Social Intelligence in the Age of Infobesity - A Verifeed White Paper

Marketing with Social

Intelligence in the

Age of ‘Infobesity’ A White Paper

August 19, 2013

What is the “gold standard” of information?

Think about what makes information valuable to you, your brand or business.

If you want to improve your bottom line you need access to data that is trustworthy, accurate, and from credible sources. It also needs to be relevant, contextual, and timely. You also want to be able to access it before anyone else so you can act faster and more confidently than your competitors.

Verifeed White Paper | August 19, 2013 | w.ww.verifeed.com

Mining for ‘information gold’ in the age of Big Data is as big a challenge as the exponential bigness of the data itself. There are 300 billion gigabytes out there at any moment – that’s three times the grains of sand on earth. Now apply Moore’s Law: The data is doubling every two years.How and where you find those tiny grains of insight and actionable intelligence is being turned on its head by social networking, mobile and the processing power of the cloud. If you’re not harnessing social media to gain insights into customer preferences, motivations and behaviors you’re ignoring a goldmine of information.


Social Data comes from real-time large-scale interactions and conversations across social networks, search, e-commerce, and media.

It’s put the customer in the drivers’ seat. There’s tremendous value in the billions of social conversations to help CMOs, brand managers, sales and customer service reps maintain and grow margins by better understanding and anticipating the needs and motivations of consumers. Facebook’s influence on buying decisions has doubled in the past year, reports Edison Research’s “The Social Habit”. Every day, there are 2.7 million ‘likes’ on Facebook and one million new Twitter accounts created. People are talking and their influence is powerful.


Many companies are overwhelmed by the challenge of clear signals accurate, actionable information amid all the repetition, irrelevancies, and rumors on Twitter, Facebook, other social networks, blogs, online comments and reviews. The good news is that social data is being put to use in innovative, profit-making, and automated cost-saving ways.

Yet so many valuable insights get lost in the deafening noise.


Google can predict flu outbreaks by cross-referencing where people search for information on symptoms, Dell was alerted to strong negative sentiment on social media to a published price and as a result changed the global price within 24 hours to avert an adverse financial impact, and the NFL uses social login technology to deliver personal emails based around individuals birthday’s and favorite teams with a call to action to buy a personalized jerseys.

The good news is that social data is being put to use in innovative, profit-making, and automated cost-saving ways.


.

Social Intelligence in the Age of ‘Infobesity’

The brands and businesses that will win hearts and minds in the ‘Age of Infobesity’ will be those that excel at unearthing and analyzing real-time relevant insights their customers and would-be customers already provide minute-by-minute across all social networks and social media as they talk about products, issues, news events and each other.


.

Top brands and many businesses have gotten good at pushing out their messages on social networks. Many now also generate contextual content to drive engagement with Facebook likes, Pinterest pins, Twitter followers, and favorable blog post mentions. Many have been aided by social media management platforms, analytics and technologies that enable them to track mentions, quantify the impact of their messaging, and identify the ‘influencers’ they can target to amplify their messages. Some of the savvier companies have integrated social data into their CRM to tie social marketing efforts directly to sales results.

Social is not only about pushing out your brand message…

…It’s about pulling in your customers’ insights


Now there’s an added imperative to pull in social data to provide valuable insights into how a campaign, product or price point is being perceived, what people think about a brand (or a competitor’s brand), or events and developments people are talking about that may impact sales, improve a marketing campaign, or effect a product.

Social data can also help identify the people who may be interested in purchasing a new product or service, or may influence others’ purchasing habits. It can also drive consumer loyalty and viral ‘word of mouth’ through the development of deeper personal, authentic relationships with targeted consumers.

The companies doing it best integrate social data in a ‘push’ and ‘pull’ strategy across all marketing channels with a ‘customer-first’ ethic it becomes that necessitates combining real-time engagement, content creation, data management, and real time actionable analytics.


But how trustworthy and useful is the social data they rely upon?

A 2012 online survey of the CMO Club, an organization of 600+ leading B2B and B2C companies, found that almost half (47.3%) of all CMOs have used social data to make predictions and forecasts, and nearly 9 in 10 say the social analytics have influenced their decision-making. Most analyze trends and sentiment to help their teams craft more effective consumer messaging. Many also say they use data to inform product development (35%), brand management (59.6%), sales strategies (36.8%) and customer service (36%).

The same survey found that CMOs believe social intelligence is “at least somewhat effective” in indicating trends or patterns that may impact the overall business, consumer demographics and psychographics, consumer sentiment towards products, product lines and brands and companies themselves, as well as the influence of individuals on purchase decisions.


The phrase “at least somewhat effective” highlights a key problem: as any executive who has interacted with social data knows, the challenge is finding – instantaneously – relevant information you can bank on as fact.

Who wants to make a multimillion dollar decision on unreliable information?


Sentiment engines can only tell you so much

More and more companies are mining the entire Twitter fire-hose to discern customer sentiment from the 500 million+ Tweets a day. But algorithmic-driven sentiment analysis has been notoriously unreliable: It is near impossible for a computer to discern the intent of the post, catch slang, put a comment in context of a longer evolving thread of a conversation, or even catch a double negative.

There are so many unpredictable variables. As Francis D’Orazio, CIO of social analytics firm FACE Group, put it, “the sentiment of a sentence only rarely lies in the sentence itself and is instead rooted in the cultural context around that sentence.”


That’s why so many social data platforms and providers still need to rely on humans as a critical part of the fact-checking equation. For instance, Datasift, the world’s largest reseller of the Twitter fire-hose, uses the crowd-sourcing marketplace Amazon Mechanical Turk to test results and three people per tweet/comment to help improve their sentiment algorithms.

“Even with humans you only get an 85% accuracy rating because views differ on whether a word is positive or negative, and one person’s view of a subject may be different to that of a specialist,” Datasift CTO Nick Halstead recently told the Guardian newspaper.

Data without context is worthless.


Though algorithmic and computing processes such as unstructured data search, unsupervised machine learning and natural language processing (NLP) text analytics techniques are improving steadily, none has yet come close to replicating the intuitive pattern recognition capability of the human brain.

How well can algorithms ‘fact check’ social data?

The need for real people to augment automated analysis of social data stands in the way of more effective and efficient filtering of social intelligence. It’s also expensive to invest in labor-intensive solutions, assuming you can even find all the good data scientists you need.


A recent survey by the Direct Marketing Association found that many CMOs, brand managers, customer service reps, and market researchers say they’re unable to handle the vast amount of data their businesses are now collecting, and agreed they needed to hire data scientists to make better use of social data.

Yet a Big Data Executive survey found 64% of respondents said it was ‘very difficult to find or hire data scientists’, while a report from the McKinsey Global Institute projects the U.S. needs up to 190,000 more workers with ‘deep analytical’ expertise and 1.5 million more data-literate managers.


Automating Actionable Information: Why the sum of the parts is greater than the wholeThe more accurate the information, the better the business decision a company can make.

Assuring accuracy with automated processes is the Holy Grail. The quality and trustworthiness of the ‘what’ is only as good as the credibility and honesty of the ‘who’. Discerning what is accurate is a complex process of qualifying sources for expertise and credibility, and exposing hidden agendas.

Only 58% of people trust

brands


Identifying the right ‘influencers’ to target is a no-brainer for any brand, business or cause seeking to amplify their message and expand audiences, followers and customers.

Developing effective and efficient automated systems to do so is not as easy.

Everyone wants to know with accurate specificity whom may be their best social ambassador to drive word of mouth marketing. Study after study shows that people are more likely to trust information that comes from their own social circle, from a trusted source.

A 2010 survey by eMarketer found that consumers trust people “like them” who have used a product or service, nearly 12 times more than information direct from the brand or service provider. Only 58% of people trust brands and businesses, and for the financial services industry, the pain is even more acute: Only 50% of people trust banks and other financial firms, according to the 2013 Edelman Trust Barometer. Verifeed White Paper | August 19, 2013 | w.ww.verifeed.com

Qualifying the right ‘trusted’ influencers depends on what behaviors you want to influence. Assuming you know the demographic and psychographic attributes you want to pursue.

A Klout score can tell you how active someone is on social networks, how many followers and friends they have, and how often their posts get re-tweeted, liked or commented on. Useful information, but it does not give you the whole picture. Is an individual with a high Klout score really influencing anyone or just posting a lot? What is the targeted ambassador’s expertise, their credibility over time? Are their posts accurate, helpful, do they have a pre-ordained or hidden agenda? You need to know this information, but the systems that process this information in real time are still in their infancy.

The ‘who’ needs to be weighed against the ‘what’, cross-referenced with multiple sources with varying perspectives, and analyzed in the context of the cumulative knowledge only an expert can provide.


Data must be put in context to ascertain its meaning and its trustworthiness. Think about what an investigative journalist has to do to be first – and right – about something that is truly ‘new’ or previously undisclosed.

A journalist talks to multiple sources, ideally in person so they can discern whether or not the source is being ‘economical’ with the truth. They have to talk to people of all different viewpoints and persuasions, and weigh their comments against agendas they must expose. They have to question each of these sources in an uncompromising way, they have to have the cumulative knowledge and research to read between the lines and know what questions to ask. And to discern what is truly important and new.

It’s not easy. A lot of time, a lot of worn shoe leather, and a big price tag because its labor intensive.Verifeed White Paper | August 19, 2013 | w.ww.verifeed.com

The massive computing power of the cloud is giving anyone the ability to filter social conversations in real time to enable sourcing of virtually unlimited amounts of information, but automated verification of the accuracy and truthfulness of the is a tough code to crack.

And assuming you know what you’re looking for, you can find it, to a point – but not as efficiently as you should be able to.

It’s even harder to automate discovery of new information you want at your fingertips in the moment … before you even know you need to search for it.


The unknown unknowns

Assume for a moment you are the CMO for a company. Variously you want to know:

:: Who’s tweeting about my brand? My competitors’ brands?:: What do they think about my brand and my competitors’ brands?:: What are people saying about organic dog food, specific ingredients and other topics that may influence my marketing strategy or product?:: How many people are interested or passionate about the space I am in?:: Who are the ‘influencers’ or thought leaders, and who are the followers?:: What are the things they care about most?


97% of relevant social posts and conversations

are missed by brand marketers

All good questions, and if you are relying on Twitter’s API, here are a few things you need to know about the limitations on what you will find:

:: You won’t have access to the entire Twitter “fire-hose” (Twitter’s API provides 5%):: You’ll only be able to search by one #hashtag or @handle

Assuming you use a platform that allows you access to 100% of the fire-hose, you will still miss 97% of all relevant Tweets. Why? A recent study by Datasift shows that only 3% of people remember to include a #hashtag, let along the right #hashtag.

Think of all the conversations you are missing! Verifeed White Paper | August 19, 2013 | w.ww.verifeed.com

You need to be able to search100% of the fire-hose with multiple search terms, phrases, keywords, #hashtags, @handles, all with with geo-location. Then you need the system to weed out all the re-tweets and repetition, and filtered out any spam, viruses, porn or worse from your search. Finally you need to be able to easily publish contextual conversations, know only the most reliable and relevant posts will be included, and recognize and reward those users with the most compelling tweets via special offers and coupons, with calls to action to invite their friends.

Engaging people in relevant and informative conversations about issues and developments contextually relevant to their brand is a value-add for your customers who want to more easily find ‘facts and friends’ relevant to them. It also inspires the sort of brand loyalty that drives word of mouth referrals and revenue. Meantime they are also giving you valuable insights for the bargain.Verifeed White Paper | August 19, 2013 | w.ww.verifeed.com

Add to that a layer that lets you analyze the reputation and expertise of the people posting each Tweet. How much influence they have with other influencers and people in general? How much a Tweet has influenced a buying decision? Or caused an adverse reaction to a brand or company? Add to that demographic and psychographic information. Tie it all up with an integrated customer-centric solution.

Now add the ability to assure the probability of what is being said is either accurate or inaccurate.

That’s rocket science, but the rocket is being primed for launch.


Why it’s all about trusted content and connection.

Actionable information is only as good as your willingness to act on it in real time to make a decision that could impact your bottom line.

Think about what it would take to put $100 of your own money down on the accuracy of a Tweet.

If you could replicate that ‘shoe-leather’ investigative and analytical process of a journalist with analytical algorithms and massive computer power to unearth valuable insights in real time, your brand or business benefit would benefit.


:: You would know faster and earlier than your competitors about the events, sentiments and issues that affect purchasing decisions.

:: You would know faster and earlier than your competitors who can most credibly and authentically proselytize your message.

:: You would be able to target and engage those ‘credible influencers’ faster than your competitors to persuade others to try and buy your product, or join your cause.

:: You would be able to improve your messaging to core constituencies by understanding, via the social data, how best to connect and communicate with them.


:: You would gain valuable ‘intelligence’ about your product (its strengths and limitations) to guide product development and enhancement, as well as sales and marketing strategies.

:: You would drive more traffic and ‘word of mouth’ network effects on your site and social media by publishing relevant and reliable real-time content streams that engage your users in meaningful and informative conversation.

:: You would be able to identify your biggest ‘opponents’ and engage with them in real time about your product, correcting misunderstandings and honestly asking for suggestions.

:: You could reward and recognize your (would-be) ambassadors with special call outs, rewards, discounts, trial offers, and more.


As social intelligence systems become more advanced, our ability to parse and verify important information from social networks will expand tremendously.

The ability of brands to “gear-up” their social data mining capabilities to assure real-time actionable information and connection will soon differentiate those who strike it rich, and those who miss the gold rush.


Author Melinda Wittstock is the founder and CEO of Verifeed. She is a serial media-tech entrepreneur, a media and content marketing executive, and an award-winning journalist (London Times, BBC, ABC, Financial Times/CNBC, Capitol News Connection). At Verifeed she brings together her social media, brand journalism and technology know-how to innovate the way we process, fact-check and utilize social data. Her vision is to replicate - with algorithms and automated processes – what an investigative and analytical journalist does intuitively to be first and right: Look for patterns, qualify sources and put data in context.

Twitter @Veriate, @NewsiTnewsLinkedIn.com/in/MelindaWittstock | Facebook.com/Verifeed

About VerifeedVerifeed filters and validates social streams in real time to provide enterprises with actionable and trusted social intelligence, contextual real time interactive content, and real-time targeted engagement.


Find out how to become a Verifeed partner at Info @ verifeed. com

Thank you!


Social Intelligence in the Age of Infobesity - A Verifeed White Paper

Business

social data

social networks

social intelligence

social networking

social habit

social marketing efforts

social login technology

age of big data