.jpeg)
Social Listening – Social Analytics – Social Intelligence
.jpg)
S-L-A-I
Social Listening, Social Analytics, and Social Intelligence - are they the same or are they integral parts of a sequential process, a so-called continuum?
Quite a few pundits have discussed this question in their articles, blogs, and essays. The most controversial of the three is social intelligence; if you Google it you will find its Wikipedia definition on first position explaining that it is “the capacity to know oneself and to know others”.
Of course, the alternative definition, the one that was coined after social media monitoring and analytics became popular only appears on page four of Google search - which you would only know if you are me and you search for social intelligence. In this secondary context, this bigram* means: the knowledge and insights that organisations can extract from online posts (mainly on social media platforms) published by their customers and other stakeholders.
The precondition to get to actionable insights is to avoid “Garbage-in” during the so-called data harvesting process, and to use appropriate machine learning models to maximise the accuracy of brand, topic, and sentiment annotation.
But let us look at the three bigrams one at a time,*combination of two words.
Social listening is short for social media listening, and to some a synonym of social media monitoring. Most people use social listening as an all-inclusive term for all online sources from which online posts can be gathered or harvested. However, in addition to popular social media platforms such as Twitter, Facebook, Instagram, YouTube, we also can harvest data from blogs, forums, news, and reviews. There are other sources that are country specific, such as Weibo for China and VK for Russia.
It is of paramount importance to harvest for all synonyms and avoid all homonyms (garbage-in) which can be over 80% of all harvested posts. A homonym is a word that is spelled the same way as the keyword for harvesting but means something entirely different, which makes it irrelevant for the project at hand. The classic example used to explain this problem is: wanting to harvest posts about Apple the company but ending up with lots of posts about the fruit or juice; let alone Apple Martin who is Gwyneth Paltrow’s daughter and if not excluded will invite a lot of noise in the dataset rendering it not only useless but dangerous for the user!
Social analytics is what happens after the posts are harvested from the various sources and irrelevant posts are removed as “noise”. For data analysis to take place, accurate intelligence needs to be added to the dataset using machine learning and/or rules-based NLP methods. The most common annotations added to each post or relevant snippet within the post (this is a longer story that requires its own blog post) are: brand, sentiment, emotions, and topics/subtopics/attributes.
These annotations can be added for text in any language, and the accuracy sought after – measured in precision, recall and F-score – should be over 75% in all cases. It is even possible to reach F-scores that are over 95% with focussed and context related training of suitable machine learning algorithms.
Social intelligence is the wisdom discovered by exploring the intelligent dataset. You see, adding brand, sentiment, emotion, and topic annotations to a dataset makes it “intelligent” but in order to find wisdom or “actionable insights” a lot more than just accurate annotations is required.
For the time being, an intelligent data cruncher and a powerful filtering or drill down tool is still needed to explore a dataset and find the gold nuggets.
We like to think of the listening247 unstructured data analytics platform also used for social listening, as the Google Maps of big data. It enables a user to navigate in a maze of millions of online posts - or other documents for that matter - safely and accurately from A to B; B being the destination or in our case the actionable insight. The path to finding the actionable insight is oftentimes a data story worth telling.
What is your Social Presence Score (SPS)?
%20(1).jpg)
Most people prefer order to mess, hardly a surprising conclusion.
A score that enables ranking in multi-player environments provides order and the ultimate gamification. Ideally it should be a composite score. This is a score that combines multiple metrics in one all-encompassing index.
Gamification does not mean turning a serious activity into a game; it is using gaming techniques to provide participant motivation and make the activity more fun overall.
I do like to think of business as a game. Some people take it too seriously, to the extent that it has a pathological effect on them – it affects their health in a negative way. Business is not a life-or-death endeavour. Winning is fun, losing is dreadful (worse for bad losers like myself) but it is not the end of the world.
So, let us break down how the SPS is calculated and its benefits for brands and for individuals; starting with the business perspective.
Brands and their parent organisations are always looking for the ideal KPIs that will drive their performance. NPS was once hailed as the single metric a company needed to measure and predict its future performance. NPS stands for Net Promoter Score and it is produced via a single question to customers, usually delivered via a survey: On a scale from 0-10 how likely is it that you would recommend Brand X to your friends and colleagues?
All those who provide a score between 0-6 are considered detractors, whereas the 9s and 10s are considered promoters; the 7s and 8s are polite negatives or passives at best.
The calculation of this composite score is as follows NPS=((promoters-detractors)/all respondents) X 100, with scores ranging from -100 to +100.
A similar idea applies to the Social Presence Score (SPS) for brands; however, its calculation is not as straightforward. Several social intelligence metrics have been considered by our data scientists and our conclusion was to use the following:
Fig.1: SPS score of shampoo brands
The SPS can have a value between 0 and 1 (see Fig.1 above); it is calculated, for a certain period of time, as one aggregated metric for multiple brand posts; it can be offered as a single score from all source types, or for individual ones such as Twitter, Facebook, Instagram, YouTube or Tik Tok (see Fig.2 below).
There is a secret sauce that even if I wanted, I would not be able to adequately describe in a mainstream article, and that is the weighting of the above-mentioned metrics in the SPS. Not only that but also the entire process, starting before the online posts are annotated, to eliminate irrelevant posts due to homonyms i.e.clean the data and remove the noise.
The benefit of having a score like this as a brand is, not surprisingly, the ability to:
Fig.2: Shampoo brands ranked based on SPS by source
The motivation to track a social presence score is not that dissimilar for individuals. Influencers and other high profile individuals want to know how their personal brand is doing compared to others and where they rank.
The metrics we use to create the composite Social Presence Score for individuals are similar but not all the same as for brands (see Fig.3 below):
The benefits for a person to know their SPS are:
Fig. 3: Influencers ranked based on number of posted brand comments
It goes without saying that even if the SPS is perfectly synthesised with its composite metrics, the accuracy of the individual metrics included must be measurable and acceptable. It is always possible to reach over 80% accuracy for sentiment, topic and brand relevance annotations.
So, what do you think?
Would you like to know your SPS in relation to others?
I know I would; I miss having one since klout score ceased to exist.
Social presence score (SPS) comes to the rescue; it is the new single KPI for brands and individuals based on all online mentions - not just a sample – that can be used to measure the overall success of their marketing efforts.
The Complete Story of listening247's NSS™ Score and Its Strategic Imperative
As businesses seek to understand their standing in the digital conversation, listening247's Net Sentiment Score (NSS™) emerges as a pivotal metric. This proprietary formula quantifies online sentiment towards brands, transforming raw social media data and unsolicited customer opinion into actionable insights.
Let’s deep dive into the NSS™ Score and explore its origins, how it operates, and its significance in today's digital-first world.
The Net Sentiment Score™ was developed by listening247 to fill a crucial gap in social media analytics. In a landscape saturated with diverse opinions, it provides a standardised way to assess and compare brand sentiment. This metric is the result of advanced machine learning models that meticulously annotate sentiments as positive, negative, or neutral, ensuring a nuanced understanding of the digital conversation landscape.
Fig.1: Graphic of listening247's advanced AI annotating sentiments as positive, negative and neutral.
Simplicity lies at the heart of the NSS™'s effectiveness. By focusing on the balance between positive and negative mentions and considering the volume of discussions, the NSS™ offers a comprehensive snapshot of a brand's online health. This approach allows for the aggregation of vast amounts of data into a digestible, numerical format, empowering businesses with the clarity needed to navigate the complexities of online reputation management.
This seemingly simple calculation belies the complexity and sophistication of the technology behind it. listening247's algorithms analyse vast amounts of online content, from tweets and blog posts to forum discussions and reviews, employing natural language processing (NLP) and machine learning to accurately capture and categorise sentiments.
The process involves more than just keyword recognition; it delves into the nuances of language, picking up on context, irony, and even regional dialects to ensure the sentiments are accurately interpreted. The result is a score that ranges from -100 (entirely negative) to +100 (entirely positive), offering a clear, quantifiable measure of online sentiment.
Fig.2: Graphic of different online content displaying positive. negative and neutral sentiment
The implications of the Net Sentiment Score™ for businesses are profound. It serves as a vital indicator for assessing the impact of social media conversations on brand perception. Here's why NSS™ is indispensable:
Fig. 3: Graphic of the value that the NSS™ serves for assessing the impact that social media has on your brand.
The Net Sentiment Score™ by listening247 represents a significant advancement in the analysis of social media sentiment. Its creation marks a strategic move towards a more informed, data-driven approach to understanding digital conversations. As businesses continue to operate in increasingly digital environments, the NSS™ offers a vital metrics for navigating the complex dynamics of online brand perception. Through its precise, insightful analysis, businesses are better equipped to foster positive engagements, adapt to consumer needs, and ultimately, drive success in the digital age.
To explore the depth of insights that the Net Sentiment Score™ can bring to your brand, we invite you to reach out and request access to your brand health dashboard. The brand health tracking dashboard offers a comprehensive view of your brand's digital presence and sentiment, enabling you to make informed decisions that drive your brand forward.
Whether you're looking to enhance your social media strategy, improve customer engagement, or simply gain a better understanding of your brand's position in the digital landscape, our team is here to guide you through the insights our dashboard can provide. Contact us today to unlock the full potential of your brand's digital narrative.
The festive season is upon us, and Christmas 2024 promises a celebration that blends cherished traditions with fresh, modern interpretations. Using listening247’s Social Listening and Analytics, we analysed over 50,117 posts from Twitter, TikTok, and Instagram between 5th October and 8th December. Keywords such as Christmas Gifts, Secret Santa, Christmas 2024, Holiday Season 2024, Gift Ideas, Holiday Shopping, Christmas Shopping, Festive Gifts, and Holiday Gifts were used to categorise posts.
These posts shed light on the key conversation drivers and trends that brands can leverage to create unforgettable holiday experiences.
Holiday cheer dominates online discussions, with a significant 68% (34,358 posts) expressing positive sentiment. Neutral conversations accounted for 29% (14,384 posts), while negative posts were low at 3% (1,375 posts). This overwhelming positivity prompts brands to connect with customers through festive messaging.
With 36,816 mentions, occasions emerged as the leading theme this holiday season. Consumers prefer shared experiences over material gifts, marking a cultural shift towards valuing time spent with loved ones.
This trend highlights the importance of framing campaigns around moments that matter for brands. Think event-driven promotions, experiential pop-ups, or even hosting holiday workshops to tap into the emotional core of Christmas. By aligning products and services with these experiences, brands can foster a deeper connection with their audience.
The allure of winter wonderlands - complete with sparkling lights, frosty landscapes, and cosy settings - is pulling audiences. This theme reflects a universal desire for escapism, nostalgia, and the joy of immersive holiday environments.
Brands can capitalise on this by transforming store displays into magical winter settings or curating themed collections inspired by the festive season. Even digital spaces can incorporate "winter wonderland" shopping experiences using AR-powered (augmented reality) winter effects.
Christmas carols with traditional melodies are being offered in interactive formats like virtual sing-alongs, community carol events, and life performances. Carols act as a unifying force, bringing families and communities together during the holidays.
Brands can harness this trend by sponsoring carolling events, incorporating carol-themed marketing campaigns, or creating playlists to complement the shopping experience.
In 2024, exceptional customer experiences are the ultimate differentiator. Shoppers seek seamless, personalised interactions at every touchpoint, from festive packaging to in-store ambience and online convenience.
This was never about selling products but building an emotional connection with the customer. Brands should focus on creating memorable experiences, such as personalised gifting services, festive loyalty rewards, or interactive in-store events that leave a lasting impression.
Sustainability continues to shape consumer behaviour. From eco-friendly holiday decor to gifts with purpose, shoppers are drawn to brands that reflect their values. Many are engaging in charitable giving, seeking ways to give back through donations and community drives.
Brands can respond by offering sustainable products, recyclable packaging, and charity collaborations. Highlighting eco-conscious initiatives in holiday campaigns resonates with consumers and strengthens brand loyalty.
The season is no longer confined to a single location. Many are opting for unique travel experiences, whether snow-covered retreats or tropical getaways. At its heart, this trend revolves around the desire to connect with loved ones, bridging geographical gaps to create cherished memories.
This opens up opportunities for brands to market travel-friendly products, such as compact gifts, travel kits, or digital gifting options. Partnering with travel companies or creating holiday content tailored to travelling audiences can further amplify relevance.
Christmas 2024 is a season of connection, creativity and purpose. Consumers are looking for experiences that resonate emotionally, whether through the carols, the escapism of winter wonderlands, or the shared joy of meaningful occasions.
The message is clear for brands: adapt to these evolving preferences by focusing on experiences, sustainability, and personalisation. With data-driven decisions, brands can create memorable and meaningful campaigns - ensuring Christmas 2024 is one to remember for customers and brands alike.
As Black Friday 2024 draws closer, retailers worldwide are gearing up for one of the busiest shopping days of the year. With billions spent in a single day, it’s an unmissable opportunity to connect with consumers, boost sales, and build brand loyalty. Thanks to listening247’s Social Listening and Analytics, we’ve uncovered the key trends and conversation drivers shaping this year’s Black Friday, helping retailers align with consumer behaviour and stand out in a crowded market.
Over 15,719 social media posts in English from platforms like Twitter, TikTok, and Instagram between 20 September and 15 November were analysed, to provide a wealth of data on what shoppers are saying, what they want, and how they feel about Black Friday 2024. Keywords such as Black Friday, Black Friday 2024, online shopping, in-store shopping, gift shopping, and Christmas shopping were used to track and categorise posts.
The buzz around Black Friday remains overwhelmingly positive:
Shoppers are excited, and motivated by the promise of unbeatable deals and the chance to save on holiday purchases. Positive sentiment is around tech deals, personalised promotions, and early access discounts, while negative sentiment is minimal, often tied to delivery delays or website crashes.
Using listening247, we identified the dominant topics shaping consumer conversations:
1. Technology – 12,881 posts: Tech products remain the crown jewel of Black Friday, accounting for nearly half of purchases. Televisions, laptops, gaming consoles, and smartwatches are the most wanted items. Major retailers like Amazon and Best Buy are expected to lead the charge with deep discounts, making this a key category for shoppers and brands.
2. Price – 9,313 posts: Pricing is a decisive factor in consumer decision-making. With financial pressures rising, shoppers are thoroughly comparing deals, looking for the best value, and waiting for Black Friday to make high-ticket purchases.
3. Discount Deal Promotions – 9,224 posts: Shoppers are actively seeking pre-Black Friday sales and exclusive member rewards. Brands offering early access or time-limited offers are seeing increased engagement, highlighting the effectiveness of targeted, exclusive promotions.
4. Gift Ideas – 7,116 posts: Black Friday is a prime moment for gift shopping. Consumers are turning to curated gift guides and social media for inspiration, with jewellery, apparel, and tech products topping the list of preferred gifts.
5. Shopping Platform – 2,094 posts: More than half of shoppers prefer online platforms to avoid crowds and enjoy convenience. Social media-driven e-commerce is particularly popular with Gen Z and Millennials, who rely on mobile-first shopping experiences.
1. Tech-Takeover: Electronics are projected to dominate Black Friday spending, driven by promotions from retailers like Walmart and Target. Smart home devices, gaming consoles, and high-end TVs are leading the wish lists. The influence of younger shoppers, particularly Gen Z and Millennials, is steering purchases online, where seamless digital experiences win the day.
2. The Power of Price: Pricing remains king. Consumers have set aside budgets specifically for Black Friday, emphasising its role as a cornerstone shopping event. The focus isn’t just on discounts but on true value, with shoppers scrutinising every deal to ensure it’s worth the spend.
3. Discount-Driven Decisions: Early-bird promotions and tiered discounts are becoming standard. Successful brands are going beyond basic sales by offering “spend-and-earn” rewards or bundling discounts to entice shoppers. This maximises short-term revenue and builds customer loyalty for future events.
4. Gifts Variety: Black Friday is the ultimate inspiration hub for gift-givers. Retailers curating well-targeted gift guides see higher engagement and conversions, as consumers look for meaningful yet budget-friendly options.
Black Friday 2024 provides a golden opportunity for retailers to connect with their audience through strategic pricing, tailored promotions, and a strong digital presence. Here’s how to capitalise on these insights:
Black Friday 2024 is more than just a sales event—it’s a cultural moment that reflects shifting consumer priorities and behaviours. By understanding what shoppers want, why they’re excited, and how they plan to spend, brands can tailor their strategies to exceed expectations. Thanks to listening247’s powerful analytics, retailers have the insights they need to navigate this high-stakes season and come out on top.
Let’s make this Black Friday a win for your business and your customers.
“Pan metron ariston” (παν μέτρον άριστον) is a quote in ancient Greek which was coined by Kleovoulos o Lindios in the 6th century B.C. and means “everything in moderation”. Some believe that the original quote was “Metron Ariston” which means “moderation is best”. Whatever the quote, ancient Greeks believed that you should live your life choosing the mean and avoid the extremes on either side, as much as possible.
Talking about extremes, I have always been fascinated by continua, I think it’s because of the order they bring to chaos and complexity. Almost every ideology or idea that matters in life, can be expressed on a continuum. A continuum has two extremes - let’s think of them as black and white with many shades of grey in between.
Here are two more official continuum definitions which are quite similar:
I do not consider myself qualified to improve on wisdom that transcended centuries (26 centuries since Lindios said “everything in moderation”) but I do have an opinion about quotes that include the words “everything” or “nothing”, “always” or “never”; incidentally these two pairs of opposite words can be the extremes of two continua; very few things are absolute, this is why the quote “everything in moderation... even moderation” may be just short of genius.
There is no doubt that being an extremist has mainly negative connotations: a fascist, a racist, a sexist, a religious fanatic, a communist… There are also some other examples like “feminist” or “atheist” that would create a debate with certain groups - as to whether they have negative connotations - that I am cowardly avoiding to mention at this time (see how I did this :)?).
Let’s first review a few random continua to familiarise ourselves on what they could look like, and after that we will go ahead and discuss the usefulness of looking at an issue through the lens of a continuum. Take the eating continuum below for example, isn’t it amazing how many types of diets there are? It has an impressive 13 elements in addition to the 2 extremes; a total of 15 elements. Kangatarian (I bet you can guess what these people eat :)) is the one that cracks me up with cannibal being a close second! I am also intrigued by how vegetarians managed to be the mean nowadays, they have come quite far from being an extreme in the not too distant past. And in case you are not familiar with ahimsa fruitarians, they only eat fruit that falls off a tree and they call pulling a carrot from the earth murder!
The God continuum with probabilities on God’s existence is not as harmless as the eating one; it is one that has been the basis for so many debates, civilised and uncivilised - and when I say uncivilised I mean the killing type if you think of the Crusaders (even though in their case it was more of a “my God is better than yours” rather than about its existence).
The selfishness continuum comes straight out of the Vedanta Treatise, a Hindu approach to life.
The colour coding means red is bad and green is good for most people.
Disclaimer: this does not always represent the author’s opinion. We will discuss more the groupings or segments of continua in the next chapter.
The continuum below communicates a thesis of mine that most people disagree with. I believe that being a patriot is the beginning of a proverbial “slippery slope”. It could progressively lead to someone becoming a nationalist and then a jingoist which is what you have to be to vote for Brexit or for someone like Trump.
The nationalism continuum can be integrated with the selfishness one at the point of loves all humans which is another way to say world citizen. One can then make interesting connections and draw conclusions about love and nationalism.
Those of you who have read other articles of mine may be wondering what all this has to do with market research, social intelligence, customer insights etc. Well the nice thing about continua is that you can conjure one out of nothing about almost anything. Case in point, digital transformation is something closer to home for a company with a name like ours; listening247. It was a sensible name 10 years ago to communicate specialism in digital market research; today however, when almost everything is digital, a name like this loses its meaning. It’s like calling a car a horseless carriage when in this day and age it is quite obvious that a car does not need horses to move (unlike the 1920s when Ford T1 was launched). But I digress... if you replace physical with ‘brick & mortar’ then this continuum becomes about retail, and if you replace it with ‘analogue’ it could be about equipment.
For market research, physical could mean in-person or telephone interviews, whilst digital means online surveys or unsolicited opinions found on social media using social listening tools.
Nothing easier than creating a 5 point continuum. The one below is about ways of gathering the opinions of customers and other stakeholders. Asking questions refers to surveys and focus group discussions whilst listening refers to unsolicited posts of people online. The discipline of harvesting these posts and analysing them is what we call social intelligence and it is mainly based on machine learning models that annotate posts for topics and sentiment in an automated way.
When you take some time to absorb the 6 examples shared above, you will realise that not all continua are created equal.
Here are some ways to differentiate them:
1. both extremes are bad (ahimsa fruitarian AND cannibal)
2. both extremes are acceptable (asking AND listening)
3. One extreme is really bad the other is really good (fascist Vs world citizen)
4. The mean is a combination of the extremes (asking & listening)
5. The mean is just a standalone option that has nothing to do with the extremes (vegetarian)
So what are they good for? They are philosophical tools that can help organise thought, clear the fog, visualise relationships, pinpoint and explain movements and trends.
Ancient Greeks believed that you should live your life choosing the mean and avoid the extremes on either side, as much as possible. Is this a good principle to follow though? If we consider the various types of continua described in the previous chapter sometimes the best choice is to adopt one of the extremes, sometimes it is indeed the mean like our ancient progenitors preached.
Thinking about moderation, can one be too much of a world citizen or too loving for all creatures?
When a continuum describes progress over time it is more likely that the most recent extreme is the best place to be. Even so, living it in moderation is probably a sound piece of advice.
I do subscribe to the notion that life is not black or white, it is mostly grey. Most of our lives are lived in the grey, only very few of us live on the extremes - sometimes by choice, but mostly not. Extremists must always be on edge, in contrast to leading a happy life, laid back, going with the flow, accepting the things they cannot control. Do let me know how you feel about continua and “pan metron ariston” @listening247_CEO or via email.
If you enjoyed reading this article you may also enjoy:
‘How does market research rank on usefulness for human life’
As the countdown to Halloween 2024 begins, it’s clear that this year’s celebrations are set to echo a blend of nostalgia and innovation, thanks to insights from listening247’s Social Listening and Analytics. Over 25,727 social media posts across Twitter, TikTok, and Instagram have been meticulously analysed to reveal what costumes will dominate and the sentiments surrounding this spooky season.
Our data reveals a robust positivity surrounding Halloween, with a striking 71% of posts radiating enthusiasm. Here’s a snapshot of the sentiment breakdown:
• Positive Sentiments: 18,336 posts (71%)
• Negative Sentiments: 1,589 posts (6%)
• Neutral Sentiments: 5,802 posts (23%)
This positivity is largely driven by families looking forward to trick-or-treating—a time-honoured tradition that promises joy and a bounty of memories for parents and their children.
Our analysis has pinpointed several key costume trends, shaped by discussions on purchase intentions, celebratory occasions, and peer recommendations:
1. Sexy Costumes: Embracing boldness and body positivity, sexy costumes are trending as symbols of empowerment. This trend highlights a shift towards costumes that celebrate individuality and self-confidence.
2. Nostalgic Revivals: A wave of nostalgia is bringing back favourites from the '80s and '90s. Expect to see a resurgence of characters from cult classics like “Beetlejuice,” alongside beloved Disney icons such as Mickey Mouse and Cinderella, tapping into the hearts of both new and lifelong fans.
3. Budget-Friendly Finds: Economic uncertainties have steered the trend towards cost-effective costumes. With a spike in searches for “budget-friendly costumes,” it's clear that affordability is top of mind for many this season.
4. Musical Homages: From rock legends like Freddie Mercury to pop sensations like Madonna, and current stars like Sabrina Carpenter, costumes inspired by musical icons are allowing fans to embody their favourite artists.
5. Couples Costumes: Doubling the fun, couples are choosing to coordinate their outfits, showcasing their relationships through creative and complementary ensembles.
6. Eco-Conscious Choices: Reflecting a growing dedication to sustainability, there’s an increased interest in eco-friendly costumes and decorations made from recyclable or biodegradable materials.
The capability of listening247 to drill down into the specifics of social media chatter has not only revealed what people will be wearing but also why certain trends are taking hold. For instance, the popularity of sexy costumes correlates with a broader cultural movement towards embracing body positivity and self-expression.
Moreover, the detailed sentiment analysis provides brands with a clear understanding of consumer attitudes towards Halloween, empowering them to align their marketing strategies with real-world preferences and expectations.
As brands look to capitalise on Halloween, the insights provided by listening247’s Social Listening and Analytics offer a strategic advantage by unpacking the complexities of consumer behaviour and trending topics.
This Halloween, equipped with data-driven insights, brands are better positioned to meet consumer desires head-on, ensuring a celebration that’s as delightful as it is insightful.
In an increasingly digital world, social media platforms have transformed the way individuals connect, share information, and express opinions. Banks can leverage the power of social media listening to manage loan default risks effectively. By gathering and analysing online posts about the companies to which they have extended loans, banks can conduct continuous commercial due diligence and identify early warning signals when debtors encounter financial difficulties. This post explores how banks can utilize social media listening as a strategic tool to proactively manage loan default risks and enhance their risk management practices.
Social Listening & Analytics involves monitoring and analysing online conversations, mentions, comments, and reviews across various social media platforms. It allows banks to extract valuable insights and gain real-time information about the financial health, market reputation, and business activities of the companies they have extended loans to. By employing machine learning for natural language processing techniques, banks can effectively identify potential red flags indicating financial difficulties and anticipate loan defaults.
A. Proactive Risk Assessment:
Traditional due diligence processes primarily focus on pre-loan assessment, often failing to capture evolving risks that borrowers may face after obtaining the loan. By incorporating social media listening into their risk management framework, banks can conduct continuous commercial monitoring in addition to their due diligence during the loan underwriting or credit extension phase. This enables them to monitor ongoing developments, industry trends, and financial indicators related to their borrowers, providing a more comprehensive risk assessment. These non-traditional indicators of credit quality and borrower’s abilities to service their debts and obligations give a much fuller picture than simple financial statements which are backward looking and often don’t provide the full story. This alternative data can also be a better predictor of borrower behaviour than historical financial statements.
B. Real-Time Insights:
Social media platforms act as virtual marketplaces where individuals freely share their experiences, opinions, and concerns. By monitoring online posts about borrower companies, banks can gain real-time insights into their operations, financial stability, customer sentiment, and market perception. Any notable shifts, negative sentiments, or concerning patterns identified through social media listening can serve as red flags, prompting banks to investigate further and take necessary actions. As we have seen recently with the collapse of Silicon Valley Bank and First Republic Bank in the US, depositor sentiment played a striking role in their demise. Due to adverse online sentiment that spread very rapidly, customers and depositors caused a digital run on the bank that had never been seen or experienced before. In the case of Silicon Valley Bank, deposits were leaving at the rate of $1 million per second for 10 hours (or $41 billion).
A. Detecting Financial Difficulties:
Social media listening allows banks to identify early warning signals of potential financial difficulties faced by their borrowers. By analysing online conversations, comments, and reviews, banks can detect signs of operational challenges, supply chain disruptions, declining customer satisfaction, or negative market perception. These signals can help banks proactively engage with borrowers, assess their financial health, and take appropriate measures to prevent loan defaults.
B. Amplifying Existing Risk Indicators:
Social media listening augments traditional risk indicators with additional insights derived from user-generated content. For example, a decline in positive sentiment towards a borrower company may coincide with a decrease in revenue, an increase in customer complaints, or a deteriorating market position. By integrating social media listening into their risk management framework, banks can enhance their ability to identify and act upon early warning signals, thereby mitigating loan default risks.
“Effective Early Warning Systems (EWS) reduce loan loss provisions by 10%-20% and required regulatory capital by 10%”
Galytix paper in association with PWC
A. Leveraging Advanced Tools:
To effectively analyse vast amounts of online data, banks can employ machine learning and sentiment analysis tools. These tools enable banks to filter and categorize information, identify patterns and trends, and extract meaningful insights. By leveraging sentiment analysis, banks can assess the overall market sentiment towards borrowers and gauge the impact of external factors on their financial health.
B. Enhancing Risk Models:
Integrating social media listening insights into risk models can strengthen banks' loan default risk assessments. By combining traditional financial indicators with sentiment analysis and social media data, banks can improve the accuracy and predictive power of their risk models. This holistic approach allows for a more comprehensive evaluation of borrower creditworthiness and provides a deeper understanding of the potential risks associated with loan defaults.
A. Privacy and Data Protection:
As banks engage in social media listening, it is crucial to prioritize data privacy and protection. Banks must ensure compliance with relevant data protection regulations and implement robust security measures to safeguard the information collected. Respecting user privacy, obtaining consent, and anonymizing data if necessary are essential steps to maintain ethical practices. This is of particular importance when used for due diligence and for the monitoring of individuals and their transactions, as is required of banks by Know-Your-Customer rules and regulations imposed upon them by the authorities.
B. Noise and Information Overload:
The sheer volume of online information can pose challenges in effectively filtering and interpreting relevant data. Banks can employ sophisticated filtering techniques and analytical tools to address information overload. Machine learning algorithms and natural language processing can help identify key topics or themes, prioritize relevant content, and provide actionable insights to manage loan default risks efficiently.
By harnessing the power of social media listening, banks can conduct continuous commercial due diligence and effectively manage loan default risks. Monitoring online posts about borrower companies enables banks to gather real-time information, detect early warning signals, and anticipate financial difficulties. However, banks must navigate ethical considerations, prioritize data privacy, and address information overload challenges. When implemented strategically, social media listening empowers banks to proactively manage loan default risks, enhance risk management practices, and ensure more informed lending decisions.
“Banks that fail to improve their EWS will also face significant regulatory pressures. The European Central Bank (ECB) has highlighted the huge variation in the quality of early warning systems and how credit assessment at a micro as well as macro level is core to risk management and processing.”
Galytix paper in association with PWC.
In today's fiercely competitive business landscape, companies are constantly seeking ways to gain an edge over their rivals. Among the various capabilities that contribute to success, unstructured data analytics capability stands out as indispensable for survival in the face of intense competition. This post explores the significance of text and image analytics specifically and argues that no company can thrive without harnessing the power of these capabilities. There is of course also audio and video analytics to consider but once the tech is available to analyse text and images the rest can be handled with voice-to-text and image-to-text technology. More details on this below.
1. Uncovering Insights: Text and image analytics enable companies to extract valuable insights from vast amounts of textual and visual data. By employing sophisticated algorithms and machine learning techniques, businesses can analyse customer feedback, reviews, social media posts, and other textual data sources. This allows them to identify emerging trends, preferences, and sentiment patterns, leading to informed decision-making and strategic planning. Similarly, image analytics empowers companies to understand visual content, enabling them to recognize brand logos, product placements, and consumer behaviour from images shared on social media platforms. The ability to uncover such insights provides a competitive advantage by allowing businesses to stay ahead of the curve.
2. Enhancing Customer Experience (CX): Text and image analytics play a crucial role in enhancing the customer experience, which is a key differentiator in today's market. By leveraging these capabilities, companies can gain a deep understanding of customer needs, preferences, and pain points. Through sentiment analysis of calls, chats, emails and social media posts, businesses can assess customer satisfaction and promptly address any concerns, improving overall customer experience and loyalty. Furthermore, image analytics can identify visual cues and sentiment from images shared by customers, helping companies gain insights into how customers engage with their products or services. By proactively addressing customer needs, businesses can establish a stronger foothold in the market and build long-lasting relationships.
3. Competitive Intelligence: Text and image analytics applied on publicly available information online also serve as powerful tools for competitive intelligence. Companies can monitor competitor activities, track mentions, and analyse customer sentiment related to competitors through textual data. This information provides valuable insights into competitor strategies, product offerings, and market positioning. Similarly, image analytics can help identify visual elements associated with competitors, such as logos or brand imagery, aiding in assessing market share and brand perception. Armed with this knowledge, businesses can adjust their own strategies, differentiate their offerings, and better position themselves to gain a competitive edge.
4. Operational Efficiency and Risk Mitigation: Text and image analytics contribute to operational efficiency by automating processes that would otherwise be time-consuming and error prone. For instance, text analytics can automate the categorization and tagging of large volumes of textual data, reducing manual effort, and improving data accuracy. Similarly, image analytics can automate the identification and classification of visual content, streamlining tasks such as quality control or identifying counterfeit products. By improving operational efficiency, companies can reduce costs, optimize resource allocation, and respond quickly to market demands, ensuring survival in a competitive environment.
At listening247, we leverage voice-to-text and image-to-text technology to efficiently process all forms of unstructured data through our social listening and analytics platform platform. This enables us to label the data with custom machine learning models, ensuring the highest possible accuracy, regardless of the original language. In contrast, some vendors offering multilingual text labelling solutions rely on translating everything to English before labelling, which is not an optimal or accurate approach.
Lately, many individuals have inquired about how the listening247 sentiment and topic labelling approach compares to GPT-4 or Bard. The answer is: the listening247 approach is unequivocally better. For a less biased and more objective perspective, I encourage you to refer to this paper. Here is an excerpt from the paper summary:
“The preliminary study shows that ChatGPT and GPT-4 struggle on tasks such as financial named entity recognition (NER) and sentiment analysis, where domain-specific knowledge is required, while they excel in numerical reasoning tasks.”
This subject deserves its own article with a proper gap analysis between LLMs and the proprietary and custom ML models that listening247 creates.
Text, voice and image analytics have become indispensable capabilities for any company striving to survive and thrive amidst fierce competition. The ability to extract insights, enhance the customer experience, gain competitive intelligence, and improve operational efficiency makes these capabilities vital for success. Companies that neglect to harness the power of unstructured data analytics will find themselves at a significant disadvantage, missing out on crucial insights, falling behind competitors, and failing to meet evolving customer expectations. Therefore, to remain competitive in the modern business landscape, organizations must prioritize the adoption and utilization of text, audio and image analytics to secure their long-term survival.
This statement, which I have shared numerous times in previous articles, encapsulates the essence:
“Over 90% of all human knowledge recorded throughout history exists in the form of unstructured data. If your company solely focuses on analyzing and comprehending structured data, it implies that you are utilizing less than 10% of the available data to inform your decision-making processes.”
What exactly is AI?
