• When using it, a sample is built based on the results of a combination of different methods for selecting texts, which are interpreted by an expert manually, without using any programs.
• Search options for texts for analysis may vary. Both search queries and continuous viewing of the content of specified Internet sites are used.
• The resulting array of texts is analyzed by an expert according to pre-formulated criteria.

With the help of manual monitoring, you can successfully solve almost any task. The analysis will be accurate, complete, and conducted according to any criteria. Naturally, it is much more expensive than automatic and makes the highest demands on the specialists who conduct it, but the results are worth it.

In their pure form, both options are rarely used. Usually one or another principle is taken as a basis, and then the necessary measure of additional involvement of either an expert or special programs is determined.

Why do you need social media monitoring?

1. Find out how you are treated

• In other words - monitoring your reputation in social networks, in particular, tracking reviews in social networks. We will find out in what context your company is mentioned on the Internet and organize the elimination of negative reviews.

1. See the effect of advertising.

• The task of advertising or PR campaigns is to cause a wide resonance, to form the right attitude. To find out if this task has been accomplished is best in the most reliable way: by seeing the resonance itself with your own eyes. Social media monitoring provides this opportunity.

1. Conduct marketing research

• There are many studies based on the analysis of Internet texts of service consumers. Thus, they often study the state of affairs of competitors, find out the influence of given sources on their audience, and much more.

There are a lot of tasks that are solved by monitoring social networks. Some firms build successfully functioning client services on its basis. You can often see how effectively the sales system through social networks works. The results of monitoring social networks are used in the search and recruitment of employees.

Our agency organizes monitoring of social networks of any complexity for you, with the involvement of the most experienced experts in the field of manual monitoring. With the skillful use of the results of our work, you will turn social networks into an inexhaustible source of new customers!

"Monitoring of social networks"

Speaking about the monitoring of social networks, it is very important to mention those people who are interested in this issue more than others. These are, of course, employers. The reputation of an employer in the labor market is a common information field for assessing the advantages and disadvantages of an employer in terms of both a potential employee and existing staff / former staff.

Monitoring can be carried out both by employers themselves and by specialized companies. The following types of monitoring can be distinguished:

• - regular monitoring, which allows you to constantly monitor the information that appears on social networks, helps to understand the trends in changing opinions, the reaction to this or that information, and even adjust the company's information policy.
• - primary monitoring is intended for companies that are just starting to use new Internet media in their communication activities. Primary monitoring will allow to determine "hot topics", places of presence of the target audience, opinion leaders. This analysis allows you to create the basis of a communication strategy on the Internet.
• - competitive monitoring allows you to determine the position of competitors in the network, their activity and the company's promo;
• - reputation monitoring - is carried out, as a rule, for a period of at least 6 months and will allow you to determine the image of the company and its products, which has developed on the Internet as a whole. The results of the analysis are the image of the company that is formed by consumers who are looking for information in the boundless expanses of the network.

When looking for a job, the Internet is one of the main sources of information about employers - job seekers are interested not so much in officially distributed information as in the opinions of current employees, former employees, and candidates who have not been interviewed. Therefore, a company that cares about its image must constantly monitor negative reviews online and work to increase staff loyalty.

More than 30 million adult citizens of the Russian Federation use the Internet, and in large cities the share of users exceeds 50%. Runet is full of job search services, the leaders of which are visited by more than 100,000 people a day. The Internet has become a familiar companion for both employers posting their vacancies and job seekers viewing job offers via the Internet. It is not surprising that before the interview, the applicant will look for information about the company where he is already used to getting information - in the Global Network. 4

The Internet has become the main media channel of our time, through which interested parties receive information about the activities of organizations. The company's reputation as an employer is an essential part of its image. The bad reputation of the employer is the negative feedback from former employees or job seekers, insiders, getting into the "black lists". At the same time, rumors and gossip discussed on the Internet leave traces in the information field of the company for a long time, being in an open search at the request of all interested users.

What actually the Internet can tell about the employer? The information field of the employer consists not only of official documents - news, press releases, white papers prepared and

distributed by the public relations service. The image of the company in social networks is also formed by expert opinions, employee reviews, insider information, rumors, gossip - real or falsified by ill-wishers. Modern social networks, developing in the spirit of the Web 2.0 concept, give users a lot of new opportunities to publish their opinions and search for information. Content created by users on Web 2.0 sites is easily indexed and gets into search engines almost instantly. This information cannot be ignored because everyone can see it. five

Perhaps, by typing the name of the company in the search box, the user will find official materials. But it may turn out that in the informal ocean of information, the official materials of the organization are only a drop. The main search engines of the Runet provide a separate opportunity to search through blogs, which greatly simplifies and speeds up the access of all interested parties to word of mouth information. The exchange of opinions and feedback within the framework of UGC sites (sites whose content is created by users themselves) has become a high-tech continuation of “kitchen conversations”. With more than a million people in Russia keeping personal diaries online, the information in which is available to everyone, the risk that the company's reputation will suffer from a careless or intentional word is greater than ever. An offended or fired employee, a manager who is dissatisfied with a bonus, an applicant who has not been hired - the statements of these persons can negatively affect the attractiveness of the company's image as an employer and play a decisive role in making a decision on employment. At the same time, there is no doubt that the applicant is more likely to believe what a person of his circle said - a former or current employee of the company, a candidate who was interviewed - rather than the official statements of the HR service or the press secretary.

There are quite a few examples of negative changes in the reputation of employers under the influence of individual reviews.

In social networks such as vkontakte.ru, odnoklassniki.ru, Webby.ru, employees can form groups based on their work in a particular company, but their statements in these groups, not limited by corporate ethics, will be informal.

Talking about an embarrassing work moment on a group page can engage a large number of participants in the discussion and become the sensation of the day on the Internet.

People write more and more, and not only about their jobs, failed relationships, but also about products and services, about their experience with this company. All this information is carefully processed by search engines and becomes available to millions across the country.

Regular work on monitoring and responding to mentions of the company on the Internet will create a positive image about it: we hear, we react, we help and it is remembered. Again, do not forget that commenting is part of the link strategy.

Social networks are practically not indexed by search engines (the only exception is some part of facebook), so tracking here takes place manually without the use of automated tools (daily viewing of the main groups, communities).

As in any marketing system, in SMM, the actual task is to monitor the attitude of users to the company and products. Moreover, if in classical marketing such monitoring requires time-consuming research, surveys, work with focus groups, then in the case of social networks, the process is much less labor-intensive and largely amenable to automation.

So, monitoring is a manual or automatic search for mentions of the company and products in social networks and blogs, as well as their further processing and analysis of the collected information.

Let's take a closer look at the key points of monitoring social networks and the blogosphere.

Monitoring tasks

The first question that usually comes up when talking about monitoring social media and the blogosphere is, “What does it make sense to use it for?” There are several basic marketing tasks that are effectively solved with the help of monitoring.

Early detection of negativity

One of the features of social networks is the high speed of dissemination of information. At the same time, not only positive statements, but also criticism instantly diverge. Often, a surge of negativity against a company can turn into a real information explosion in a few hours, and with each hour the tension will only increase.

Thus, the sooner you can start a campaign to neutralize the negative, the more effective and less costly it will turn out. However, in order to start it, it is necessary to quickly trace the centers of negative information, and it is this task that can be solved by monitoring social networks and the blogosphere.

Competitor analysis

Analysis of the prevailing opinion of the target audience about competitors, studying their marketing policy, as well as comparing their performance with yours (number of mentions, tone of mentions, audience response) are important marketing tasks.

A properly configured monitoring system allows you to automate the collection, systematization and analysis of key data, as well as compare the performance of competitors with your own in dynamics.

Evaluation of promotion effectiveness

Another important monitoring function is to evaluate the overall effectiveness of ongoing PR campaigns both online and offline. An increase in the number of mentions, as well as a change in the ratio of sentiments in favor of positive ones, indicates that the ongoing campaign increases your information field and heals it. If the indicators remain unchanged, then this, first of all, indicates that the campaign does not solve the tasks assigned to it.

In addition, by analyzing the data obtained over a significant period of time, one can trace the evolution of the audience's attitude to the brand and how the ongoing PR campaigns influenced this.

Feedback

It is often important for a company to get feedback from customers about themselves and their products. This allows you to adjust the positioning, marketing and PR strategy, pay attention to the nuances that are important for the consumer.

For example, after the launch of a new model, automakers monitor opinions to understand how people perceive it: what strengths and weaknesses they highlight, what arguments they have for and against buying, what models from other manufacturers are compared with.

Search for topical discussions

Sometimes it is important to take part in thematic discussions. So, for example, if there is a discussion of various Internet service providers somewhere, then this can often be used by a specific provider to convey to the interested audience information about themselves and their offers, to emphasize advantages over competitors, and also to neutralize the negative, which may appear in this thread.

Accordingly, monitoring for well-chosen queries makes it possible to detect such discussions soon after they start.

Requests for monitoring

The next important point is to determine the range of requests that will be monitored. This is the formation of the so-called semantic core. The number of requests can vary greatly depending on the business area, product line and the general saturation of the information field. As a rule, the average size of the semantic core is 20–50 units. There are several main categories of requests used for monitoring.

• Company brand. All major spellings are highlighted, including those in different languages ​​and in transliteration. Most systems have a heuristic semantic analyzer, so you should not separately enter spellings of the name with errors and in different cases.
• Product brands. Product line names - for example, tariffs for mobile operators, model names for automakers.
• Surnames, names and patronymics of top managers. Here a classical linguistic problem arises: if the surname and name are rare, then it is quite easy to find their references, but if this is a common combination (Ivan Sergeev, Alexander Petrov, etc.), then the information noise will be too great and most of the references will not will concern the target figure. Therefore, it is necessary to add an additional query to the surname and name. Usually this is either an indication of professional activity (“Alexey Ivanov, pharmaceuticals”), or a position (“Alexey Ivanov, CIO”), or the name of the company (“Alexey Ivanov, "Pharmacor") This will significantly reduce the number of inappropriate mentions in the monitoring process.
• competitor brands. It makes sense to track these requests if the task is to compare your own information field with that of competitors. However, it should be taken into account that the monitoring of each additional competitor will greatly increase the size of the semantic core and require the involvement of large labor resources.
• Industry thematic queries. Such requests must be monitored if the task is to search for thematic discussions. In this case, phrases are highlighted that users can use in discussions on a specific topic. For example, monitoring queries such as "choose an ISP", "which ISP", "recommend an ISP", etc. will allow the ISP to find targeted discussions.

Monitoring Platforms

Another topical issue is which specific platforms it makes sense to monitor. In most cases, it makes sense to keep the following sites in sight.

In addition, most automatic monitoring services allow you to add additional sites that are not in the database, but that need to be monitored: online media, thematic portals, comments on selected sites, etc.

Monitoring methods

There are two main monitoring methods: manual and automatic. Both in the first and in the second case, the process is carried out through special services, however, automatic monitoring has the following advantages.

• Permanence. Tracking is ongoing, at any time you can get a cutoff as of a given time.
• Saved settings. Unlike manual monitoring, in the case of automatic monitoring, there is no need to re-enter the monitoring parameters (requests, platforms, etc.) each time - the monitoring campaign is configured only once, and then only adjusted if necessary.
• Deep monitoring. In the case of manual tracking, monitoring with a chronological depth of several years can be problematic. Yandex.Blogs often prevents you from viewing a large number of pages of search results. Automatic systems allow you to bypass this limitation.
• Automatic systematization of data. With manual monitoring, the information received must be entered into tables independently.
• Analysis of the received data. Automatic systems themselves are able to analyze the data, in particular, to determine the tone of references.
• Formation of reports. In addition, in most monitoring systems, it is possible to generate reports based on the received data and their analysis. In particular, these reports include a graphical representation of the information field, diagrams that allow you to visually demonstrate the change in its volume and quality in dynamics, comparison with competitors, the ratio of positive and negative information fields.

As a result, automatic monitoring allows, firstly, to significantly reduce labor costs by automating routine processes, and secondly, to achieve high accuracy due to better data systematization and the use of a significant number of analytics tools.

However, in some cases it may be more appropriate to use manual monitoring, in particular:

• if the number of mentions of the company does not exceed five per day;
• if the only actual task of monitoring is an immediate response to the emerging negative, while there is absolutely no need for data collection and analytics;
• if there is no money to buy an automatic monitoring system.

Manual monitoring

The most relevant tool for manual monitoring is Yandex Blog Search (//blogs.yandex.ru). This is a unique service, which has no full-fledged analogues in the world. Yandex.Blogs indexes several tens of millions of Russian-language blogs, microblogs and forums. At the same time, entries from frequently updated resources are indexed several times per hour, rarely updated blogs are indexed several times a day. Consequently, the Yandex.Blogs database always contains the most up-to-date information on blog posts, which makes it an ideal monitoring service.

An important feature of Yandex Blog Search is that, by default, search results are ranked not by relevance to the query (as in a regular search), but by chronology. That is, the less time has passed since the publication of the post, the higher it will appear in the search results. Thus, on the first pages of search results, as a rule, the most recent mentions are collected. This is consistent with the general logic of the blogosphere: the relevance of published information is rapidly declining.

Yandex.Blogs has several special tabs that allow you to filter search results.

Among other things, Yandex.Blogs can use a Yandex-specific query language to make the search as accurate as possible. You can read more about this language on a special page: //help.yandex.ru/search/?id=1111313.

Thus, the algorithm for manual monitoring through the Yandex Blog Search service is as follows.

• Step 1. Queries for monitoring are sequentially entered into the search window.
• Step 2. For each of the requests, the data for the period that has passed since the previous monitoring is analyzed.
• Step 3. Mentions that require an immediate response are highlighted.
• Step 4. The received data are entered into a special table. Here are the main fields for such a table:
  • date of;
  • inquiry;
  • link;
  • area;
  • blogger;
  • post title;
  • tone of mention;
  • newsbreak;
  • audience reaction (number of comments);
  • the need for a response.

I want to emphasize that such a detailed summary is needed only if the data received is planned to be analyzed. When the only task is to detect negativity and respond to it, pivot tables are not needed.

Automatic monitoring

Today, the market for automatic monitoring systems for social networks and blogs is experiencing a sharp growth: new services are constantly appearing. Of course, this is due to the active entry of business into social networks and the gradual complication of the tasks that it sets in this environment.

Most automatic monitoring systems work on the basis of the Yandex.Blogs API, that is, in fact, using their own algorithms, they analyze and modify the data provided by Yandex. However, there are a number of services (for example, Wobot and Kribrum) that create their own index of blogs and forums, and then monitor it.

As an example of an automatic monitoring system, you can take Brandspotter - one of the most popular Russian-language services of this kind. Let's take a look at the main features it provides.

• Summary of mentions. New mentions for given queries appear almost in real time.
• Summary data. Here, again, graphs are built in real time with the dynamics of mentions, the ratio of positive and negative statements, as well as showing a comparison of key indicators with those of competitors.
• Information about opinion leaders. In the system, you can trace the history of interaction of each opinion leader with the brand: when, how many times and in what tone he mentioned the company and products. It also indicates the popularity of the blogger, which allows you to understand the significance of his statements. In addition, Brandspotter's special algorithms allow you to identify potentially interested users, such as friends and readers of opinion leaders loyal to you.
• Information about communities. Brandspotter collects information about each community where the brand has been mentioned. In particular, indicators such as the number of subscribers, the total number of posts and comments, the average number of comments per post, the average number of comments per participant are monitored.

This data allows you to make a decision: does it make sense to respond to the publication on this site or the audience coverage will be minimal and no action is required.

• Analytics. Information about all found mentions is stored in the system and, if necessary, is available for sorting, filtering by specific criteria (for example, mentions on a certain platform or only a certain tone) and further analysis. Brandspotter has several built-in unique metrics: attention (the number of users who received information about the brand through social networks); engagement (the extent to which users are involved in active activities around the brand - for example, in communities or special promotions); influence (how loyal the audience is to the brand). All of them can be tracked both in the cutoff at the current moment and in the dynamics of change.
• Reports. Brandspotter has a function to automatically generate reports in a convenient format (for example, .xls or .pdf). At the same time, it is possible to configure what specific data will be presented in the report.

In addition to Brandspotter, the following examples of automatic monitoring systems can be listed:

• IQBuzz;
• YouScan;
• Wobot;
• Babkee;
• "Cribrum".

In general, the automatic monitoring algorithm can be reduced to the following steps.

Step 1. Selection of the semantic core.

Step 2 Setting up a monitoring campaign.

Step 3 Campaign launch.

Step 4 Tracking results at a specified frequency.

Step 5 Analysis of results.

Building a monitoring system

The question arises: how to properly build a monitoring system so that it effectively and systematically solves all of the above tasks?

One of the most effective systems is a three-level one.

First level: permanent monitoring

There is a category of mentions that require an immediate response. In most cases, these include negative mentions of a company or products, as well as urgent consumer questions that will no longer be relevant in a short time. Therefore, it is necessary to constantly review newly appearing references.

Depending on the number of mentions, such monitoring is performed from three to ten times a day. At the same time, first-level monitoring does not require deep analytics. A number of automatic services provide the ability to receive notifications when publications appear in which the brand is mentioned with certain phrases.

For example, for Internet providers, these can be the phrases “not working”, “disconnected”, “can’t connect”, etc. If the service detects such a review, you receive a notification in the form of a letter, Skype message or SMS to the specified phone number .

Second level: periodic monitoring

The main task of this level of monitoring is to analyze trends in the information field over the past period (as a rule, a period of time is taken from one to four weeks).

The task of monitoring social networks today is faced by any brand. With the help of social media, you can find out the most complete information about the audience of a product or service, its opinion about the work of the company. Cossa.ru offers you a selection of Russian systems that carry out such monitoring.

Babkee

The system is positioned as a unique service for evaluating the significance of a message - a calculation of what its potential audience is. Corporate tariff users are provided with the services of a personal consultant.

We decided to use the free version and monitor the mentions of "cossa" and "cossa.ru" over the past 5 days. The result, to put it mildly, turned out to be unexpected.

From it, in particular, it follows that over the past 5 days we have never been mentioned on Facebook, and there were a measly eight links to Cossa.ru posted on VKontakte.

Cost: from 0 to 29,900 rubles.

brand spotter

The system offers a standard set of services - determination of the emotional coloring of statements, statistics on topics, platforms, authors, comparison with competitors.

Several demo examples of static reports are available on the site. They look like this:

The project is being implemented by people with outstanding biographies.

Cost: from 31,500 to 42,000 per month. There is a one month free trial period.

buzz look

The system is a panel for processing mentions in social media, forums, YouTube and Flickr. With its help, you can monitor the reputation of your own company and the activities of competitors, collect suggestions from customers and support the brand's online community. The service provides a free 14-day testing opportunity, after which you can choose one of 3 service tariffs:

A search is available on the site that allows you to find mentions of interesting words in 6 social networks in real time.

So, at the request of "Cossa.ru" we managed to find references to our materials on Facebook and Twitter, but problems began with YouTube - in addition to the video from the "Influence Index" project, the system highlighted videos about an elite hotel and a massage parlor as containing references to our portal in Italy.

buzzware

The system, developed by the research company Net Mind, allows you to monitor brand opinions expressed by users in social networks and blogs, tracks data in retrospect (up to 1 year).

“Acceptance of applications for invites is temporarily suspended,” the announcement on the main page of the site says. However, the creators are ready to provide accounts on an individual basis after a written request.

IQBuzz

The system conducts round-the-clock monitoring of social networks, which allows receiving information in real time. There is the possibility of collective work with the service, as well as the opening of general access to the analyzed data for everyone.

The system determines the tone of user messages, analyzes the socio-demographic characteristics of their authors based on information from social media profiles.

The website of the system indicates that there is a possibility of free testing for 7 days.

An example of a summary report - Cossa.ru mention rates for the last month:

Cost: from 3,500 to 21,000 rubles per month.

Kribrum

The system allows you to track and analyze brand mentions, products or services, key persons, events, geographical names. An automatic algorithm allows you to evaluate the emotional coloring of statements and build interactive reports. The system screens out spam messages and those in which the brand is mentioned in passing, for example, posts like “Let's meet tomorrow at 8 at Alfa Bank.”

The distinctive quality of the system, the creators call a special linguistic technology, which:

• correctly handles phrases with a negative (for example, the phrase “I thought it was a good bank, but it turned out that it wasn’t” will be perceived as a negative review, despite the presence of the word “good” in it);
• highlights the characteristics of each object in a complex sentence (for example, in the phrase “this provider has high speeds, but technical support is no good” both positive and negative reviews will be marked);
• perceives amplifications and comparisons (“not the best choice”, “the worst choice”, “Beeline is better than MTS”, “MTS is the best”).

Data begins to be displayed in the system 2-4 hours after its publication, which allows you to work with mentions almost in real time.

Cost: set individually, based on the selected parameters. There is no free demo period.

semantic force

Distinctive features of the system are a linguistic algorithm that ensures high relevance of the received data, monitoring of unstructured sources, such as comments in online media and online stores, more than 20 types of analytical reports. SemanticForce is integrated with external systems: KLOUT, Copiny, GoogleAnalytics.

Here is the data we were able to get on social media mentions of Cossa.ru over the past week:

Chart showing the ratio by media type:

Dynamics of mentions of the resource over the past month (in the main social networks):

Wobot

The creators of the system claim that with the help of Wobot, you can find the maximum number of mentions. This is explained to a large extent by the fact that the service allows you to trace the retrospective of opinions up to 2005.

A wide range of metrics is available - engagement, the social graph of users, the size of the audience. The mechanism of the system, which allows you to determine the tone of the message, is self-learning. It is possible to add personal metrics upon request. A 14-day free demo plan is available.

The report in the free demo version shows the number of mentions of Cossa.ru over the past week:

Unlike the Babkee service, YouScan saw several mentions of Cossa.ru on Facebook:

Cost: $1,990 to $14,990 per year.

According to Alexey Orap from YouScan, paid accounts provide the ability to monitor more topics, access to them is provided to more users. Also, in free systems, there is a limit on the number of topics that can be saved in the system, while paid ones accumulate a huge array of data.