The most complete guide to analyzing social media
In the modern dynamic world, we, humans, are characterized by everything and everyone, quickly sorting out and shoving them in the usual drawers. Otherwise, do not survive.
From here and "got": while one shove them into the old boxes, the leaders - create new ones. Who first got up - that and sneakers (drawer). To help future leaders of marketing, PR and Big Data Scientist, we have made another useful translation with a detailed explanation of the difference between “Social Media Analysis” and the Social Media Monitoring box.
')
PS Preliminary note from the translator: the term “Smart / deep analytics” is used (in the Social Media Intelligence article) implies active saturation with external metadata of source texts / photo / video, and, most importantly, so close to the smart work of Stirlitz and James Bond - using data to analyze and predict the development of events. No wonder that the intelligence services of England are called “Intelligence Service”.
The text is large, for connoisseurs there will surely be obviousness-boring, but it is chewed on the shelves and with examples for each social network. Stock up on tea and go deep into smart analytics of humanity.
=====================
The most complete guide to analyzing social media
The eyes reflect the soul, and social networks reflect the decision-making process of the buyer. At least it can be reflected. The mutual relationship between the brand and the customer has changed greatly, becoming a dialogue in which the degree of influence of the customer is higher than ever before. Nowadays, digital marketing allows you to establish direct links between the brand and the user, and the final goal was not just the purchase of a product by the consumer. Each brand is obliged to use social media to maintain customer loyalty and, most importantly, to encourage the spread of brand positives.
The impact of social networks on business can be divided into two types:
1. Indirect, through the analysis of the collected data
2. Direct - the “Buy” button is now directly accessible to numerous users of social networks.
What is Smart Social Media Analytics?
In the Merriam-Webster dictionary, the term “Intelligence” is defined as: The ability to recognize or absorb information.
Based on this definition, it can be said that simple data collection or monitoring of social media is not yet Analytics. Intelligence (Smart Analytics) process becomes only when certain conclusions have been made from the data, or the data have been used to adjust the course of action.
For example, you can simply see on the graph an increase in the number of subscribers, say that there has been a sharp change in the rate of inflow. But in order to turn this process of “simple vision” into an analysis of social media, it is necessary to understand why such changes have occurred.
Difference between Monitoring and Social Media Analytics
Social media and their users generate huge amounts of information every day. And nowadays, most companies naturally search for information that affects them on the Internet - references and user opinions about their brand, products and services. Social media monitoring is absolutely necessary for brands of all sizes to know what is being said about them, for which they are praised and scolded.
But simple Monitoring of social media, as mentioned above, cannot be considered as Analytics by itself. If the brand receives 8,000 positive mentions during the week, then this is certainly very happy information, but it is not clear, what were the specific references that were caused? If you do not analyze the data, then simply collecting data from several sources does not imply Analytics of the dynamics of the process.
Difference between Reporting and Social Media Analytics
Social media reporting is a pre (a) presentation of events that have already occurred, but in order for it to become an Analyst, you must ask the question WHY. Returning to our previous example - a graph showing a sharp increase in the growth of subscribers is usually the result of some existing factors. When reporting social media, you need to think about “And what follows from this?” To reflect the answer to this question in the report.
With a deeper analysis of the process, it can be found that the increase in the rate of growth was caused by just one particularly interesting message that you published. We have made certain conclusions, and therefore the process can be called “Analytics”. For even more in-depth understanding, you should find out, but what was so interesting about this particular message? Was this a picture or video that the public was actively sharing? How exactly is your brand used in this message? Was it advertising in its essence, or was it a congratulation on a seasonal holiday? In order to spread among the broad masses, brand content is being promoted in a variety of ways, and therefore it is useful to understand which methods of promotion have led to positive results.
By itself, reporting on data collected from social media is not a particularly difficult task (if you have access to a high-quality system for collecting data from social media - Translator's note), so if you want to add importance to your report by adding analytic information , you should think about the causes of the incident.
Go ahead. The average response time to messages in social networks addressed to the brand is certainly a very important indicator, but a more detailed study of the data may open up additional possibilities. For example, saturation with meta-data (geo, bundles of profiles, etc.) will reveal that a particular brand of beer most often refers to the evening while watching football in bars, which means that you need to respond promptly to tweets with the mention of @brand written after 15 hours.
How can Social Media Analytics help my brand?
Embedding smart social media analytics in many (or even all) aspects of your business can have far-reaching consequences.
An important problem for brands is also the delivery to all interested departments and departments of operational information obtained from social media, and tools to work with it: these can be specialized systems, dashboard, alarm notification system, etc.
How is the analysis of social media?
For a start, the main thing is that you need source data.
Social media generates such a large amount of data (billions of messages per day) that it is almost impossible to work with such volumes without special programs and algorithms for sorting and grouping data into a form convenient for analytical work.
[ There are specialized companies that aggregate data from various sources of social media: American Gnip (bought by Twitter), English DataSift , Russian Brand Analytics . Aggregation companies can supply, upon request, the necessary data slices taking into account the language, search phrases, geo, etc. for analytical modules and systems. ]
When parsing data and further imposing other parameters, there are charts and diagrams that can be used for analytics. For example, a comparative diagram of all the brands of a particular industry.
Facebook and VK
Each social network has its own indicators illustrating successful activity, and the data from each social network can play a different role in your marketing strategy. For many countries, Facebook is the ideal network for building and maintaining customer loyalty, as well as expanding brand coverage. For Russia, VKontakte is an equally important network. Facebook is the most popular social network in the world (1.5 billion active users per month), and represents an important goal for brands working in social networks.
For such a large audience, brands should pay increased attention to previously inaccessible types of analysis. Yes, it would certainly be convenient to stop at counting the number of likes, but you should not give out a brand mailing to a “too tricky” advertising agency. The main goal is to establish contact between the brand and the public, where the chosen advertising agency plays only the role of an intermediary. And if the agency’s participation is really necessary, then you wouldn’t like your audience to leave you to the mediator? The basic indicators - the number of likes and reposts - these data are not particularly useful if they are not deeply analyzed.
There are several ways to get useful information from this basic data. For example, you can simply monitor the above parameters to ensure that your subscribers are “consuming” the content you generate. The higher the consumption, the higher the consumer loyalty. Even if readers do not purchase your product today, this information will help them make a choice in your favor in the future. To contribute to this, you should constantly monitor the indicators of involvement both at the level of this campaign and at the level of the whole industry in order to understand what kind of content is most in demand by the public.
From the perspective of current business challenges, data derived from social media has a wide range of uses. Advertising of current promotions and offers, announcements about the launch of new products and other activities allow to receive feedback, including interaction with readers on choosing a new brand name. Throughout the history of mankind, it has never been so easy to get the opinions of huge masses of users, and if you do not use social media in order to better know your consumers, then you miss a lot of crucial information.
Twitter data analytics
Twitter is very popular with customer support workers. Although it is also used to distribute content, the main distinctive quality of Twitter remains the simplicity and speed of receiving feedback. Many large brands even have a separate Twitter page, which is not led by advertising specialists, but by a customer service team. The data collected on Twitter will help you more to meet the desires of consumers, because the faster you respond to them, the better they will be impressions of your brand.
American Airlines is considered one of the most effective companies in user support, which is also facilitated by the response rate on Twitter. American Airlines responds to more tweets than any other airline — and they do it much faster.
Analyzing Twitter data helps you keep abreast of your audience, and also keeps you abreast of what is being said about your brand. It is certainly pleasant to count retweets, but it is much more important to understand what exactly caused such a reaction. The information collected in social media will help you to be aware of what people say about the brand, you can prevent or extinguish scandals at the very beginning, as well as be aware of the activities of competitors.
A metric such as Share of Voice (share information presence in the competitive field) can help to find out the amount of your brand presence in discussions of key topics and concepts - for airlines, for example, understanding who is often mentioned along with such terms as 'kilometers' , 'rewards' or 'points' will help coordinate changes in strategy to maintain a leading position in the number of discussions, and ultimately influence the choice of potential buyers in favor of your company.
Analytics data on Instagram / Pinterest
Digital marketing is expanding rapidly - instead of textual information and messages, it now includes more and more visual elements - photos and videos. Social networks with a focus on posting photos, such as Instagram and Pinterest, are rapidly gaining popularity among both users and brands. While Instagram is still thinking about ways to monetize the platform, Pinterest has already introduced features such as “Recommended Publications” and “Buy Button”.
With Pinterest, analysts can identify both potential customers and people who have already bought a product. Tools that help identify leaders of opinion will prompt the most relevant of potential partners and will help significantly expand the coverage of your brand.
Data Analytics on YouTube
Every minute more than 300 hours of videos are uploaded to YouTube. With more than a billion users and a wide range of advertising opportunities, this channel is an ideal intermediary on the way to the consumer. YouTube provides its own tape showing the success of your videos, but when it comes to competitors' videos, you often have to rely on chance.
The system for identifying video clips gaining popularity can greatly assist in creating relevant content. It is important for analysts to take into account the growth rate of popularity and compare it with the average value - then any excess can be considered a signal to action.
Another key indicator is the duration of the video. Although it may differ depending on the type of content being lit, there is some ideal duration for different topics, after which viewers tend to lose interest. For example, we found that in the hotel business, long videos perform better, conducting detailed tours, while in travel and tourism, shorter videos show the best result, in which only key places are shown.
Although an increase in the number of views is in itself an important goal, it is equally important to increase the number of subscribers to the brand’s channel. The growth rate of views can be extremely high, while the number of new subscribers will be much less. Do not forget to compare these two indicators, and also try to find averages for your industry.
Studying YouTube data will help increase the popularity of your channel, and at the same time monitor the state of affairs of competitors.
Social Media Analytics Job
The ability to in-depth analysis of data obtained from social media is an important skill demanded by large companies. Today, many firms are gathering teams of social media specialists, and one of the possible roles is a social media analyst - one who will not only collect data, but try to delve into it and understand how it can be used to benefit the company.
Just one search query on LinkedIn with the words ' social media intelligence ' in the job title or job description will perfectly illustrate how important this profession has become.
How to choose social media analysis tools
There are many different tools for analyzing social media. There are tools for simple “listening” to social media (which can call themselves monitoring tools), there are tools for analyzing (which can also be labeled as monitoring tools) and tools for benchmarking social media (adapting the effective functioning of a company to improve performance). And at the same time all the systems in one voice claim that they will help in the analysis of social media. A logical question: what to choose?
First you need to decide what specific information from social media you want to analyze?
Social media monitoring (“listening”) tools can help you with the first two items. Standardly, collection and monitoring is conducted not over the entire information field (this is done by aggregator companies ), but on a specific set of social network accounts, blogs, forums and industry sites. Tracked given keywords that you specify. Such systems view a dedicated set of accounts and sources, and group messages by keywords.
To solve the full list of the company's objectives, including such complex ones as improving the brand's marketing strategy, it is necessary to use advanced social media analysis systems.
A source
==========================
The article is large, fairly complete, including data collection issues, which is rarely mentioned, but very expensive, if we are not talking about a simple API like Twitter. The only thing that caught my eye was the absence of such an important topic for Intelligent Analysis as linguistics: automatic tonality (especially important for analytics of mass real-time events), clustering, automatic selection of entities (NER).
It is clear that a full-fledged social media analysis system should include high-speed linguistic modules. To estimate the cost of technology, we can recall the recent purchase by IBM of AlchemyAPI , which has a high-speed linguistic system for some Western languages, for $ 100 million.
For English and Western European languages, BrandWatch English is a good example of a complete social media analytics system; for Russian, German, French and Eastern European languages, Central Asia and Arabic regions, Russian Brand Analytics . For in-house systems, HP Autonomy can be recommended, but data acquisition from aggregator companies will be required.
From here and "got": while one shove them into the old boxes, the leaders - create new ones. Who first got up - that and sneakers (drawer). To help future leaders of marketing, PR and Big Data Scientist, we have made another useful translation with a detailed explanation of the difference between “Social Media Analysis” and the Social Media Monitoring box.
')
PS Preliminary note from the translator: the term “Smart / deep analytics” is used (in the Social Media Intelligence article) implies active saturation with external metadata of source texts / photo / video, and, most importantly, so close to the smart work of Stirlitz and James Bond - using data to analyze and predict the development of events. No wonder that the intelligence services of England are called “Intelligence Service”.
The text is large, for connoisseurs there will surely be obviousness-boring, but it is chewed on the shelves and with examples for each social network. Stock up on tea and go deep into smart analytics of humanity.
=====================
The most complete guide to analyzing social media
The eyes reflect the soul, and social networks reflect the decision-making process of the buyer. At least it can be reflected. The mutual relationship between the brand and the customer has changed greatly, becoming a dialogue in which the degree of influence of the customer is higher than ever before. Nowadays, digital marketing allows you to establish direct links between the brand and the user, and the final goal was not just the purchase of a product by the consumer. Each brand is obliged to use social media to maintain customer loyalty and, most importantly, to encourage the spread of brand positives.
The impact of social networks on business can be divided into two types:
1. Indirect, through the analysis of the collected data
2. Direct - the “Buy” button is now directly accessible to numerous users of social networks.
What is Smart Social Media Analytics?
In the Merriam-Webster dictionary, the term “Intelligence” is defined as: The ability to recognize or absorb information.
Based on this definition, it can be said that simple data collection or monitoring of social media is not yet Analytics. Intelligence (Smart Analytics) process becomes only when certain conclusions have been made from the data, or the data have been used to adjust the course of action.
For example, you can simply see on the graph an increase in the number of subscribers, say that there has been a sharp change in the rate of inflow. But in order to turn this process of “simple vision” into an analysis of social media, it is necessary to understand why such changes have occurred.
Difference between Monitoring and Social Media Analytics
Social media and their users generate huge amounts of information every day. And nowadays, most companies naturally search for information that affects them on the Internet - references and user opinions about their brand, products and services. Social media monitoring is absolutely necessary for brands of all sizes to know what is being said about them, for which they are praised and scolded.
But simple Monitoring of social media, as mentioned above, cannot be considered as Analytics by itself. If the brand receives 8,000 positive mentions during the week, then this is certainly very happy information, but it is not clear, what were the specific references that were caused? If you do not analyze the data, then simply collecting data from several sources does not imply Analytics of the dynamics of the process.
Difference between Reporting and Social Media Analytics
Social media reporting is a pre (a) presentation of events that have already occurred, but in order for it to become an Analyst, you must ask the question WHY. Returning to our previous example - a graph showing a sharp increase in the growth of subscribers is usually the result of some existing factors. When reporting social media, you need to think about “And what follows from this?” To reflect the answer to this question in the report.
With a deeper analysis of the process, it can be found that the increase in the rate of growth was caused by just one particularly interesting message that you published. We have made certain conclusions, and therefore the process can be called “Analytics”. For even more in-depth understanding, you should find out, but what was so interesting about this particular message? Was this a picture or video that the public was actively sharing? How exactly is your brand used in this message? Was it advertising in its essence, or was it a congratulation on a seasonal holiday? In order to spread among the broad masses, brand content is being promoted in a variety of ways, and therefore it is useful to understand which methods of promotion have led to positive results.
By itself, reporting on data collected from social media is not a particularly difficult task (if you have access to a high-quality system for collecting data from social media - Translator's note), so if you want to add importance to your report by adding analytic information , you should think about the causes of the incident.
Go ahead. The average response time to messages in social networks addressed to the brand is certainly a very important indicator, but a more detailed study of the data may open up additional possibilities. For example, saturation with meta-data (geo, bundles of profiles, etc.) will reveal that a particular brand of beer most often refers to the evening while watching football in bars, which means that you need to respond promptly to tweets with the mention of @brand written after 15 hours.
How can Social Media Analytics help my brand?
Embedding smart social media analytics in many (or even all) aspects of your business can have far-reaching consequences.
- HR professionals can use Analytics to understand the mood of the company's brand and adjust their hiring policies.
- The development department will be provided with the views and wishes of users about what features and features should be added to the company's product.
- The customer service will determine the times when the largest number of users need technical support.
- Marketers Social media analytics is needed to understand the effectiveness of messages and advertising, as well as to compare the results with the effectiveness of competitors
- In the sales department, Analytics will allow you to understand what are the most frequent problems experienced by potential buyers and how to solve these problems.
An important problem for brands is also the delivery to all interested departments and departments of operational information obtained from social media, and tools to work with it: these can be specialized systems, dashboard, alarm notification system, etc.
How is the analysis of social media?
For a start, the main thing is that you need source data.
Social media generates such a large amount of data (billions of messages per day) that it is almost impossible to work with such volumes without special programs and algorithms for sorting and grouping data into a form convenient for analytical work.
[ There are specialized companies that aggregate data from various sources of social media: American Gnip (bought by Twitter), English DataSift , Russian Brand Analytics . Aggregation companies can supply, upon request, the necessary data slices taking into account the language, search phrases, geo, etc. for analytical modules and systems. ]
When parsing data and further imposing other parameters, there are charts and diagrams that can be used for analytics. For example, a comparative diagram of all the brands of a particular industry.
Facebook and VK
Each social network has its own indicators illustrating successful activity, and the data from each social network can play a different role in your marketing strategy. For many countries, Facebook is the ideal network for building and maintaining customer loyalty, as well as expanding brand coverage. For Russia, VKontakte is an equally important network. Facebook is the most popular social network in the world (1.5 billion active users per month), and represents an important goal for brands working in social networks.
For such a large audience, brands should pay increased attention to previously inaccessible types of analysis. Yes, it would certainly be convenient to stop at counting the number of likes, but you should not give out a brand mailing to a “too tricky” advertising agency. The main goal is to establish contact between the brand and the public, where the chosen advertising agency plays only the role of an intermediary. And if the agency’s participation is really necessary, then you wouldn’t like your audience to leave you to the mediator? The basic indicators - the number of likes and reposts - these data are not particularly useful if they are not deeply analyzed.
There are several ways to get useful information from this basic data. For example, you can simply monitor the above parameters to ensure that your subscribers are “consuming” the content you generate. The higher the consumption, the higher the consumer loyalty. Even if readers do not purchase your product today, this information will help them make a choice in your favor in the future. To contribute to this, you should constantly monitor the indicators of involvement both at the level of this campaign and at the level of the whole industry in order to understand what kind of content is most in demand by the public.
From the perspective of current business challenges, data derived from social media has a wide range of uses. Advertising of current promotions and offers, announcements about the launch of new products and other activities allow to receive feedback, including interaction with readers on choosing a new brand name. Throughout the history of mankind, it has never been so easy to get the opinions of huge masses of users, and if you do not use social media in order to better know your consumers, then you miss a lot of crucial information.
Twitter data analytics
Twitter is very popular with customer support workers. Although it is also used to distribute content, the main distinctive quality of Twitter remains the simplicity and speed of receiving feedback. Many large brands even have a separate Twitter page, which is not led by advertising specialists, but by a customer service team. The data collected on Twitter will help you more to meet the desires of consumers, because the faster you respond to them, the better they will be impressions of your brand.
American Airlines is considered one of the most effective companies in user support, which is also facilitated by the response rate on Twitter. American Airlines responds to more tweets than any other airline — and they do it much faster.
Analyzing Twitter data helps you keep abreast of your audience, and also keeps you abreast of what is being said about your brand. It is certainly pleasant to count retweets, but it is much more important to understand what exactly caused such a reaction. The information collected in social media will help you to be aware of what people say about the brand, you can prevent or extinguish scandals at the very beginning, as well as be aware of the activities of competitors.
A metric such as Share of Voice (share information presence in the competitive field) can help to find out the amount of your brand presence in discussions of key topics and concepts - for airlines, for example, understanding who is often mentioned along with such terms as 'kilometers' , 'rewards' or 'points' will help coordinate changes in strategy to maintain a leading position in the number of discussions, and ultimately influence the choice of potential buyers in favor of your company.
Analytics data on Instagram / Pinterest
Digital marketing is expanding rapidly - instead of textual information and messages, it now includes more and more visual elements - photos and videos. Social networks with a focus on posting photos, such as Instagram and Pinterest, are rapidly gaining popularity among both users and brands. While Instagram is still thinking about ways to monetize the platform, Pinterest has already introduced features such as “Recommended Publications” and “Buy Button”.
With Pinterest, analysts can identify both potential customers and people who have already bought a product. Tools that help identify leaders of opinion will prompt the most relevant of potential partners and will help significantly expand the coverage of your brand.
Data Analytics on YouTube
Every minute more than 300 hours of videos are uploaded to YouTube. With more than a billion users and a wide range of advertising opportunities, this channel is an ideal intermediary on the way to the consumer. YouTube provides its own tape showing the success of your videos, but when it comes to competitors' videos, you often have to rely on chance.
The system for identifying video clips gaining popularity can greatly assist in creating relevant content. It is important for analysts to take into account the growth rate of popularity and compare it with the average value - then any excess can be considered a signal to action.
Another key indicator is the duration of the video. Although it may differ depending on the type of content being lit, there is some ideal duration for different topics, after which viewers tend to lose interest. For example, we found that in the hotel business, long videos perform better, conducting detailed tours, while in travel and tourism, shorter videos show the best result, in which only key places are shown.
Although an increase in the number of views is in itself an important goal, it is equally important to increase the number of subscribers to the brand’s channel. The growth rate of views can be extremely high, while the number of new subscribers will be much less. Do not forget to compare these two indicators, and also try to find averages for your industry.
Studying YouTube data will help increase the popularity of your channel, and at the same time monitor the state of affairs of competitors.
Social Media Analytics Job
The ability to in-depth analysis of data obtained from social media is an important skill demanded by large companies. Today, many firms are gathering teams of social media specialists, and one of the possible roles is a social media analyst - one who will not only collect data, but try to delve into it and understand how it can be used to benefit the company.
Just one search query on LinkedIn with the words ' social media intelligence ' in the job title or job description will perfectly illustrate how important this profession has become.
How to choose social media analysis tools
There are many different tools for analyzing social media. There are tools for simple “listening” to social media (which can call themselves monitoring tools), there are tools for analyzing (which can also be labeled as monitoring tools) and tools for benchmarking social media (adapting the effective functioning of a company to improve performance). And at the same time all the systems in one voice claim that they will help in the analysis of social media. A logical question: what to choose?
First you need to decide what specific information from social media you want to analyze?
- Development
- Product problems and technical support
- Analytics of successful ways of working with social media
- Evaluation of marketing and PR policy
Social media monitoring (“listening”) tools can help you with the first two items. Standardly, collection and monitoring is conducted not over the entire information field (this is done by aggregator companies ), but on a specific set of social network accounts, blogs, forums and industry sites. Tracked given keywords that you specify. Such systems view a dedicated set of accounts and sources, and group messages by keywords.
To solve the full list of the company's objectives, including such complex ones as improving the brand's marketing strategy, it is necessary to use advanced social media analysis systems.
A source
==========================
The article is large, fairly complete, including data collection issues, which is rarely mentioned, but very expensive, if we are not talking about a simple API like Twitter. The only thing that caught my eye was the absence of such an important topic for Intelligent Analysis as linguistics: automatic tonality (especially important for analytics of mass real-time events), clustering, automatic selection of entities (NER).
It is clear that a full-fledged social media analysis system should include high-speed linguistic modules. To estimate the cost of technology, we can recall the recent purchase by IBM of AlchemyAPI , which has a high-speed linguistic system for some Western languages, for $ 100 million.
For English and Western European languages, BrandWatch English is a good example of a complete social media analytics system; for Russian, German, French and Eastern European languages, Central Asia and Arabic regions, Russian Brand Analytics . For in-house systems, HP Autonomy can be recommended, but data acquisition from aggregator companies will be required.
Source: https://habr.com/ru/post/298150/
All Articles