, without the participation of the webmaster.

Go to WFS settings:

Turn on FireWall

And we configure the FireWall itself as in the picture below:

What do these settings mean?

1. If anyone's requests exceed: if more than 480 requests to the site are recorded in one minute, then such an IP will be blocked;
2. If a crawler's page views exceed: if more than 480 page views are recorded in one minute from bots, then such an IP will be blocked;
3. If a crawler's pages not found(404s) exceed: if 120,404 errors are generated per minute from any IP (an attempt to view a non-existent page), then such IP will be blocked;
4. If a human's page views exceed: if a person views 120 pages of a site per minute, then his IP is blocked;
5. If a human’s pages not found (404s) exceed: if a person tries to view a non-existent website page 120 times a minute, then he is blocked;
6. If 404’s for known vulnerable URL’s exceed: By analogy with the previous point, we set 120 hits per minute;
7. How long is an IP address blocked when it breaks a rule: for how long to block IP. In my case, it's one hour;

That's all, actually. Now any IP address that exceeds the specified limits will be blocked for an hour. Or for a day, as you configure.

Important: do not set too fanatical rules, as there is a risk of blocking search robots Yandex or Google.

But let's get back to our sheep =)

That is, to refspammers who generate fake referrals from garbage sites using bots.

Suppose you notice this beauty:

At first we were happy, but then we realized it was refspam. It is USELESS to ban these sites by IP address. They regularly change their IPs and such blocking is of little use.

But even in this case, WordFence Security will save us.

It's easier to show with a specific example how it works.

Let's say I don't like transitions from my blog site to my test site, win4blog.ru. And I want to block such traffic. Go to the WFS settings on the website win4blog.ru in “Advanced Blocking” and in the field “Referer (website visitor arrived from) that matches”

I enter the name of my blog like this:

Of course, you CANNOT register your blog on your own blog =)

By the way, you must indicate the reason for blocking in the “Enter a reason you’re blocking this visitor pattern” field in English letters.

For example, something like “blokirovka”. And click on the big blue button =)

You can see for yourself =)

Instead of a total

I advise you to immediately block refspammers as soon as you notice them in your statistics, so as not to distort the PF data collected by search engines.

Yes, representatives search engines they write that they are able to detect such traffic and not take it into account in general statistics. But you know that it’s better to play it safe, especially at the very beginning of site development.

And further important point. What I talked about will not help fight off a serious DDOS attack. But from automatic scripts who are looking for vulnerabilities, refspammers and petty online hooligans - will help.

If something serious happens, you will have to connect a CDN and other fail2bans, which I may talk about later.

To fully assess the impact of your website, you need to get the most accurate analytical results possible. Let's take conversion as an example.

Let's say a hundred users come to you every day, which you will find out using metric data. This is three thousand users per month. Of these, 450 people ordered the product. We calculate conversion by dividing users who performed the target action by the total number of visitors and multiplying by 100%. That is, 450:3000*100%, that’s 15%. However, you are also among the total number of visitors, because you also visit the site to check its performance. Therefore, the numbers you calculated earlier will not be absolutely correct.

That is, to correctly calculate the conversion, you need to subtract the visits that you and your employees make, that is, not the target audience. Let's say there are 400 such visits in total. Then we calculate: 450/(3000-400)*100% = 17%. That is, we cut off non-targeted visits and got a real conversion.

In order to exclude your own visits, you can cut them off from the general data. For this purpose it is provided special opportunity. Let's stop at this functionality more details.

We cut off unnecessary visits to Yandex. Metrics

First, you need to log in and go to the main page of the counter we need. Click “Settings”.

Now go to the “Filters” section.

Your visits will not be displayed, you just need to check the box next to “Do not count my visits.” For the changes to take effect, click the “Save” button.

The operation associated with excluding your visits is tied to cookies stored in the service’s memory. Now Yandex. The metric will not count you as a new user. However, if you clear your cookies, your new visit will be considered a new user's visit.

We’ve sorted out our own visits, now let’s talk about other possibilities of “Filters” that take into account the source data.

Using the specified functionality, you can filter the following data:

Visitor's IP address. This is relevant if not only you, but also your employees look at the site. You need to know the IP of their computers or the router that distributes to the department or office. Click “Add filter” and in the fields that open indicate:

Traffic type – Exclude traffic;

Field – IP address;

Condition – Contains;

Equal to or Starts with (depends on the number of addresses and the most optimal criterion for adding);

The value is the IP address itself or a chain of them. Set this value and click Save. Below is a good example.

You can remove the filter by simply clicking on the cross.

Page URL. With its help, it is possible to filter those website page addresses where analytics is not needed. Let's say you don't need data from a certain transition external source, since these transitions are rare, the traffic from them is not targeted. Here you need to pay attention to the Conditions field, which contains a number of sub-items: “Contains”, “Starts with...”, “Equal to”, “Only site and other domains.

If the first points do not raise questions, then the last one needs to be dealt with. After the specified value is set, the report will include only those visits to the web resource and its domains that were specified when editing in the “General” tab. Below is a visual example.

Page title. Using this filter, you can prevent some analytics data from being displayed, for example, some pages or groups of pages with a certain title. Installing this is similar to the case that we considered earlier. Let’s set all the parameters and don’t forget to click Save.

Disabling filters without deleting

Everything is simple here - just use the checkboxes that are placed opposite the filters that will not be used.

Let's summarize: filters created for Yandex Metrics make it possible to exclude your own visits, as well as not show sources and data that do not have of great importance, allowing you to focus on your target data.

Now let's talk about bots and how to cut them off.

We cut off bots in Yandex. Metrica

You can view the Robots report by completing next transition: Reports - Standard reports - Monitoring - Robots.

There is another way - use the search.

How to cut off search bots from the displayed data? Go to the “Filters” section. It was said earlier how to get there.

The last point at the bottom of the page deserves special attention. It’s called Robot Filtering and contains three sub-items:

Take into account the visits of all robots - in all reports that you receive, visits from bots will be noted;

Filter robots according to strict rules - selection according to specified patterns per IP address or group of addresses, or User-Agent. If the filter is enabled, there will be no “false” positives, and only a small part of the robots that perform indexing will be shown.

Filter robots only by strict rules and behavior - allows you to find the majority of search engine bots and show them in reports. This is relevant if you noticed a sharp surge in traffic that aroused your suspicion.

Speaking about the Filters section, I would like to go over the Operations item. Why is this option needed?

Firstly, it is needed to create strict/common rules that allow modification of the URLs of the pages that are shown in the reports. Rules are added and set according to the same principle as when adding filters. There are several operations, each with its own function.

Cut parameter - removes part of the page address. Useful when several pages have the same address, similar content, and need to be combined so that the report shows one page rather than two.

Cut all parameters- there is nothing to explain here;

Replace https protocol on http- using this rule in the displayed page report secret protocol connected to the regular one and shown as a single parameter;

Replace domain- replaces the domain in reports (page addresses);

Convert to lowercase- alphabetic characters in the URL are written in lowercase, excluding capital letters.

Operations are highly customizable so you can customize and optimize these metrics to suit your goals.

When setting up Filters and Operations, consider the following:

⦁ You can create no more than 60 Rules: 30 for filters and 30 for operations;

⦁ Changes in settings take effect only from the moment the options are added and saved. The old data will not change and will be displayed as before.

Final word

So, we see again that Yandex Metrica is a cool service that has great potential and huge opportunities. The system of filters and setting rules makes it possible to adjust analytics to your goals, choosing only what is needed to monitor and calculate the information data you need.

In August, Yandex updated questions not only for, but also for Yandex Metrica. The exam has become more difficult, new types of questions have been added, and their number has increased.

Today, the Yandex Metrica exam is:

✅ 50 questions in the test

✅ 6 topics to go through

✅ At least 80% correct answers in each topic

✅ 50 minutes to complete the exam

✅ 4 attempts to pass*

*If you failed to pass the exam the first time, then the second attempt will open a week later, and subsequent attempts will be made at intervals of 3 months, so you need to prepare for passing the exam.
Don't have time to take the exam and get a Metrics certificate on your own? We will do it for you, simply and find out how to pass the Yandex.Metrica exam using our website.

25 %

Answers to questions about Yandex.Metrica

1,200.00₽

Answering questions on Yandex Metrica allows you to successfully pass the exam in the Yandex.Expert system and become a certified specialist in your field. Updated database! docx format(WORD).

Buy now

Yandex.Metrica exam – multiple-choice questions

Theory: how Metrica processes data

1. Does Metrica have a standard report on robots?
Yes, in the Visitors section.
No.
Yes, in the Technologies section.
Yes, in the Monitoring section.

2. What type of filtering allows you to filter greatest number robot visits?
According to strict rules.
Through operations.
According to strict rules and behavior.
By referrer.

3. What are “new visitors” in Metrica?
Visitors who visited the site for the first time in the entire history of data collection during the reporting period.
Unique users who visited the site during the reporting period.
Visitors who visited the site for the first time this year.

4. The “Attendance” report has a column “Visitors”. What does this column show in the “Totals and Averages” row?

5. What is a grouping?
A set of Metrics reports.
A group of statistics objects of the same type.
A group of counters for one account.
A set of goals of the same type.

6. What does Metrica use to reduce the time for calculating indicators when loading reports?
Sampling.
Groups.
Informer.
Segmentation.

8. What are “visitors” in Metrica?
Users who made at least one visit to the site during the reporting period
Authorized Yandex users who visited the site during the reporting period
People who visited the site for the first time in the entire history of data collection during the reporting period

10. For what period does the meter store data?
For the last year.
Webvisor data is stored for 15 days; there are no other restrictions on data storage time.
There are no restrictions on data storage time.
For the entire period of work, provided that the rights to the site are confirmed to Yandex.Webmaster.

11. Is it possible to find out in Metrica the phone models from which visitors accessed the site?
No, in the reports you can only see data on mobile device manufacturers.
No, you can only find out the usage percentage of mobile and standard devices.
Yes, the phone model can be seen in the “Technologies - Devices” report.

12. What factor is most significant for recognizing visitors in Metrica?
Visitor behavior.
IP address.
Availability of JavaScript in the browser.
Visitor cookies.

13. The visitor entered the “Electric saws” section at 16:00. At 16:15 he went to the page with delivery conditions and then, without closing the browser, walked away from the computer. Returning at 16:30, the visitor opened the return conditions page and closed the site at 16:45. What data will Metrica record?
1 visit.
2 visits.
3 visits.

14. Link to the report. Dates: April 2018. Which gender visitors have the highest percentage of conversions for the “3+ hits” goal?
The conversion percentage is the same for both categories of users.
Female.
Male.

15. After what time from last action will the visitor's visit be considered completed?
In 30 minutes.
In 30 minutes. The inactivity time can be changed in the counter settings through the “Visit timeout” parameter.
After 30 minutes, if the user did not come through advertising systems. According to advertising systems, the visit ends after 60 minutes of inactivity.

16. The same Metrica counter is installed on the website site.ru and on the website webpage.ru. Arkady first went to site.ru, and after 10 minutes he moved from there to webpage.ru. Will a new visit be formed at the same time?
A new visit will not begin.
Yes, there will be a new visit with the “Sites” source.
Yes, there will be a new visit with the source “Internal Transitions”.

17. The site has a standard Metrics counter configured by default. During the day, 2 visitors came to the site, they made 4 visits lasting: 5, 10, 20 and 30 seconds. For each visit, one page of the site was viewed. Calculate your bounce rate per day.
0.5
0.75
0
1
0.25

18. On what basis is information collected in the Demographics group reports?
Data obtained using Crypt technology.
Social network data.
Yandex user registration data.

19. The “Attendance” report has a column “Visitors”. What does this column show in the Totals and Averages row?
The number of visitors with a unique IP address who visited the site during the reporting period.
The sum of unique visitors for each day of the reporting period.
The number of visitors with a unique browser cookie who visited the site in the reporting period.
The number of people who visited the site for the first time in the entire history of data collection during the reporting period.

20. Why do I still see previously collected robot visit data in reports after applying robot filtering?
Filters are applied some time after activation.
Filtering only applies to data collected after the filter is activated.
After activation, it was necessary to update the counter code on the website.
Filtering can only be applied to non-standard reports.

21. The site was launched on Friday. On Saturday there were 50 visitors, on Sunday 70. On Monday you look at the “Attendance” report for Saturday and Sunday. How many visitors will be in the “Total and average” line in two days?
From 50 to 120 visitors.
60 visitors.
From 50 to 70 visitors.
From 70 to 120 visitors.
120 visitors.

22. What is the lifespan of cookies that Metrica uses to identify users?
90 days.
One day.
Two month.
Two years.

24. The visit timeout is set to 60 minutes. Visitor path: from Yandex Search to the main page, after 50 minutes - transition from the Direct ad to the product page, after 90 minutes - transition from Yandex Search to contacts. How many visits will there be in Metrica?
2
4
3
1

25. What is a visit timeout?
Sources that, when clicked, will start a new visit and end the old visit.
The time of visitor inactivity on the site, after which Metrica will consider the visit completed.
The number of pages viewed after which the visit will be considered completed.
Duration of the visit after which the visit will be considered completed.

26. How can I see robot visits in regular reports?
Robot visits cannot be seen in regular reports.
Apply segmentation to visits.
Apply segmentation by “For people who have – characteristics – roboticity – robots.”
Select the appropriate metric.

27. What filters can be used to filter robot visit data?
Take into account the visits of all robots.
Filter search engine robots.
Filter robots by strict rules and behavior.
Filter robots only according to strict rules.
?
29. What can be done to reduce report loading time?
Reduce accuracy.
Use segmentation.
Increase accuracy.
Take advantage of the offer to prepare a report.
30. Link to the report. Dates: April 2018. How many visitors returned to the site in April 2018 from those who first visited the site in March 2018?
2730
15957
2222
497
Link to the report. Dates: April 2018. How many page views have there been for pages whose URL contains “features”?
2489
733
510
612

Setting up analytical data collection
1. Under what logins can you delete the Metrica counter?
Under the login of the counter owner.
Under the representative's login.
Under a login with guest access for viewing only.
Under any login, if the “Public access to statistics” option is selected in the access settings.
Under a login with guest editing access.

2. Statistics on the counter are provided public access. Which reports are not accessible to unauthorized users?
Target call reports.
Goals.
Search phrase report.
Webvisor.
Reports via Direct.

3. Is it possible to change the Metrica counter number?
Yes, you can change your number yourself at any time.
No you can not.
Yes, upon request to the support service.

5. What counter code blocks need to be installed on the site for Metrica to correctly calculate and display site traffic data on the informer?
Only block – Yandex.Metrika counter –.
Both blocks of code.
Only block – Yandex.Metrika informer –.
Either of two blocks of code.

6. Clicking on a product on the website opens a gallery with photographs of the product from different angles. The opening of a gallery should be counted as a transition to separate page. Which counter method should I use?
hit() method.
notBounce() method.
trackLinks() method.

7. Is it necessary to create a new counter for each site?
Yes, definitely.
No, you can install an existing counter on other sites.
No, only if it's a subdomain.

8. Vasily enabled the counter settings option from the “Basic” section – “Including subdomains”. How will data be collected now?
Automatically collect from subdomains, even if the Metrics counter is not installed there.
The option filters the data: the reports will contain data only from domains from the “Site Address” and “Additional Addresses” blocks, as well as from all subdomains.
Automatically collect from subdomains if the Metrica counter is installed there.

9. Using which parameter can you check the correctness of the counter installation via a debug message in the console?
debug=1
from=test
_ym_debug=1
utm_source=test

10. Counter settings No. 33980360: This counter is installed on all pages specified in the answers. Select the pages that will appear in Metrics reports with the specified settings.
site.ru
site.com
site.ru
site.biz
site.rf

11. Counter settings No. 44147844: This counter is installed on all pages specified in the answers. Select the pages that will appear in Metrics reports with the specified settings.

metrika.yandex.com/new
metrika.yandex.com
metrika.yandex.com/about/new
metrika.yandex.com/about/2
metrika.yandex.ru

12. As in Ecommerce Metrics is called an event of type “View”?
choice
info
view
detail

13. Is it possible to set the currency value in E-commerce?
The currency value can only be changed by changing the counter currency.
You can pass the value of another currency.
This is not possible, all numbers are transmitted without currency.

14. Is it possible to combine data collected on two different Metrica counters in one report?
No you can not.
Yes, using counter labels.
Yes, Metrica automatically combines data from several counters if the same domain is specified in their settings.
Yes, if they are under the same login.

15. Which field in E-commerce can be used to transmit information about the type of product?
name
Such information cannot be transmitted via Electronic Commerce.
category
variant

16. Is it possible to transfer to Metrica additional attributes visits?
Yes, but only information about whether the visitor was authorized on the site.
Yes, you can. Names additional parameters must be predefined in the meter settings.
Yes, you can, but only when transferring e-commerce data.
Yes, you can using Visit Settings.
No, there is no such possibility.

17. Who can restore a counter that was deleted more than three months ago?
Login with editing access.
Owner.
Owner's representative.
Such a counter is permanently deleted and cannot be restored.

18. Under what logins can you edit counter settings?
Login with guest editing access.

Login of the counter owner.
Representative login.

19. The site has pages with addresses like http://site.ru/?order=by_size, http://site.ru/?sort=by_price. How to use operations to combine all these pages into a single URL - http://site.ru?
Add the “Cut Anchor” operation.
Add 2 operations: Cut parameter – page URL – “order” and “sort”.
Add 2 operations: Cut parameter – page URL – “by_size” and “by_price”.
Add 4 operations: Cut parameter – page URL – “order”, “sort”, “by_size” and “by_price”.

20. How will data be taken into account when installing one counter on several domains?
The data will be collected incorrectly. For correct operation counter for each domain, you need to create a separate counter.
The statistics will take into account visits to all domains.
Statistics will only take into account visits to the domain specified in the “Site Address” field when registering the counter.

21. What does connecting the Informer function do?
Ability to view visitors' actions on the site.
The ability to add a graphic block to the site that displays traffic data.
The ability to receive notifications about problems related to the operation of the site.

22. Do I need to specify a protocol for the site address when registering a meter?
It is necessary in any case.
No need.
Necessary if it is ftp.
It is necessary if it is https.

23. Is it possible to install several counters on one site?
It is possible, but the Webvisor only needs to be connected to one of the meters.
It is possible without any restrictions.
You can only install one counter per page.

24. Given filters, indicate what data will be filtered?

Views of any pages except the page with the title “test”.
All visits.
Page views with the title "test".
Visits from the meter owner.
All visits except visits from the meter owner.

25. What is the maximum number of visit parameters that can be passed in a visit?
1000
200
20
500

26. Which field applies only to the action “Purchase a product”?
quantity
actionField
products
variant

27. Which option must be selected in the counter settings to correctly count transitions to pages like site.ru/photo#1, site.ru/photo#2?
For XML sites.
Hash tracking in address bar browser.
Asynchronous code.

28. Will data be collected if the counter code is installed on the site, but the counter is deleted in the Metrica interface?
Data will be collected, but it will be possible to access it only after restoring the counter in the interface.
Data will continue to be collected as usual.
No data will be collected.

29. What needs to be done after creating a new Metrica counter so that data on the site begins to be collected?
Site traffic data will begin to be collected immediately, additional actions not required.
You need to set its code to home page site.
You need to install its code on all pages of the site.

30. The counter code contains the following information: try ( w.yaCounterXXXXXX = new Ya.Metrika(( id:XXXXXX, accurateTrackBounce:true )); ). How will such a counter count failures?
A refusal will be considered a visit during which the user visited one page of the site and viewed it for less than 15 seconds.
A refusal will be considered a visit during which the user visited one page of the site and viewed it for less than 5 seconds.
A refusal will be considered a visit during which the user visited one page of the site and viewed it for less than 5000 seconds.

31. Who can grant guest access to the meter?
Login with guest access to edit the counter.
Representative.
Login with guest access for viewing only.
Owner.

32. How does the “Accept data only from specified addresses” option work?
Allows you to collect data from all sites on which the current Metrica counter is installed.
Allows you to collect data only from specified traffic sources.
Allows you to collect data only from the site specified in the “Site Address” field and additional site addresses.
Allows you to protect data from spam if the counter is copied and installed on other sites.
33. What types of notifications are there in the Yandex.Metrica interface?
By email.
In the form of a phone call.
In the form of a message in instant messengers.
In the form of SMS.

34. Which container is used to store E-commerce data in Metrica by default?
layerData
eCommerce
dataLayer
mCommerce

35. What is the default currency used in E-commerce?
Units without currency specification.
Dollars.
Rubles.
Counter currency value.

36. You want to see in the reports which area of ​​the city visitors choose to deliver to when ordering. How can this be done?
Using target tracking.
Using the visit parameter with the appropriate value sent at the time of placing the order.
Using the visitor parameter with the appropriate value sent at the time of ordering.
Via the Popular report.

37. What activities can be tracked using E-Commerce? Select all appropriate titles.
detail – view the product card
add – add to cart
remove – removal from the cart
view – view the catalog page
purchase - purchase

38. Is it possible to independently transfer a counter from one login to another login in Yandex?
No, to do this you must contact support.
Yes, this option is available to the meter owner.
No you can not.

39. Is it possible to independently provide access to all counters installed on the login?
Yes, using the representative system.
No, you can only provide separate access to each counter.
No, to do this you need to contact support.

40. Will the counter collect data about visitors if you install only the informer code on the site pages?
The meter will not collect data.
The metric will only collect general data, but not data on targeted visits.
Data will be collected correctly.

41. How many days is it possible to restore a counter after it has been deleted?
15 days.
1 Week.
A deleted counter cannot be restored.
90 days.

42. Are statistics saved after transferring the counter to another login?
Yes, it is completely saved.
It persists, but only for the last three years.
No, it is not saved.

43. The web viewer is disabled in the counter settings. The owner of the counter turned on Webvisor to collect data across the entire site. Do I need to reinstall the counter code on the website?
No, the counter code will be automatically updated on the website.
Yes, you need to reinstall the counter code on all pages of the site.
You can reset the counter code only on the main page. Data will be collected in full.

44. In the counter settings, filtering of robots by strict rules and behavior is enabled. What happens to robot visit data?
They will disappear from all reports except the robot report.
They will disappear from all reports.
They will disappear from all reports except Webvisor and the robot report.

45. What does the message in the browser console about viewing Metrics look like?
PageView
Ya.Metrika
Hit
Reach Goal

46. ​​Which counter tag is loaded when the user has JavaScript enabled?

Our site uses cookies to improve the performance and efficiency of the site. By continuing to use the site, you agree to our use of cookies and privacy policy. OK

A few months ago, our Google Analytics bounce rate increased significantly. We have done standard set actions that are recommended to be done on the Internet: created a “no spiders and bots” view in analytics (the “Robot Filtering” setting in the view), checked the quality of the settings Analytics code, checked and adjusted the session duration, and so on. All this took time, but did not produce results. The failure rate on some days exceeded 90%. At the same time, the quality of the content on our website or the structure of incoming traffic did not change in any obvious way. It just “happened overnight” and that’s it. Since I couldn’t find anything similar described on the Internet, I decided to describe how we found and fixed the problem and reduced the failure rate to an acceptable 42-55%.

Here's a screenshot to illustrate the original problem:

Because everything standard circuits They didn’t give any results, we had to think for ourselves and look for the problem. Analytics didn’t help, so I started checking through Yandex.Metrica. The overall indicators according to Metrics were quite acceptable (up to 10% failures). After reading several articles about why failures in Metrica can be acceptable, but in Analytics they can go off scale, it became clear where to look for the problem. In short: Metrica considers all visits that lasted less than 15 seconds to be bounces, while Analytics considers all visits after which there were no other visits to the page. So I started looking at the report on the duration of visits in Metrica and Analytics and realized that it was unrealistic for me large percentage visits with a duration of 0:00 seconds, up to 50% of daily sessions. A few more articles I read allowed me to discard the hypothesis about the code not working and bots making their way through the Analytics filter. I also didn’t have any signs of referral spam.

In the end, I simply filtered out visits with a duration of 0:00 in Webvisor and decided to try to find a pattern. This is what I received:

Each “visitor” came from his own subnet, with an explicitly specified User Agent, screen resolution and operating system, that is, for Metrics and Analytics he was in no way perceived as a bot.

The only thing that gave it away was the rhythmic nature of visits every 1 hour 1 minute and 0 viewing duration. I sent the screenshots to our system administrator Andrey and asked him to look at what it was in the server logs. The first IP member made us surprised: someone from the League.Zakon subnet came to us.

I have an IP address for 4:56, now I’ll check other calls
inetnum: 193.150.7.0 - 193.150.7.255
netname: LIGA-UA-NET2
remarks: LIGA ZAKON

In total, the logs revealed 43 IP addresses that accessed different pages of our site from different providers, with different User Agents.

We didn’t find anything interesting on RIPE, unremarkable IPs from ordinary subnets. Apart from the fact that the bots walked at intervals of 1:01, there was nothing else that was obviously common between them.

We blocked the entire list found in iptables.

During the day, we also caught several new IPs, sketched out an algorithm for automatically filtering such bot traffic in case, after blocking some bots, new ones come to replace them. However, no one else like him was found. There were only a couple of curious User Agents, but no one with an interval of 1:01.

Another day later, the bounce rate according to Google Analytics began to return to normal and dropped sharply from 89% to 42.75%.

Today, almost a week after the events described, the failure rate remains within an acceptable range of 42-55%; the overall dynamics can be seen in the graph. Where there is a sharp decline, we have filtered out the bots.

We have only two hypotheses about “what it was.”

The first is that one of us configured some monitor bots incorrectly. At one time we were playing with different software to check the state of the server. They could have turned something on and forgotten. The downside of this theory is that I don’t remember a single service that would declare that they send requests from different subnets to different pages site from different User Agents. Therefore, this is most likely not the case.

The second hypothesis: this is some form of bot attack unknown to the general public, perhaps aimed precisely at increasing the bounce rate and, as a result, pessimization in the search engine. Google results.

If you have also encountered this, I would be glad to hear your comments. If you need more detailed instructions Write about how we looked for it and blocked it, too.

Filters are applied to the raw data collected by the meter in a specified order, leaving or removing records that meet certain conditions in the reports.

Changes caused by filters are irreversible. If there are no filters, all records will be included in the reports. The checkbox to the left of the filter allows you to disable it without deleting it.

Filters can be applied to the following types of data:

Visitor IP address

For example, if you select the filter type sequentially “Leave only traffic”, field “IP address”, condition “equal to” and value 192.168.0.2 - only transitions from this IP address will be included in the report.

in the interval

Examples of setting IP address intervals: 192.168.0.0/24; 192.168.*; 192.168.0.0–192.168.255.255.

Page URL

The filter type can be specified by the following conditions:

• contains
• begin with
• equals
• only site and additional domains - to leave in the report only page views of the site and its additional domains, which are indicated on the counter editing page (General tab).

Page title

The filter type can be specified by the following conditions:

• contains
• begin with
• equals

To prevent Yandex.Metrica from taking your visits into account, enable the option Don't count my visits. Using this setting, the service remembers the cookies of visitors who accessed reports or the Settings section of this counter, and does not take them into account when calculating statistics.

Operations

Operations are performed on the original data. They modify the page addresses of the reporting or referring site that are displayed in reports.

The following operations are available in the settings:

• Cut parameter- cuts from URL given parameter. For example, if the operation is selected Cut parameter and the name sort is specified, then in the reports the URLs http://example.com/?sort=by_price and http://example.com/ will be combined into a single URL - http://example.com/.
• Cut all parameters- cuts out all available parameters from the URL.
• Cut anchor - cuts the anchor from the URL (#anchor).
• Replace https with http- in reports all pages will appear as accessible via HTTP protocol.
• Replace domain - replaces the domain displayed in page addresses in reports.
• Convert to lowercase- all characters in the URL are displayed in lowercase in reports.

Operations are applied to the page address of the reporting site, as well as to the referrer. For example, operation Cut parameter specifying the name sort will exclude the sort=by_price parameter from the referrer http://source-example.com/?sort=by_price .

Note. For one counter you can specify 30 filters and 30 operations. The setting does not change previously collected data.

Robot filtering

Robot filtering is applied to the original data. The setting allows you to filter information about robot visits in service reports. Regardless of whether filtering is enabled, information about robots that visited the site will be reflected in the Robots report.

The following modes are available:

Count visits of all robots

Robot visits will be taken into account in all Yandex.Metrica reports.

Filter robots only by strict rules (default)

This filtering is carried out using fixed templates for IP addresses and User-Agent and does not provide false positives. Under strict rules, only a small fraction of robots can be detected.