Recently, I set out to figure out at what time you need to publish Vkontakte posts in order to get maximum coverage. To find out, I did a little research.

It's about the group of copywriters. Other areas have their own audience with their own preferences and schedule.

Theoretical part

Interested subscribers already regularly visit the group to read new posts. There is no point in subjugating them. The more interesting you write, the more people will remember the community and view its wall.

But you can influence the number of people who see the post in the news. To do this, you just need to publish content when a lot of your subscribers are online. or just before rush hour. These are the numbers I measured.

Experimenting

For experiments, I chose 3 communities: Overheard | Copywriting, My blog Web.txt and Panda copywriting (the liveliest of the major groups). Purpose: to determine the time with the maximum online subscribers-copywriters.

The study lasted a week - from 14 to 21 July. Every hour, starting at 9 am, Alexey and I checked online in these three groups, and entered the indicators into a table.

In total I made 7 simple tables

And then I calculated the average values, divided the results into groups: from the warmest (peak online) to the coldest.

Everything was calculated according to Moscow time.

Early in the morning people are still sleeping - it makes no sense to publish posts. The revival begins at noon, by 13-00 it reaches its peak value and remains at a high level for another 4 hours. After 17-00 working copywriters go home - online falls. And at 21-22 hours people come again.

What conclusions can be drawn from this:

• The best time to post is just before the lunchtime peak. For example, at 12-30.
• The next post can be published right after lunch - online will be big for a few more hours.
• Evening posts publish around 20-30 - before the night peak.

Days of the week

At the same time I will share the results by day of the week. This will be useful for those who do not write posts daily.

Of course, statistics based on one week's results are not very accurate. Therefore, I do not advise you to listen to it especially. Now I'm working on automating measurements. When everything is ready - I will collect readings for a couple of months - and then I will publish new, more accurate data.

But here the statistics of the Overheard group come to my aid. There, people themselves offer content, new posts appear regularly throughout the day. And we can look at the activity statistics by day of the week.

See the "pits"? It's the weekend. Don't post anything important on weekends. Nobody will read this. The coolest content should be released on Monday-Tuesday. Then you will get maximum coverage, and the group will grow faster.

Social poll

I conducted a survey with almost 200 people. The purpose of the survey is to find out what time people prefer to view the news feed.

Conclusion: posting in the evening (around 20-00) is a great idea. Firstly, at this time in groups there is a good online. Secondly, people are already free from work and ready to read your content. Thirdly, almost no one writes at night - so in the morning your publications will still be closer to the top of the news feed.

Of course, 200 people is very little. Don't take this poll too seriously.

conclusions

I repeat. I did research in groups of copywriters. It only applies to copywriters. If your the target audience- not copywriters - do not use these data. They may not work for you. It’s better to set aside a week and follow your subscribers - it will immediately become clear to you when the content should be released.

What can be learned from the study:

• The best posting time is before lunch. This way you will get maximum coverage in the news.
• The second place goes to the evening - 20-21 hours. At this time, good online, besides, your publications will still be visible in the morning.
• Do not post anything important on weekends - coverage will be minimal.

Thank you for taking the time to read this post. I hope my little research will help you increase your audience reach.

Rate this article:

Subscribe to newsletter

Any analytics system collects polygamy of site parameters, one of the most basic is the average visit time, but should it be taken literally? Let's look at the example of Yandex.Metrica.

Definition given parameter in YM: “Time on site is the average time (in HH:MM:SS format) spent on the site by visitors. It is calculated as the difference between the time of the last and first registered page view by the visitor within one visit.”

The key word is "between". Few people understand the real meaning! This is not the average time a user spends on the site, but the time between (!) The first and last pages viewed. The more pages a user views in one session, the closer this time is to real time, but what if the site usually has 1-2 views per visit? Just forget about the existence of these statistics.

If you have a site on which the user gets to the landing page - reads necessary information, makes a phone call, fills out an AJAX form, clicks-clicks, but does not reload the page (counter) and ... leaves, then the average time spent will be very far from the real one. I'll give you an example. The average time according to Yandex.Metrica of users from search engines was 31 seconds.

Now let's go to Webvisor and see how things really were:

Those. the average residence time was actually 5 minutes 52 seconds. The reason for this difference lies in the different approach to collecting statistics.

The reason for this situation lies in the way the duration of the user's visit is determined. It is defined as the time difference between downloading the Metrica counter code on one page and downloading the Metrica counter code on another page. Thus, if the user was only on one page, then we cannot accurately determine the visit time in a regular report. Webvisor reports record the user's entire visit, so the time of the visit is accurately determined in such records.

Enabling the option to collect an “accurate bounce rate” in the counter settings will help reduce the bounce rate - by excluding visits longer than 15 seconds and one hit, but not increasing the accuracy of the “view time”.

Yandex.Metrica support service: When this option is set, an additional hit is sent after the first 15 seconds of stay. Further hits are not sent.

What to do?

If you need to fight for an increase in viewing time, then you need to motivate the user to walk around the site as much as possible, but if there is no such task, then just forget about this metric.

Many SMMs are very often interested in what time is the best time to post in their group. Western colleagues wrote a lot of articles about this, drew a bunch of infographics. But all this is about facebook, twitter, pinterest, and not about our VKontakte, where we spend a lot of time.

In fact, this article will not talk about publishing posts to your group or public page, but about the best time to advertise in VK communities to attract users to a group or to commercial sites (you sell something)

It is probably worth starting with the fact that, like on television, the Internet has prime time - the period of time in which the maximum number of users are on the Internet. There is a prime time and VK. It looks like this

• Activity - users online in social network In contact with
• Color - the red, the more users online in a given period of time
• spaces- at some points in time it was not possible to receive data from VK
• Data - Russia, Moscow, different age groups (gathered separately)
• Time- according to MSC

From the picture "14-80" you can see that the general "prime time" starts at 20:00 every day of the week (slightly less on Friday and Saturday) and ends at 00:00.

Activity (the number of users online in VK) starts growing from 8 am to 4 pm, where it hangs until 8 pm and the growth continues after which, at about 01:00 am, it starts to fall rapidly until 04:00 am. From 04:00 to 07:00, there are fewer people from Moscow on the social network.

Few people are interested in the overall picture of activity, so let's consider the activity of individual age groups

We are interested in the age groups 22-30, 31-45 and 46-80 and the general group 22-80 ... with schoolchildren, children have not yet learned to work, except perhaps in applications.

Based on activity, different age groups, you can build template models of user behavior. Approximately it looks like this

… woke up, had breakfast, checked through a smartphone if anyone had written on the social network during the night, the way to work, turned on the computer at work, looked at My News, received the first stream of information for discussion with colleagues at lunch, worked until lunch , sometimes being distracted by “My News” and a personal message, had lunch, after lunch, to enter the working mode, he again checked the social network for new messages, and also quickly ran through news feed, worked, before leaving work, checked the social network again, on the way home, came tired, washed, ate, turned on the radio or his playlist in the social network in parallel, continued to do household chores, sometimes was distracted by answers in a personal, shared a couple funny pictures for relaxation and feeling moral satisfaction, watched some series or movie on a social network, fell asleep ...

Of course, for a more accurate construction of the template, it is necessary to survey the group that advertising on the social network will target (target audience).

My personal experience promotion of the meta-search engine for air tickets showed that social network users react to information about air ticket sales (put likes, repost, follow links) better from 11 to 15.00, the peak of transitions from advertising was exactly at 11:00-12:00. At the same time, the placement of advertising in "prime time" did not give the same effect, despite the high activity of social network users.

The heat map shows that the majority of requests for advertising in groups are placed from 13:00 to 16:00 on weekdays. Most advertisers have come to the right conclusion - VKontakte's "prime time" is not suitable for advertising, users want to relax and do household chores, and not receive information about tire sales in a group about cats.

In fact, based on the information above, we can also draw conclusions about effective posting to your group or public page:

• Understand who your user (TA) is: at what time he is active on the social network, what he does, what communities he subscribes to - create your own user behavior template to determine at what moment and what information he is ready to receive;
• Experiment and analyze. Track everything you can. Data will help build a pattern of behavior

General conclusion

Today is not yesterday, and tomorrow will not be today. Next Tuesday is only a week away and it will be different because it will snow. At a certain point, someone will share a post, and then his friend with 100,000 subscribers will notice and he will share it too, but tomorrow is not today, and this will not happen, which means last Tuesday was not so effective compared to this one, because it didn't snow.

The dependence of attendance on the day of the week and time of day is one of the most obvious and stable patterns in Internet statistics. If we do not take into account the topics of specific resources, but take an average picture for Runet, then we can talk about a drop in attendance by about 40% on weekends and about a fluctuation in attendance during the day by almost 4 times (from a minimum around 5 am Moscow time to maximum between 2 and 6 p.m.).

However, it is obvious that the combination of two factors - time of day and day of the week - will inevitably lead to a significant complication of the picture. It is natural to assume that the hourly dynamics of attendance on weekends will differ markedly from the dynamics on weekdays.

This article analyzes the hourly traffic dynamics over a weekly period. The week from February 5 to February 11 was taken as an example. All the data given in the article is calculated for a sample of the 500 most visited (according to the statistics system SpyLOG) Runet resources. In sum, the Top 500 resources collect more than 75% of all traffic on the Runet, and the results obtained on the basis of this sample adequately reflect the processes common to the entire Russian Internet.

It should be added that the data presented in the article characterize the behavior of the entire Runet audience as a whole, and not just Russian users; accordingly, users from other CIS countries, as well as the Russian-speaking diaspora far abroad, make a significant contribution to the overall average picture.

Weekdays and weekends: two models

The main difference is that the daily rise in traffic is much weaker on weekends. As a matter of fact, it is the relatively low attendance in the middle of the day that predetermines the overall lag of weekends in terms of audience activity.

On weekdays, already at 5 am Moscow time, attendance begins to grow due to the Far Eastern and Siberian audience. At 8 a.m., when Muscovites wake up, growth accelerates sharply, and by 11 a.m., audience activity is 4 times higher than the minimum nighttime level. At the same time, the peak of attendance on weekdays is extended in time: only around 6 pm, with the end of the working day for visitors from the Moscow time zone, attendance begins to decline. Until 9 pm every hour the audience is reduced by about 30-40 thousand people, and after 9 pm the decline slows down to 10-20 thousand people per hour.

Weekends paint a different picture. From midnight to 5 am (Moscow time) attendance is at the same level as on weekdays. However, weekend morning growth is much weaker than on weekdays, and besides, it starts later - at 7-8 in the morning. If on weekdays in the interval between 5 and 11 o'clock in the morning attendance grows by 3.9 times, then on Saturday it is 1.8 times, and on Sunday it is only 1.6 times.

Since the majority of users access the Internet from home on weekends, there is no sharp decline in the evening - high activity remains until 8-10 pm, and after that, traffic decreases much more slowly than on weekdays.

Share of "office" audience

The difference in daily attendance between weekdays and weekends is mainly those users who have access to the Internet only from work or from the institute. Accordingly, it is possible to approximately estimate the dynamics of the activity of the “office” part of the audience (Fig. 2). This group (without taking into account the "round-the-clock" audience - those who have access to the Internet both at work and at home) according to our study is about 41%, which corresponds to the estimates obtained from the results of online and offline studies. In the daytime, the share of the "office" audience is 50-60%, and about 25% of the audience during these hours forms access from the office to the "round-the-clock" audience.

As expected, the maximum activity of the "office" audience is observed at the beginning of the working day, around 11-12 o'clock - many users prefer to first learn the news, check their mail, read fresh jokes, and only then get down to work.

The given schedule of activity of the “office” audience on weekdays makes it possible, paradoxically as it may sound, to understand the reasons for the difference in the schedules for Saturday and Sunday (Fig. 1). In the period from 5 am to 5 pm on Saturday, attendance is 7-20% higher than on Sunday; at the same time, the differences are maximum around noon - i.e. at the same time when the "office" audience is most active. It can be assumed that the higher attendance on Saturday during the day compared to Sunday is determined mainly by the fact that more people come to work on Saturday.

Weekdays: deviations from the mean

The traffic dynamics model described above for weekdays is an averaged trend. Of course, the differences in traffic dynamics between individual working days are significantly smaller than the differences between weekdays and weekends, but they still exist, and in general, these deviations on the Runet can reach 15 thousand visitors per hour.

On Monday night, the activity of visitors is noticeably lower than the average nightly level: Russian users sleep off after the weekend, and in America Sunday is still going on (with the corresponding negative influence for attendance). But on Monday afternoon, when users come to work after the weekend, they use the Internet more actively than usual - in this case, we can talk about the effect of "laid" attendance, compensating for the absence of Internet access during the weekend. Moreover, on the night from Monday to Tuesday, the Runet traffic is maximum for night time (the second "night" peak falls on the night from Tuesday to Wednesday). The period from mid-Wednesday to mid-Thursday can be considered a reference period - the deviation from the average attendance on weekdays does not exceed 0.5% (Fig. 3).

The approach of the weekend begins to be felt already with the beginning of the working day on Thursday: albeit only by 1%, but the average attendance is declining compared to Wednesday and Tuesday. On Friday, the decline reaches 7% (around 6 pm) - the working day ends much earlier than usual.

It is interesting to trace the "leaders" and "outsiders" for individual hours. From midnight to 7 am, attendance is lowest (from weekdays) on Monday, i.e. immediately after the weekend, and from 7 am to midnight, the lowest attendance is observed on Friday (the eve of the weekend). As for the highs, from midnight to 4 am the most visitors are on Saturday, from 4 to 11 am on Tuesday and Wednesday, from 11 am to 10 pm on Monday, and from 10 pm to midnight on Tuesday.

Time of day and work schedule

Summing up the results of the study, we can state that the hourly dynamics of attendance during the week is determined by the influence of two factors. This is, firstly, the time of day and, secondly, the work schedule of users. The influence of the time of day is obvious - at night most of them sleep, during the day they are awake. Since more than 50% of the Runet audience is concentrated in the Moscow time zone, the pace of life of Muscovites significantly affects Runet traffic: even on weekends, nighttime traffic is 2.5 times less than daytime traffic.

As for the work schedule, its influence is determined by the fact that, as before, most of the Russian audience accesses the Internet from work or from the institute. Accordingly, during business hours in Moscow (approximately from 10:00 to 18:00 on weekdays), the Runet audience doubles compared to the "background" daytime attendance on weekends. The influx of "office" audience also increases the maximum gap in attendance between day and night hours - from 2.5 times on weekends to 4.5 times on weekdays.

The trend reverses after 11 pm on Friday - on Saturday night attendance is higher than the average nightly level: users stay at computers longer than usual, because they do not have to go to work or study the next day. And vice versa - on the night from Sunday to Monday, attendance is minimal: people are trying to have time to sleep off before the new working week.

Dynamics of attendance: comparative characteristics of the days of the week

Attendance, thousand people per hour

"If we want to understand Information society we have to measure it. Without measurement, we won't be able to track progress or identify gaps that need our attention."

Director of the Telecommunication Development Bureau of the International Telecommunication Union, Brahima Sanou

Internet audience

“We rank first in Europe today in terms of the number of users global network. There are already more than 90 million of them in Russia.” (from the speech of the President of the Russian Federation V. Putin at the International Cybersecurity Congress in Moscow, 07/06/2018)

18 years and older

On September 17, 2018, the All-Russian Public Opinion Research Center (VTsIOM) conducted a survey about "Do you use the Internet, and if so, how often?"

The share of Internet users in Russia - 81% of citizens. Including 65% go online daily. Among Russians aged 18 to 24, this figure is 97%.

Also, among the most active audience (using the Internet daily) are highly educated (78%) and financially secure (72%) Muscovites and Petersburgers (76%).

1,600 Russians aged 18 and over took part in the survey. The data are weighted by probability of selection and by socio-demographic parameters. For this sample maximum size errors with a probability of 95% do not exceed 2.5%.

According to the Foundation " Public opinion”(December 2017 - February 2018) total users (surfed the Web at least once a month) - 83.8 million people (72%). 74.7 million people (63.8%) use the Internet daily.

Data in million people

• General population - 72


• Central Federal District - 72
• Northwestern Federal District - 77
• Southern and North Caucasian Federal District - 71
• Privolzhsky Federal District - 68
• Ural Federal District - 70
• Siberian Federal District - 72
• Far Eastern Federal District - 75

Data in million people

Data in million people

Winter 2017-2018, percentage data

• General population - 64


• Central Federal District - 64
• Northwestern Federal District - 71
• Southern and North Caucasian Federal District - 63
• Privolzhsky Federal District - 60
• Ural Federal District - 62
• Siberian Federal District - 64
• Far Eastern Federal District - 68

Data in million people

Data source: Leaked data from FOMnibus weekly surveys of Russian citizens aged 18 and over. The surveys were conducted from December 2017 to February 2018, 24,000 respondents.

16 years and older

The Russian branch of the research concern GfK (Gesellschaft fur Konsumforschung) Group, on January 15, 2019, published the report "Internet Penetration in Russia".

The audience of Internet users aged 16 years and older was 90 million people (75.4% of the adult population of the country), which is 3 million more than a year ago.

73 million(61% of the adult population) go online on mobile devices.

Of them, 32 million Russians use the Internet only on mobile devices.

13% of Internet traffic in Russia falls on mobile devices (smartphones generate 10% of traffic, tablets - 3%). This is reported by "The World of Apple in one site" with reference to the data of the analytical company StatCounter.

In October 2016, for the first time in history, global mobile internet traffic outpaced desktop traffic. Tablets and smartphones accounted for 51.3% of global Internet traffic, while desktop PCs and laptops accounted for 48.7%.