Social Media Analytics

Impact of social media marketing on brand attributes for a travel portal

The travel platform client wanted to boost brand loyalty using social media. They wanted to understand the impact their social media marketing activities were having on the brand attributes. Building and maintaining brand loyalty was one of the central themes of their social media marketing activities. They also wanted to know the factors driving brand attributes, especially – engagement campaigns, promotions and activities on multiple versus single social platform.

The content of social posts drives brand loyalty

The scope of the study included users who made at least three posts every week on social media and only those users who followed the online travel brand.

Average age of followers was 28 with a high percentage being students

Popularity of content among first degree connection biggest driver

Participative, game based content drives maximum engagement (relevance)

The study identified that users avoid sharing social content that is promotional or seen as advertisements.

Try to sell or make connections?

The goal of any marketing activity is two fold – boost branding e.g. awareness, engagement, loyalty and generate sales. It is therefore easy for a brand to assume social media marketing is no different – and treat it as any traditional marketing channel. Other unique positioning of social media as a marketing platform is the dynamic nature of the platforms – vying for user’s attention. Is it better to grab user’s attention than trying to sell?

One of the challenges of measuring brand attributes on Social platforms is the difficulty in mapping likes and shares To sales. We proposed to extract the attributes and the emotions associated with these attributes to map them with the branding variables – like awareness, loyalty, quality of offerings. Appropriate sales proxy variable was identified to be used as dependent to measure branding.

Brand attributes on a dynamic platform

Brand attributes being driven by social media include engagement, awareness, relevance and quality.

We categorised the social media marketing posts based on content into – promotional, games, awareness, fun and participative. These posts were then scored on relevance, engagement and similar other metrics.

We then extracted the sentiment attached with each of these posts and the branding attributes to which each of these posts could be mapped.

Structural equation modeling was then utilized to measure the impact of latent, direct and indirect factors on branding attributes.

Popularity of content with first degree connections and relevance were the top two drivers.

Social Media Analytics

Deep insights from Vietnamese conversations on social media

The Vietnamese privately held bank was planning to engage with it’s fans on social media. They did not have a clue of the best social media channels to begin their social efforts. Besides, there are local forums that are more popular than Facebook and Twitter, given the cultural fit and the presence of more active and visible Vietnamese user base. They wanted to look at their competitors and identify the forums most suitable for a focused social media marketing efforts.

Social media doesn’t just drive opinion

The banking client used the analysis to generate insights – the most important being – social platforms are much more than opinion driving platforms.

Identified six social platforms – including four local and more popular ones

Reduction of 40% in turn around time (TAT) on user complaints

Competitor analysis and benchmarking at product and service level

The bank has a better understanding of the strength (and weaknesses) of each platform, and has used our analysis to build an automated reporting system for social media.

Non-English conversations. Local slangs.

The Vietnamese client is one of the Top 3 privately held banks in Vietnam. They are growing at a healthy pace and desire to engage with fans on popular social media platforms to boost brand visibility. They also wanted to look at competitors and establish themselves on platforms where competitors were already active and had a fan base.

There was one problem though – to build a manageable, scalable social media marketing function, they needed to utilize technology driven solutions. And all major solutions did not provide support for Vietnamese language – the primary language of communication on social platforms in Vietnam. Another significant problem was the use of slangs, spellings and language that was not standardized and could not be processed through an off-the-shelf solution. A third, but minor problem was the popularity of local social platforms.

One more window to look at the world…

The Vietnamese banking client was able to leverage our strengths in designing the Vietnamese natural language processing (NLP) based text analysis solution from scratch.

A major challenge for us was the identification on non-standardized language – slangs, local context and misspellings. We navigated these through local language experts. As a solution, we developed an automated monthly reporting system that allows the client to:

a. Identify posts and sentiments associated with the posts across global and local social platforms

b. Monitor the activities of the competitor

c. Extract product and services attributions in the posts and monitor sentiment on those

Social Media Analytics

Sentiment analysis for a retail brand to measure customer perception

The retail brand has a very active and engaged fan base on Twitter and Facebook. They keep the fans engaged with regular fun and participative videos and social posts. The users themselves create a lot of posts discussing various offers, events, products and brands. They also provide some feedback on products like comfort, pricing and sometimes tag competitors in their posts. The client decided to organize the user generated content to generate actionable insights.

User generated content is an insights goldmine

The client used social media mostly as a customer service platform addressing user complaints. They now use social media for valuable insights on their customers.

The brand and its top competitors have no difference in sentiment distribution

Negative sentiments are associated with variety and price

There are significant number of posts mentioning more than one brand (36%)

The analysis was setup as a monthly report that enables the client to monitor opinion over time and analyse the impact of corrective strategies.

High volume. High velocity.

The retail brand is doing a good job engaging users with participative and fun posts. They have an active fan base, who share their opinion and tag the brand. The retailer was making limited use of the data being generated on social media to address user complaints and queries. Given the high volume of social posts, it was impossible to look at them manually, and a technology driven solution was needed to extract the social posts and extract sentiment.

A big challenge in setting up a sentiment mining solution was the volume of data being generated on a daily basis. The client has a limited need for a state of the art data Infrastructure till date, as they mostly dealt with sales and marketing data. They were not big on digital marketing. With the shift in focus on social media data driven insights, they opted for a cloud based solution.

He, who has a longer bait catches the bigger fish.

Sentiment analysis helps extract the opinion and polarity of emotions from social media posts.

With state of the art supervised learning approach, the retail brand is not able to extract valuable insights including:

a. A better understanding of their brand strength and where they outperformed competitors

b. Insights on negative opinions associated with price and the variety of product offerings

c. Insight into the path to purchase and how users compared similar products from competitors to make a purchase decision

Sentiment analysis is the first step in using social media to generate user level and aggregated insights.

Social Media Analytics

Predictive analytics using data from social platforms

Social Media allow users to create, publish and share digital content. Overtime, specific platforms have evolved for specific activities – Tiktok and YouTube for video content, Instagram & Pinterest for visual content etc. Since there are diverse people sharing their opinion, likes, dislikes and experiences on social media, there is an aggregation of different viewpoints. It is worthwhile to note that social media viewpoints have a shelf life – as viewpoints may change over time. And this provides us with another opportunity – to study the propagation of viewpoints and identify people/organizations that make them popular.

If the right data is being collected at the right time, the social media data can provide useful insights and predictions. Such predictions are useful across business functions – marketing, sales, after sales service etc. Some industries have more intuitive applications of social media e.g. Hospitality, Media, Governance. However, if done right, social media predictions are useful for all businesses.

Social Networks

Let us refresh our understanding of Social Networks.

A structure comprising of people or organizations, usually represented as nodes, together with the relation between them – represented through the link between nodes. The relation between nodes could be explicit e.g. classmates, employees of a company or implicit e.g. interest for NBA, brand advocacy for iPhone. Below is a generic representation of a social network:

A social network is scale free and the degree of distribution follow power law. Social networks on social media platforms are created to share and discuss user generated content.

Social as a Media Channel

Social Media differs from traditional media in many ways:

1. Anyone can publish content at a minimum cost

2. Provides opportunities for two way communication between brands and customers

3. Opportunities for word-of-mouth promotions and instant feedback on shared content

4. Different social platforms offer different opportunities e.g. LinkedIn for business communications, Blogs for descriptive, information dissemination based content, Twitter for quick updates and so on

Social Media and Predictive Analytics

Even today, humans are better at interpreting social content compared to machines. That is not to say we haven’t made progress. Within the last decade, we have been able to look at text data far beyond simple sentiment analysis. Image & Video based content can now be processed through machine learning algorithms to generate insights that are very useful for business – and would not be humanly possible. We discuss some of the use cases of predictive analytics with social media data.

1. Sales Prediction: The volume of related blogposts have a high degree of positive correlation with sales rank. With the growth of digital and e-commerce, this relationship will get stronger. It is however important to note that predicting tomorrow’s sales numbers based on blog mentions today is a complex problem to solve. Blogs (and other social content) have a reach mechanism as well as carryover, impact and lagged effects.

2. Sentiment and stock market performance: Efficient Market Hypothesis (EMH) and random walk theories try to establish that stock prices are unpredictable. However, the movement of stock prices can be predicted with a fair degree of confidence. Multiple studies have established a high degree of correlation between the stock price movements and the sentiments on finance and news boards

3. Virality Prediction: Social Media platforms have huge potential in creating (or destroying) brand reputations. People pay more attention to their peers’ opinions and views. And the ease of sharing on social could play havoc to a brand’s reputation – atleast in the shorter term. It is therefore important for brands to know the virality potential of social posts, and then either use it to ride on popular content or mitigate brand risks at the earliest.

I have purposefully not touched on few other predictions being utilized on social content – movie performance predictions, election results, economic confidence etc. as those are very specific use cases. Prediction models built on social media data pose additional challenges of unstructured data – text, image and video. In addition, a lot of interactions are not in one specific language, as well as do not adhere to grammar rules in most cases. Having said that, social media has created a new way to collect, process and utilize the user generated insights with low cost and high accuracy.

Social Media Analytics

How can Financial Services use Social Media

Almost all financial services (FS) have some form of social media presence. However, majority of them are not making the most of it, and focused on generic marketing content through their channels or responding to tagged customer communications – usually unsolved queries through traditional customer service channels.

Strategic use of social media can provide valuable insights about your audience – not just the customers. It has the potential to impact audience’s view of your products and services, reputation as a brand, marketing of products and services, risk management, competitive analysis and educate people about financial matters

Why Financial Services firms need to Social?

1. A significantly high percentage, if not all customers and prospects are already on one social media platform or the other

2. There are conversations about your brand – regardless of whether you are present (and active) on social media or not

3. Social Media provides you the opportunity to have a conversation with your customers and prospects – one to one

4. If done right, social media can reinforce your brand persona

5. Provides customers with a 24*7 access to your services – helping build the trust that you are always there to listen to them

6. Allows you to monitor competitor activities

Social media is all about conversations, and with the increasing penetration and trust on digital, more and more customers (and prospects) feel comfortable interacting with the brands on Social. But before you embark on your social media journey, you need to understand the beast that social media is. It is like taming a wild horse, and if not done right, has the potential to harm your brand reputation.

The team at Xtage Labs has worked with multiple financial services institutions – mainly banks, and helped them navigate the tricky waters of social media. We present ways social listening and drawing insights from social contents can help financial services:

1. Keyword and Hashtag based listening: The simplest to implement, keyword and hashtag monitoring enables FSs to know what is being talked about. With some deeper implementation of sentiment and polarity monitoring, along with summary reports, social listening allows you to know what is happening

2. Marketing: All social platforms allow targeted promotions. Digital as a marketing channel is getting more budget allocated in every planning cycle. Marketing on social platforms is different, as it can provide several targeted promotion options – by age, gender, geography, interests etc. With these information, you can offer specific products and services to specific segment of customers, and can reach them through a language that resonates with your audience.

3. Customer Service: There is a reason customers prefer raising complaints on social media – they do not have to wait in a queue or find the right executive to handle their issue. They trust their brand to hear them, and do everything possible at their end to solve their problems. And when brands resolve their customer complaint on a platform, say Twitter, you have the opportunity not only to improve satisfaction level of a single customer – but also to show the customers in the complainant’s network how quick, efficient and customer centric you are. Customer servicing through social media isn’t a risk, but an opportunity to improve customer satisfaction and brand equity. But for this to happen, you need to invest in text analytics driven complaint allocation and automated monitoring of resolution rates

4. Conversation Clustering: Or topic modeling as some may call it, allows financial services firms to identify the multiple themes of conversations happening at a given time. Identifying these themes, and then engaging with your audience with the right content will allow you to converse with the audience and foster a meaningful relationship with them. It also provides insights on your customers from multiple products and services, and those of your competitors

5. Risk Monitoring: Social media is a double edged sword. If you are not on top of the social game, it may backfire and do more harm than good. It is therefore important to have a robust social risk monitoring mechanism before you delve into the world of social media. And social risk monitoring starts with listening, and being able to parse the un-ending stream of social posts to generate insights and handle risks efficiently

Social media, if used correctly, is one of the best channels to grow influence organically and connect with people in novel, personalized manner. It has the potential to multiply your brand reputations and improve customer experience. People are on social platforms to interact with their friends, relatives and share content of personal interest – and if they allow you to interact with them means they are allowing you a peek into their personal space. It is upto you as a brand to maximize the opportunity being presented to you everyday – to be viewed beyond yet another business just focused on deriving profit out of them

Social Media Analytics

Measure value driven by influencers using impact analysis

The haircare brand was running aggressive always on campaigns across social platforms like Facebook, Insta, Tiktok. They had engaged global as well as local influencers to boost awareness, engagement and consequently the sales of different products and the variants. They wanted to look at metrics beyond likes, comments and shares and measure the impact of these campaigns against other investments on digital marketing on make a more effective campaign plan

Right insights. Better campaign outcomes

Through our impact analysis study by measuring the correlation of influencers’ reach, engagement and post attributes, we generated following insights:

Global influencers generate ~ 2.4x more (brand) awareness than local influencers

Global influencers generate ~1.7x brand engagement, but with about 63% lower purchase intent

The RoI on local influencers is 1.8 times that of global influencers

The insights from our study surprised the campaign team. They have reaped the rewards through better marketing performance using those insights

Proactive use of MMM to plan future spends

Our client was allocating campaign spend on influencers based on general gut, and mis-aligned KPIs of likes, shares and comments. They were spending more on more expensive global influencers, who were generating more likes and equivalent metrics. The implicit driver being more fans because of global reach. The local ones, having lower fan base were being less favoured.
What they needed was better KPIs to measure if the More expensive global influencers were driving the KPIs that mattered i.e. brand awareness, brand engagement and consequently higher sales in the long term. With the custom analysis, Xtage Labs team was able to overlay the impact each influencer type (global vs. local) and each influencer was having on the branding attributes and the return on investment (RoI) being generated

Influencers’ impact is more than just likes, shares and comments

If you were to ask a person whether celebrity endorsement nudged them to buy a product – the popular answer would be a resounding ‘No’.

And that is what we discovered in the study we executed for our client. While being able to generate

lower numbers on the vanity metrics – likes, shares and comments – they were more effective in driving brand awareness, brand engagement and ultimately sales.

The icing on the cake is – global influencers are more expensive to onboard. In the case of our client, they could onboard more than 20 local influencers on the onboarding one global influencer.

The client has used the insights well, realigning their influencer budget and focus, thereby generating better RoI and lower investment on influencers and more on other digital marketing avenues.