The Political Capital article was created at the request of the NIOK Foundation, in which we present in detail the basic functions and possibilities of community and non-community media monitoring tools that can be used to monitor narratives involving civil organizations.
What purpose can media monitoring tools serve in the life of an NGO?
For civil society actors such as nonprofits, advocacy groups, and community initiatives, social media monitoring tools can be a valuable tool for increasing their communication effectiveness and audience reach. In addition, the tools allow them to gain insight into the evolution of the attitudes of the public discussing matters and topics relevant to them, to recognize the defining communication trends and disinformation, and then to respond in time to communication challenges or crisis situations based on these. This is one of the ways for civil organizations and the civil sector in general to remain competitive in the currently radically transforming information and online technology environment, the main driving force of which is generative artificial intelligence applications for workplaces, individual skills, work processes and the economy in general. respectively provided by its revolutionary effect on services.
The software can provide answers to the following questions:
What narratives dominate your organization's cause?
What disinformation is circulating about your topic?
What characters support these narratives? Are there dominant characters? Who are they?
On which channels and platforms do narratives affecting the organization appear the most?
How does your own target group talk about your case? How does social listening help you understand your own target group?
What narratives emerge about organizations like yours?
How can social listening help your campaign?
How can you mobilize your own target groups along your cause?
Below, through practical examples, we present in detail how civil society actors can use these tools in their daily operations. In this article, we describe the functions of media monitoring tools, and we deal with their practical uses in a separate article.
Definition of media monitoring tools
Media monitoring tools are online paid or free software that enable the monitoring and recording of written and audiovisual communication in the most varied forms - such as articles, Facebook posts, comments or interactions - on the most diverse interfaces and technical platforms of the Internet, the data AI -based analysis and evaluation in reports. In this way, users or organizations can, in real time or based on historical data:
They can follow the communication of key media, news sources and influencers in order to better understand the evolution of narratives or public affairs that are important to them, gaining deeper insight into the attitudes of media consumers and the evolution of public opinion.
Based on the data, analyzes and reports provided by media monitoring tools, NGOs can better manage their brand, PR campaigns and conduct research.
In addition, they can act more effectively against disinformation affecting them or their sector.
Compared to less complex tools, such as Google Trends, advanced media monitoring software is not only capable of simply summarizing search trends over time or geographically, but also has functionality that can be specifically tailored to organizations or projects. In other words, the specific - preselected - platforms and news sources of these programs are capable of extensive news production with an advanced analysis toolkit that also uses artificial intelligence (e.g. trend, source, platform, emotion, interaction, audience analysis) simultaneously in several languages or in several analyze in real time across countries. They can present the results through diverse and interactive data visualizations or use them to trigger online responses (e.g. news alerts or engagement management operations).
Functions of the devices
The devices we present usually have three basic functions and two additional functions. The basic functions are related to the collection of data, their processing and the preparation of statements, while the two additional functions contribute to maintaining contact with users or clients, as well as the application of artificial intelligence. The functions and their use are explained using the examples of the Gerulata Juno and SentiOne media monitoring software.
3.1 Basic data collection settings ("Perspectives"/"Projects")
All software has a setting function that defines the initial framework of media monitoring, which is called "Perspectives" in Gerulata and "Projects" in SentiOne. Basically, depending on the research goal or project, they define the range of media and messages to be observed along the following parameters:
Search focus or "Search query"/"Keywords": one of the central, if not the most important, functions of data collection is the setting of search keywords and keyword combinations that allow us to search for specific messages, narratives, organizations, etc., and the discourses associated with them in the media space. If we do not set keywords, the search results will display the full range of data generated in real time or for the period we specify - we can further filter them, for example focusing only on the most viral or recent messages.
Time interval ("Date"): all software provides the possibility to limit the collection of media data by time using the "date" function. Historical or past data are usually available in the databases going back to the last 1-2 years - i.e. trends can be examined retrospectively for such a period of time.
Range of media or news sources ("Targets"/"Source"): by defining Gerulata's "Targets" or "goals" and SentiOne's definition of "sources", it enables the definition of a predefined range or lists of media necessary for monitoring. Delimiting them is especially useful if, for example, we want to compare news from "reliable" or "unreliable" news sources - software usually allows us to manually or with the help of operators add new media and news sources relevant to our data collection to the database. If we do not define the range of media or news sources to be monitored, the tools will use all the media available in the database on the day of use for our data collection.
Type of platforms or messages ("Types"/"Type of mentions"): for the type of messages or platforms, we can set which media content of which platforms we want to collect. In the case of Gerulata, we can respond to 13 message types, from Facebook posts to Internet articles to YouTube videos. It depends on the device as to which content of which platforms it makes available, for example one or two devices do not have access to the comments of Facebook groups. If we do not set the type of messages to be monitored, the tools will use all content types available in their database during the specified time interval for data collection.
Language: advanced media monitoring tools are now available in many languages. On the one hand, the language settings allow only the relevant content of the given language-specific public to appear as search results, and on the other hand, they provide an excellent opportunity for the implementation of projects or comparative researches covering several countries and taking place in several languages.
Other additional functions of the basic settings: depending on the device, the basic settings ensure the subtlety of data collection with other functions. In the case of Gerulata, these include the filtering of data or messages based on tags or tags ("Tags"), the number of interaction types ("Performance"), the type of audiovisual content ("Features") or deleted content ("Deleted"). SentiOne also makes it possible to filter search results based on specific words ("Included words"), authors ("Name of the author" or "Facebook author categories"), Internet addresses ("Website address - URL").
3.2 Analysis function
The data extracted or filtered by data collection can be analyzed using qualitative or quantitative methods and tools. Qualitative analysis means a manual analysis based on reading, which can be done either by displaying the type of content you want to read on the online interface of the devices - for example X posts, Facebook comments - or you can also download the same content as a file in Excel or CSV table format to the for further offline investigation. For quantitative analysis, on the other hand, there are many interactive tools available online, while downloading the data as a file is also optional here. Some of the analysis tools are either available directly on the interface of the data collection function, such as in the case of Gerulata Juno as the sub-function "Inspect" or "Inspektion" of Perspectives, or form a separate analysis module - for this in both Gerulata and SentiOne the "Analysis" or "Analysis" modules are used. The most important and most frequently used tools of analysis interfaces and modules are the following:
Summary of results ("Observed Data"/"Result summary"): depending on the software, summarizes the number of messages or content selected by the data collection and used for analysis, the numbers of interaction types associated with the content, the reached or estimated audience ("Views"/"Reach").
Ranking ("Ordered by"/"Sort by"): different types of messages can be ranked based on publication time and number of interactions, which also provides the opportunity to quickly review the latest and/or most important discourses and narratives.
Trend analysis ("Trends"/"Mentions in time"): trend analysis, i.e. the distribution of the occurrence of data over time, can be requested for a wide variety of data depending on the given device, or these trend display tools are already basically given and ready to use in the analysis modules. In Gerulata, trend analysis can be used to examine the temporal distribution of the number of messages or content analyzed, the number of interactions, top or leading sources, Facebook or Telegram groups, leading platforms, publication dates, sentiment analysis, and the evolution of words in the word cloud. to use. The trend analysis itself enables the numerical analysis of the evolution of the selected data type over time, the identification of outstanding "peaks" of data or news trends, and further quantitative or qualitative examination of the messages, narratives and sources or influencers driving the peaks. In addition, it is possible to compare the trends of different searches, such as, for example, the trends regarding the mention of the names of two NGOs.
Source and platform analysis ("Sources, Platforms"/"Sources breakdown", "Top authors"): media monitoring software automatically and by default displays the leading sources and platforms that reach the most content or interactions. In this way, it can be determined which sources, mediums or influencers and on which platforms the examined topic or narrative is presented or dominated, or on which surfaces the given NGO can improve its presence or the visibility of its affairs. Some software, such as Gerulata, also compiles a separate list of the best performing links ("Links") and domains ("Domains").
Audience estimation ("Views/"Reach"): the audience estimation function usually provides data on the size of the audience reached by the analyzed contents using some predetermined formula or artificial intelligence.
Sentiment analysis: sentiment analysis assigns a positive, negative or neutral value to the analyzed data based on the categorization of the positive, negative or neutral meaning of the words found in the messages. In Gerulata, these only include media content, but in SentiOne, they also include the emotional evaluation of sources. With the help of sentiment analysis, we can gain insight into the attitudes behind the topics and narratives that are the focus of our investigation, which in many ways determine the reaction of the audience and the number and type of interactions with the content.
Based on the sentiment analysis trend of SentiOne above, it can be established that the "Family is Family" campaign launched by the Rainbow Families Foundation in 2020, or the public discourse about "rainbow families" was generally calm and neutral in the year 2023. The only exciting moment was Prime Minister Viktor Orbán's visit to Pope Francis in April, when the Prime Minister and the media controlled by the government twisted the campaign slogan and interpreted it as referring to the traditional family model, causing the discourse to shift in a minimally negative direction.
Publication time distribution ("Publish Times"/"Tops days, hours"): the analysis tool displays the pattern of the temporal distribution of messages in a trend-like, daily or hourly distribution - pointing to the contingency or coordination of the investigated communication. The function can be used primarily in action against disinformation or in research to identify observed communication patterns.
Network analysis: with Gerulata's "Connection explorer" function, we can map the connection system between individual media content or mediums, the spread of false information, if they share each other's content - articles, posts, links, etc. - referred to or cited. This function is particularly useful if, for example, an organization wants to take action against mediums or influencers carrying out coordinated disinformation attacks.
Keyword cloud and hashtags ("Keyword cloud"/"Hashtags"): the keyword cloud visually displays the most frequently occurring words and phrases in the selected database, i.e. the more frequently a given word occurs, the larger it appears on the word cloud diagram. Through this function, the most important topic foci, narratives, and discourses of the examined period can be identified. Hashtags also contribute to the analysis by quickly selecting the most relevant content.
Tagging: Tagging allows us to classify or "tag" various media contents into predefined categories. This function is particularly useful for research, as it enables the organization of relevant media content, e.g. to identify narratives.
Alarms or warnings ("Alerts"): the function makes use of the real-time media monitoring of media monitoring devices to enable users to receive an immediate alert about an important change in the monitored communication. In this way, organizations can immediately become aware of sudden, negative or positive changes in trends, topics, narratives or audience interactions that are important to them. All this enables them to respond quickly to a wide variety of communication challenges, for example in crisis situations. SentiOne's "Alerts function" alerts you when the number of pre-defined message types with a high impact factor on specific media platforms exceeds a threshold we define.
3.3 Reporting
Some software makes it possible to summarize the data visualizations produced during data collection and subsequent analysis and the analysis outputs displayed on the interface in an online report or format. Thanks to this, the monitoring results become available quickly, in one place, in a clear form, so that they can be shared digitally, downloaded or integrated into a longer, offline report. Of the two software, only SentiOne has such a feature called "Reports", which includes, for example, brand, source and sentiment analysis. Reports can be downloaded in PDF or PPT format for further use.
3.4 Commitment or "engagement" management
The toolkit of media monitoring tools listed above enables NGOs to monitor reactions related to their organization, their campaigns and the causes they represent, or respond in a timely manner to questions or problems raised by their (target) audience. Engagement can be managed either through general analysis tools, such as monitoring comments or reactions, or more specifically through contact functions that enable direct communication with the platforms. Unfortunately, the two software we presented do not have the latter.
3.5 Use of artificial intelligence
Artificial intelligence (AI) has now become a central element of a significant part of the software, which supports the processing of the languages and language databases they manage in the background, and also contributes directly to the operation of some analysis tools. In Gerulata and SentiOne, these include, for example, the "summary of the most important events" that determine the trends, the definition of the "general focus of the discourses", as well as the summary of the clients' "negative and positive" opinion of the brand (SentiOne) and the emotional analysis (Gerulata). With other media monitoring tools, machine learning helps to recognize and predict new trends, or crisis management based on the successful detection of rapid information flow.
After reviewing the functions, we recommend our article, in which we give examples of the areas in which and how we can use such software.