– by ORKASH Labs, Copyright: ORKASH Services Pvt Ltd

Given the unprecedented growth and reach of social media, monitoring what a company is doing on social media channels and what the people are talking about its products and services, has a huge potential for competetive intelligence gathering.  The forecast for such monitoring and the consequent business potential, is without doubt exciting for any corporate. The case for increased usage of social media analytics is getting more compelling, with every passing day.

To this end, Orkash has put together this blog post, which is an amalgamation of figures and case studies.  If you are interested in competetive intelligence, simply read on.

Some Trends

A study was conducted by the Center of Marketing Research at the University of Massachussets, to identify how Fortune 500 companies are using Social Media, as part of their marketing and customer service strategies.  77 percent have active Twitter accounts and 70 percent have an active Facebook presence (both up by 4% compared to the previous year, 2012). These companies are not just promoting their products & services and creating a digital marketing campaign, but have identified Social Media for customer engagement, for example, as an effective means of interacting with disgruntled customers as well.

Further, HubSpot’s 2013 survey shows that Social Media produces almost double the marketing leads of trade shows, telemarketing, direct mail or PPC ( Pay Per Click, an internet advertising model).

Case Study 1 : Predicting Sales – Automobile Sector

Social Media analysis of the Indian automobile sector was undertaken. The activities of a well known international car manufacturer (Brand X) were monitored on various social media websites by usings ORKASH’s social media mining and analytics tool.  Over the course of a few days, various types of information were extracted from Facebook and Twitter pages. This included data extration on various metrics such as trends and demographics associated with the ‘Likes’ on Facebook pages of various brands, ‘mentions’ count of the car brand in social media conversations, influence networks of the top users, sentiment and comparative analysis of comments, location specific trends, etc. This was also done simultaneously for a rival car brand (Brand Y) to better understand the user sentiments related to various parameters, and the sales pattern.


Fig : Sales figures obtained from Open source


Fig : Mentions on Facebook. This is an Orkash Technology output

A direct correlation between the sales figures and number of mentions was found on a month-on-month basis. However, at the time of the study (July 2013), sales figures till the month of May 2013 only were available, with the subsequent month figures awaited. The pattern of the sales could however be predicted since the social media activity, in terms of mention counts, for these months were already available. Once the sales data was released by the car manufacturer, it was found that the sales of Brand X increased in the same pattern as predicted by the Orkash tool – there existed a direct co-relation between the quantum of social media activity and sales. Such information is of tremendous value to business.

The snapshot below, obtained from the Orkash Clustering Engine technology, shows the grouping of important keywords and phrases associated with the brand, as they appear on social media platforms. For example, the specific problems, complaints and opinions on these issues, forms clusters based on the number of occurrences, showing the inter-connections. This kind of visual representation gives a comprehensive overview of the most common topics being discussed about a brand and the inter-relationship of these topics.


Fig : A cluster identifying Brand X problem areas. Orkash Technology Output

The Orkash tool was also able to classify which of the negative comments were complaints, and further subdivide them into various categories like quality, post-sales service, customer service etc. The locations of the origin of these complaints, in terms of which dealership or workshop in which city had a recurring problem, could also be identified by the system. The information was retrieved through the contextual analysis of text and content of the comments.

Orkash technology

Fig : Orkash Technology Output

Case Study 2 : Competetive Public Engagement – London Olympics  

In the build up to the 2012 London Olympics, Adidas struck a reported £40 million three-tier sponsorship deal to be the official sponsors for the games and launched their #takethestage campaign. In a classic case of competitive garnering of mileage on social media, Nike quickly countered the Adidas campaign by creating an all new #findgreatness campaign.   Figures from Socialbaker’s CheerMeter revealed that between July 27 to August 2 there were over 16,000 tweets associating the keyword Olympic with Nike, while Adidas received 9,295 tweets in the same period. Adidas gained over 80,000 new Facebook fans during the Olympics compared to more than 166,000 for Nike. The biggest sporting event in the world was a great platform to showcase the brand and increase awareness, but, as the numbers suggest, the battle on Social media was won by Nike, even though it was Adidas that spent big bucks sponsoring the London Olympics.

Case Study 3: Response to Digital Marketing : Food and Beverages Sector

The Orkash Clustering Engine technology has emerged as a very powerful tool in identifying brand activities like promotional campaigns and events, upcoming products and accumulating user sentiments and producing clusters showing their interconnections. A representation of this sort enables the brand to quickly identify problem areas, apart from ascertaining the outcome of their marketing strategies. The screenshot below, of a popular Indian liquor brand, shows the clusters obtained from social media conversations, just after the launch of a digital campaign. Key phrases like ‘Global Beer’, ‘Just want to Drink’, ‘New Brews’, ‘High Prices’ have the largest clusters and the maximum connections. This information culled out from Social media, which happens to be the platform consumers first turn to, can be of huge value to companies producing consumer goods.

Digital Marketing

The Future of Social Media Analytics in India

With its huge ‘youth demogrpahics’ driving India towards becoming one of the biggest Social Media market in next couple of years, it is becoming increasingly essential for companies to not rely solely on traditional media for the purpose of deriving business  insights. The approach of mining information from Social Media pages and conversations and using analytics to create deep insights into the activities of the brand, the user perceptions and trends related to various parameters is increasingly becoming highly relevant.

Unlike many ‘first world’ markets, India has a huge social and cultural diversity that has a very significant impact on consumer behaviour and consumptions patterns. This makes social media analytics and intelligence collection even more relavant, but with the caveat that such analytics must take social, demographic and cultural context into consideration for meaningful trends and insights to be arrived at.

In conclusion, unlike global organizations, Orkash, being a 100% Indian company, understands the local issues and dynamics, which is supported by in-house existence of social, behavioural, political, analytical and technical subject matter experts. This enables us to develop the technology keeping the local context in mind.  Thus, social media analytics, customised to the Indian market, is what Orkash can deliver.

, , , , , , ,


– by ORKASH Labs, Copyright: ORKASH Services Pvt Ltd

This blog post applies the concepts of game theory to Left Wing Extremism (Naxalism) in India. Game theory has gained importance in recent past in the study of unconventional conflicts. It is a tool for framing and analysing scenarios of strategic importance. The business world uses the principles of game theory to analyse and make competitive sense of strengths and weaknesses of the players involved for optimal decision making with respect to competetion, negotations and pricing strategies.  Corporations have been using this concept to model the mergers & acqusitions, price wars, trade union negotiations, divisional relationships, market dynamics, strikes, product launches etc. in order to make informed decisions and strategic moves.

A ‘Game’ derives its uniqueness from its rules and the way it is being played, be it a price war between two competitors or an unconventional conflict between the State and the insurgents.

To that end, extensive form of modelling has been used in this post to understand the dynamics and arrive at an equilibrium point for the left wing insurgency.

Political Angle – Arriving at Nash Equilibrium

It is understood that a politician will win an election if he aligns himself to the majority aspirations and brings about good governance. In LWE affected regions, the consensus among a proportion of the people is of an anti-State nature, with a belief that the Maoists would some day be instrumental in bringing prosperity to them. Confronted with such a situation, the plausible alternatives available to the politician in order to maximize their chances for wining the election are:

Naxal Political Payoff3

1. He/she aligns to the majority aspirations and therefore supports the Maoists/Naxals, either directly or indirectly (given that Maoism ideology has a large vote bank in the form of an existing support base), or

2. Over the long term, he undertakes developmental initiatives and works for the upliftment of the people to create a genuine support base amongst the masses. (For this to happen the politician has to compete against the support base of the Maoists/Naxals)

In the first scenario, the politician will likely reap the benefits of political allegiance immediately. In the second premise, the process is long-drawn and requires a commitment to the community. The politician has to, through his continuous efforts of bettering their lives, win over the people of the community in the LWE affected region. It is apparent that the first scenario offers the quickest results with the least efforts – the best and more efficient outcome for a politician solely interested in electoral victory.

Now, assuming that two major politicians contest an election, the outcome can be predicted by modeling the situation through the Game Theory. If both politician A and B don’t align with extremists and instead undertake initiatives and set good examples of governance to generate a genuine support base amongst the masses, then either may win the election. They have equal chances. However, it is promoting prosperity and development over the long term and is better for society as a whole. Thus, they both get a positive payoff each, from this situation, and the total of their payoffs is the maximum in this scenario – signifying the Best Possible Outcome for society as a whole. We depict this with the payoffs “A:8, B:8”.

Alternatively, if one politician aligns with the extremists while the other doesn’t (A aligns B doesn’t OR B aligns A doesn’t) then the politician who aligns with them, will win the election in this short run. This is the maximum possible payoff a politician can get as he definitely wins the election, and a better option for the individual than the earlier one (A:8, B:*) described above. The politician who does not align with the extremists is sure to lose the election. Thus, he gets the lowest possible payoff, in this scenario. We depict these payoffs as below:

If A aligns and B doesn’t – A:10, B:0

If B aligns and A doesn’t – A:0, B:10

As we can see, neither of these scenarios are best for society as a whole.

The last scenario we must look at is if both politicians decide to align with the extremists. In this case again both politicians have equal chances of winning the election. However, the important point to note is that this scenario does not lead to prosperity and development in that society. Thus the total of the payoffs of both politicians in such a situation would be less than that in the Best Possible Outcome case. We call this the ‘No Regrets Option’ since neither politician can feel regret for not having played to the interest of the majority to attempt to win. We depict this by the payoffs A:5, B:5.

Now, to predict what is likely to happen we must compare these scenarios and look at what decisions the politicians are likely to make.

In India, the primary goal is election victory, while good governance and development initiatives are unfortunately secondary goals. A major driver that creates this situation is that cast and community act as a dominant force in vote banks alignments, and the divison in vote banks resulting from three to four way election contests (two national parties and one or two regional parties being in the electoral fray in most constituencies). As a consequence governance based politics has increasingly emerges as  a lower priority. Electoral victory – a short term goal, becomes the priority, forcing the political system to align with the supporters of predominant ideology that can add to the vote banks – here, LWE.

The same is depicted in terms of the payoffs of each politician in different scenarios. No matter what Politician B does, A gets a higher payoff by aligning with extremists (If B aligns – 5>0 for A, If B does not align – 10>8 for A; so A will always choose to align). The same is true vice versa that no matter what A does, B gets a higher payoff by aligning with the extremists (If A aligns – 5>0 for B, If A does not align – 10>8 for B; so B will always choose to align). Thus, with both parties interested in maximizing their own well being, they both come to the scenario where both align with the extremists. This is the ‘No Regrets Option’ and is the Nash Equilibrium of this game.

This is why we see the political system attempting to align with the supporters of the predominant ideology of a region, so that it can add to their vote banks.

Operational Angle – Playing the Game – Changing the Rules from Sequential to One time

Politicians are the key players in the Game being played by the State (the counter insugent) and the anti-State extremists (the insurgent). This has emerged as a sequential game – with one party winning at the times when the momentary strength of the other party is weakened. Each party may take turns in winning the game (or temporarily emerging as the dominat player) since both believe that they would be the ultimate winner they continue to play numerous cycles of this sequential game. However, as a result the sequential game is a long-drawn-out and process which would likely hurt resources – both human and financial. Changing the way the game is being played, from sequential play (repetitive game of strike and counter strike) to a one-time game would probably yield more results. Thus the ‘surge operations’ in Iraq and the Greyhounds’ counter insurgency operations in Andhra Pradesh do prove a point here. Insurgency affected regions can be  generally saturated with large number of troops for area domination and population/territory denial to insurgents to bring about a situation conducive for better governance, and  while at the same time relatively smaller numbers of special forces are used in ‘strike role’ to cause attrition on the insurgents.

However, it is to be noted that tactical operations are only a means to achieve a safe environment that facilitates the restoration of the functioning of the civil administrative machinery. The surge operations need to immediately be followed by developmental programs and the creation of employment opportunities for the local populace which would bring about economic prosperity in the region and hinder any possible future extremist infusion and propogation of extremist ideologies.

In a seprate study we have calculated the quantum of resources and budgets needed (including the force levels for the counter-insurgency grid in LWE affected districts), and the capacity building and timelines that it would entail. Suffice here is to say that this would require a minimum of five year plan to just create the required resources and capcity building, and that the resources needed are large scale.

Winning the Game – Creation of a Counter-Narrative for Changing the ‘RULES of the GAME’

Owing to the current nature of law and order management in the Indian state being pre-dominantly an incident response system, the administrative machinery and law enforcement agencies typically focus more on the zones affected by violent forms of extremism.  While certainly effective to some extent, this may not be the most appropriate manner in which to eliminate the extremist movement in its entirety. In order to avoid a possible revival of the movement, the support bases need to be dismantled, as these will otherwise remain fertile grounds for germination of the underlying protest movement of the insurgency.

A community welfare based approach, especially in the peripheral zones, counters the predominant one that has been propounded by the extremists in LWE affected regions – that of State apathy and indifference. The creation, adoption and implementation of a   counter narrative in the affected and surrounding areas would be best experienced through governmental initiatives that facilitate economic prosperity and development.

Case of Northern Ireland

The outside-in devlopmental approach tends to change the Rules of the Sequential Game. It invades, weakens and breaks the support base for the extremist movement, destabilising their hold on the population. It will encourage the groups within the local communities to gradually align themselves and their resources with economic growth and development, which in turn assists the State in its counterinsurgency operations and in quelling any future onset of extremism in the area.

The British counterinsurgency experience in the Northern Ireland insurgency (as also the Greyhounds’ example in Andhra Pradesh state) is a worthy example in the study of the effectiveness of adopting a counter narrative approach to defeat the extremist movement. Through a revamp of their operational tactics from a full-fledged military onslaught to an inclusive community based approach, United Kingdom achieved greater success in margenalising the Northern Ireland secessionist movement propounded by the Provisional Irish Republic Army (IRA).


The tackling of LWE mandates a multifarious approach, focussing on development and security related interventions. The Politicain’s role remains the key. It is the Politician that can set the agenda for the governance and developmental angle to be the pre-dominant form of counter insurgency. This would need to delve into a host of forces at play (in the Game), inter alia, unemployment, poverty, land acquisition, forced displacement, distress migration, propaganda etc. Security related intervention will encompass police organisational structures, equipment, specialised knowledge of terrain, intelligence and a host of other measures for community oriented policing. In pursuing these twin approaches, we need to change the underlying drivers of the insurgency, i.e the Rules of the Game; use of game theory based modeling and maangment can shed light on underlying dynamics and sharpen the decisions, which can be a game changer in tackling LWE.