Social Media Predictive Analytics

Social Media Predictive Analytics

Social Media Predictive Analytics

I believe that the future of social media holds a wealth of opportunities to increase engagement with your customers and your employees, drive efficiencies and encourage innovation. But where there is an opportunity, there is a risk.

 

Social media analytics will evolve and allow organizations to predict more accurately who will purchase their products and services and when they are most likely to purchase them. We will see the decline of email as the de facto communication tool, with more social-type systems taking their place.

 

Social media analytics

Social media analytics is a topic that already attracts much interest and hype. The ability to understand which products your customers buy, which promotions and advertisements resonate with the best, and which social networks they spend time on is really useful to organizations looking to drive sales.

 

A lot of work is going into improving sentiment analysis, which is the ability to automatically measure the positive, negative or neutral sentiment of social media posts.

 

Predictive analytics takes analytics to the next level and instead of offering insight into events or behaviors in the past, uses large amounts of historical data trends to predict what might happen in the future. There are already many examples of predictive analytics correctly predicting future events.

 

One of the best known is when US retailer, Target, ‘predicted’ that a girl was pregnant before her father had found out based on the products that she had purchased. The girl received coupons for baby products such as clothing and cribs which the father thought inappropriate until he realized that she was in fact pregnant.

 

In this example, the retailer used information about the products she had purchased to predict that she had recently become pregnant, and similar predictions can also be made by analyzing social media posts or internet activity.

 

Predictive analytics is particularly interesting and powerful because of the vast amounts of data we share on our social networks every day. Not only do we share posts about our personal lives, we often geo-tag them with the location and ‘tag’ our friends in posts when we’re ‘checking in’ to a nice restaurant for dinner.

 

Furthermore, it’s also possible for the social networks to track which posts a user has seen and how long they spent looking at them. All of this information is extremely valuable to companies who are looking to target their advertising more effectively.

 

Consider the key events of a couple’s love story. They meet, get engaged, get married, get pregnant, have a baby, celebrate the baby’s birthdays, their wedding anniversaries, and so on.

 

Now, while not all of these things will happen in the same order (or indeed, happen at all), there is a typical trend a company can analyze.

 

If our social networks know the date of our wedding anniversary, a company could make a prediction that adverts for flowers, chocolates or jewelry might be more effective during the weeks leading up to the date.

 

So, a company may pay more for placement of their advertisements during these times, then cut the advertisement on or after the date.

 

Many analytics platforms have been built for a specific purpose to give insight into social media data and offer predictions using pre-built algorithms. The more technically advanced and accurate way to perform predictive analytics is to build your own algorithms.

 

The advantage of this is that it allows you to pick what data you want to input and choose exactly what you want to predict instead of having to rely on off-the-shelf analytics packages that are available to everyone.

 

The disadvantage is that there is a fairly high barrier to entry into the world of predictive analytics. You’ll need to employ specialists such as statisticians or data scientists who will use advanced tools such as ‘R’ or ‘python’.

 

Building your own algorithms is likely to give you an advantage over your competitors because they will allow you to predict future trends more accurately.

 

There is already high demand for job candidates who have skills and expertise in analytics and this trend will continue for years to come. Incidentally, Harvard Business Review named Data Scientist the sexiest job of the 21st century!

 

There are a number of ‘use cases’ for social media predictive analytics, such as:

 

Recruitment.

By using data from multiple social networks, including LinkedIn, it will be easier to predict if an employee is more likely to leave for another job. This can be useful for headhunters when targeting candidates or for companies looking to retain their staff.

 

Industry trends.

Extrapolating trends in social media data that discuss certain industries, such as utilities, could be used to predict spikes in energy usage.

 

Crime prediction. Data that tracks where a crime takes place might help authorities to increase their presence if incidents are predicted in certain areas. Likewise, the monitoring of social networks, however controversial, could alert authorities to individuals or groups who are planning to commit crimes such as terrorism.

 

Law enforcement agencies around the world already use social media data to track fraud or predict where it is likely to occur.

 

Public opinion.

No longer will we need to rely on the polls reported during elections, instead relying on data from social networks to get a more accurate prediction on the outcome of elections.

 

Users are already becoming more aware of how companies mine their data in order to advertise products and services they might like. While this is good for customers, as it means the adverts that they see will be more tailored and relevant to them, users may view it as an invasion of privacy.

 

The EU General Data Protection Regulation is going to have a huge impact on all businesses in the EU as well as companies around the world who process data on EU citizens.

 

Because predictive analytics relies on vast amounts of personal data, we will see legal cases brought against companies who exploit predictive analytics to sell their products. Businesses will need to ensure that they are processing data ethically and in a manner, users would not object to.

 

Farewell, email!

Although not everyone will agree, I believe email is wildly inefficient. We use it every day and many of us have to sift through hundreds of emails to work out which ones are for our information, which ones need us to do something, and which ones are just junk offering us a promotion for something that we just don’t want or need.

 

The big problem with email is that we’ve grown so reliant on it that it’s difficult for us to imagine a world without it.

 

Have you ever considered how difficult and long-winded it can be to make a decision among your team through email alone? Let’s say that you need to agree on a decision with your team members, so you email all of them.

 

One member replies to you only with their comments, the others reply to all and one of the team also forwards the email to someone else for their input.

 

The person who received the forwarded email replies to you and all, but has missed some of the emails from the other members. You then compile all of the responses and reply to all with a synopsis.

 

This is complicated and could go on and on with multiple email trails breaking off and bringing other people into the conversation without your knowledge.

 

Many of the so-called digital natives, the people who grew up with social media, smartphones, and tablets, don’t use email. They communicate using social networks or instant messaging apps. Often, email is just one of those things that they need in order to sign up for certain services in the same way that we all need a postal address so that packages can be delivered.

 

However, many online services have started allowing users to sign up to their services by connecting their social network, rather than requiring an email registration.

 

Doing this is a quick and easy way to set up a new profile on an online service as it only requires a few clicks and all of your information is pulled to the new service automatically. These systems will continue to contribute to the decline in an email.

 

When I first mentioned my prediction to my wife that email will eventually die she said that I was mad. However, just a few days later, she asked her niece (a 19-year-old from Russia) to email her a document. The niece’s response was ‘email? Nobody uses email anymore! I’ll send to you it via Facebook’.

 

At this point, my wife realized that maybe email isn’t the great tool that we’ve become so over-reliant on over the years and that perhaps something else will eventually replace it.

 

Of course, I don’t think that email will completely die altogether. Good old snail mail has not completed died. After all, having a washing machine delivered to your house is far better than receiving a photo of one via social media, or indeed better than receiving instructions on how to 3D-print your own!

 

Some people do take pleasure out of using a real-life pen to write a real-life letter, put it in an envelope, affix a stamp and send it to a loved one.

 

But, you must agree that this sort of communication has declined significantly over the last few years. Many of my family members are abroad so at Christmas time I now expect electronic greeting cards that have a corny animation and cheap music instead of an actual physical card.

 

Although receiving a physical card would be a nice surprise, it would probably mean that I would be obliged to send a physical card in return, and I’m far too digital for all of that paper nonsense!

 

Other advantages of social collaboration over email include:

Centralization: everything is not stored across multiple mailboxes and is instead in one location, making it easier to refer back to at a future date.

 

Audit trail: changes to documents or discussion on an enterprise social network are all logged with the time that they were changed as well as the person who made the change.

 

Time-saving: rather than sending emails and collating the responses, social collaboration does all the work for you.

Space savings: If an email includes a large attachment, that attachment may be recreated every time a reply is sent, clogging up the inboxes of all recipients. If an enterprise social network were used, the attachment would be uploaded to the discussion thread and would be downloaded only by those who wanted to view it.

 

Hopefully, I’ve convinced you of the inefficiencies of email and that something more ‘social’ will surpass it. The problem, however, is that I don’t believe social networks, as they stand at the time of writing, are ready to replace email.

 

Traditional social networks are public, meaning anyone can use them. But, each person needs to be a member of the same social network or platform to be able to send messages to each other.

 

Email, on the other hand, allows a person to send a message to anyone else and it doesn’t matter which email provider or email software they are using. Internally, people within organizations are already communicating using their company’s enterprise social network, but this only allows communications within that network.

 

So, if they want to contact one of their suppliers, they will have to resort to email and all of its inefficiencies. There is no standard system that everyone uses to get the benefits of social collaboration.

 

I believe that the solution to this is a new protocol that will emerge and allow social-type communications to be sent from person to person without the need for everyone to use the same social network.

 

A new authority or network will emerge that will handle all social-type messages and become a hybrid between a traditional public social network and an enterprise social network.

 

When this happens, I believe it will herald a whole new form of communication. Email is definitely wildly inefficient and there are already tools that go some way towards addressing these issues but they do not have the required security or ability to share content outside of the corporate network.

 

Education

Social media offers people, young and old, opportunities to stay in touch with friends and family, to collaborate and to share their thoughts and ideas. Many social networks and online services offer users the ability to hide their identity.

 

The unfortunate side effect of this is that anonymity can lead to some unsavory behavior. Some users hide behind anonymity and act in a totally different way online than they would in the real world.

 

Social media and other digital technologies are usually adopted by the younger generations first. Only later do adults follow suit. At which point, the platforms become too ‘uncool’ because the parents have joined. This makes it difficult for parents to keep up with the technology that their children are using and makes it hard for them to determine where the risks lie.

 

Cyberbullying

Education is key and we will need to see far more of it to ensure our young people know how to use the internet responsibly and what to do if they ever encounter inappropriate material or behavior. There have been a number of examples where both children and vulnerable adults have been exploited online or faced cyberbullying.

 

The scary difference of cyberbullying to playground bullying is that cyberbullying can give the bullies anonymity, enabling them to be even harsher than in the playground.

 

Another issue is that while playground bullying may take place while the child is at school, cyberbullying can take place relentlessly at all hours of the day and night. There have been a number of sad cases where young people have taken their own lives as a result of cyberbullying.

 

Unfortunately, the internet has long been used by people to share indecent images, sometimes involving children. We’re now seeing a worrying trend whereby children themselves are being exploited directly.

 

Some of the new social networks allow users to send messages and photos that will allegedly be deleted a few seconds after they have been read or viewed.

 

The system isn’t 100 percent effective though, as it’s easy to take a screenshot when a temporary message or photo is received, thereby making a permanent record. Sometimes, children have sent indecent images of themselves to other users thinking that the other user is an attractive boy or girl of a similar age.

 

Unfortunately, in these examples, the person sending and receiving the image has been a criminal who then uses the indecent image to exploit the child by threatening to post the photo online unless they pay money.

 

In February 2015, Google announced that it would be launching a version of its popular video sharing site, YouTube, for children below the age of 13. At the time of writing, only children who are 13 or over can create accounts on YouTube.

 

This represents an interesting shift, in that it may signal a change where other technology companies begin to offer more services for children under the age of 13.

 

Of course, it’s easy for children younger than 13 to circumvent the age restriction by simply checking a box or lying about their age. But these age-specific versions of popular sites will rely on parental and network permission, which is clearly preferable.

 

If successful it should make the internet safer for children as the content will be monitored to ensure it is age-appropriate. It also presents opportunities to advertisers who will be able to market their products and services to a younger age group, in the same way, that they already do on television.

 

Oversharing

It’s not just children who face the risk of cyberbullying. There have been a number of cases of adults being ‘trolled’, which is a term used to describe when someone is bombarded with offensive messages on online platforms. We covered an example of this in blog 9, which resulted in the Twitter troll being jailed due to a campaign of abuse on Twitter.

 

Many people are guilty of so-called ‘oversharing’, when they publish so much information about their personal lives that they put themselves at risk of stalking or burglary. Where stalking used to involve physically the following someone, in today’s world those people can track their victims using social networks far more effectively.

 

On a more positive note, social media is an effective tool to support distance learning. University fees are high in many countries, which often means that young people have to take out huge loans to pay their way through studies. Because of this, distance learning has grown in popularity.

 

People who use distance learning engage with content, tutorials, video and other rich learning materials through the internet. They don’t need to sit physically in a classroom and instead can form virtual groups with fellow students and work together over the internet using social media and other learning and messaging platforms.

 

Education will not only help raise awareness of the risks of the internet among young people, but it will also give them the skills needed to work in a more connected and fast-paced world of technology.

 

Research from London First, a non-profit organization with a mission to make London the best city to do business in, shows that 80 percent of companies in London’s ‘tech city’ cite a lack of skills as the biggest single barrier to growth.

 

The UK school’s curriculum was changed in 2014 to ensure that children are equipped with the new skills that they will need when they enter work, such as programming and graphical design skills.

 

CASE STUDY

Teenagers commit suicide because of cyberbullying

In 2013 tragic reports emerged in the media about teenagers aged between 12 and 17 years who had taken their own lives due to cyberbullying.

 

Many of the teenagers were users of Latvian-based social network ASKfm, which allows users to ask questions and post responses anonymously. While this may seem innocent at first, it turned out that many teenagers were being bullied on the site.

 

One such teenager who took her own life was 14-year-old Hannah Smith from Leicestershire in the UK. Hannah’s father, Dave Smith, said that he found posts on ASKfm from people asking her to die. The messages on the site urged Hannah to ‘cut herself, drink bleach, and kill herself’. Mr. Smith has called for tighter controls to be placed on social networks.

 

In another case in September 2013, a 12-year-old girl from Lakeland, Florida, committed suicide after receiving text messages saying ‘You’re ugly’, ‘Why are you still alive?’ and ‘Go kill yourself’. On the day that she killed herself she changed her name on a social network to ‘That Dead Girl’.

 

In total, ASKfm has been linked to six suicides.

  1. UK Prime Minister David Cameron commented that social networks must ‘clean up their act’ or face boycotts from users.
  2. Peter Wanless, chief executive of the UK’s National Society for Prevention of Cruelty to Children (NSPCC), said:
  3. The cruel nature of cyberbullying allows perpetrators to remain anonymous and hide behind their screens.

 

This is something that must be tackled before it gets out of hand. We must ensure young people have the confidence to speak out against this abuse, so they don’t feel isolated and without anywhere to turn.

 

These examples of cyberbullying are clearly worrying and shocking. Children need to be taught about the internet and how to stay safe while on it. Companies also need to be aware of this risk.

 

Advertising is often purchased through agencies in ‘bundles’, which place their adverts on a wide range of websites. Because of this, some companies were unaware that their adverts were appearing on ASKfm beside such worrying content.

 

Many advertisers later pulled out from advertising on ASKfm as a result of the revelations in the media about child suicides.

 

Democracy 2.0

The revelations of Edward Snowden in 2013 caused a public backlash against the way that technology companies were passing information about their users to governments.

 

The technology giants claimed to have been under pressure from governments to release information about their users so they publicly fought back by introducing the publication of so-called ‘transparency reports’.

 

The transparency reports detail the number of requests from governments for user data to be handed over or for data on the networks to be removed. The reports also show how many requests were fulfilled by the tech giants and how many were rejected.

 

In February 2015, Twitter reported that it had seen a 40 percent increase in the number of requests from governments since its last report in July 2014.

 

Twitter received a total of 2,871 requests from governments across the world asking it to reveal data about 7,144 of its users during the second half of 2014. Twitter reported that it had fulfilled 52 percent of the requests.

 

Governments in non-democratic states have to work harder to control their citizens’ online activity. Social media is an inherently open place where people can post whatever they wish.

 

A government censoring the internet will keep a constant watch on social media to understand public sentiment and to work out which pieces of content pose a threat and which it might, therefore, remove or censor.

 

Bitcoin and the dark web

It’s difficult to tell what the future will hold for social media in the context of internet censorship. We’ve already seen social networks and other online services get banned in certain countries; for example, in 2013, Twitter was blocked in Turkey. Savvy users in countries where censorship is more pertinent know ways to get around these blocks. Two of the most common are:

 

Virtual Private Networks (VPNs).

These are often paid services which route internet access through another country and encrypt the connection to the user’s computer. This makes it far more difficult for a government to monitor a user’s internet usage and it is also an effective way of bypassing local internet censorship.

 

The Onion Router (TOR).

TOR is a tool that heavily encrypts a user’s connection by routing it through hundreds of other computers (like the layers of an onion). Using TOR gives users access to the so-called ‘dark web’, websites that are accessible only through the TOR network, like a parallel internet.

 

Using TOR allows users to avoid internet monitoring and bypass internet censorship. There are legitimate reasons to use TOR; for example, journalists might use it to report news from somewhere the internet is heavily restricted or censored. However, some of the dark websites that operate on the TOR network have become infamous black markets for illegal goods and services.

 

Transactions on the dark web marketplaces usually make use of crypto-currencies, such as Bitcoin. Crypto-currencies are decentralized digital currencies that use cryptography to secure transactions and control the creation of new units. The security that crypto-currencies offer means that criminals can make transactions on black markets anonymously.

 

Since TOR is all about anonymity, the ‘social networks’ that exist on the dark web tend to be messaging boards where users are identified by their username or alias, rather than their full name. Many hackers use message boards to discuss vulnerabilities in security systems and to organize cyber attacks.

 

If you want to find out more about the dark web, once connected to the TOR network you can visit the ‘Hidden Wiki’, which is a guide to the dark web.

 

It’s easy to understand why some governments want to control internet usage because we’ve already seen the effect that it can have. In July 2014, Russia announced a reward of 3.9 million roubles ($110,000; £65,000) for anyone who can crack the identity of users of the TOR network.

 

The Arab Spring was a revolutionary wave of demonstrations and protests that began on 18 December 2010 and swept through countries of the Arab League and its surroundings.

 

During it, people used social media to organize demonstrations and circumvent state-operated media channels. In August 2011 the London Riots were said to have been coordinated through social networks.

 

Incidents like these have led some governments to censor the internet and monitor their use. But, this move towards internet censorship and monitoring has prompted public scandal and opposition. It has also had a knock-on effect on how companies use data to track users and serve advertisements.

 

New social networks have started to appear, such as Ello, which is a social network that claims to never sell user data to advertisers and to never show adverts. There’s also Telegram, an alternative to Whatsapp developed by Pavel

 

Durov, who is sometimes described as ‘Russia’s Mark Zuckerberg’ after having launched Vkontakte (VK mobile version, Russia’s equivalent to Facebook). Durov stepped down from his company in 2013 amid wide speculation and rumors that he was forced out by the Russian government.

 

His new venture, Telegram, claims to be highly secure and rely on complex encryption algorithms which make it far more difficult for governments to track usage.

 

Encryption

I believe that we’re only at the start of the movement towards greater encryption and protection for user data in social networks. We will see more high-profile incidents and examples of excessive monitoring and censorship of the internet by governments and companies.

 

This will increase the public’s interest in internet and social media monitoring and will result in more social networks and platforms being created in order to safeguard free expression and protect user data.

 

Encryption will be key to this protection. Once data has been highly encrypted, without the password (or key) it takes an enormous amount of computing power to crack. It’s possible to encrypt data so strongly that it would take even the most powerful supercomputers in the world years to crack the encryption.

 

Once social networks have developed their security to a high level it will make it extremely difficult for governments to monitor or censor the internet.

 

This will mean that governments will find it more difficult to control or influence public opinion, which will, in turn, lead to more transparency, more freedom and thus a more pure form of democracy – what I like to call democracy 2.0.

 

This isn’t without risk though, because without effective monitoring it will be easier for terrorists to use encrypted communications to plan attacks. Because of this, encryption and anonymity remain hotly debated.

 

Identity verification and biometrics

As we discussed in the previous section, one of the issues with social media is that it’s easy to hide behind anonymity. Many social networks try to stop users from being anonymous by imposing terms and conditions as well as by asking users to verify their identity by connecting their email address or phone to their account.

 

However, it’s easy to circumvent these types of controls. Many governments have ambitions to introduce digital methods for identity verification which, if successful, could even allow voting to take place electronically rather than in a polling station.

 

Estonia is one such government that has already introduced electronic voting which builds on the Estonian ID card.

 

A secure database will authenticate someone, similar to the way that we authenticate ourselves to gain access to our internet banking.

 

The big difference in the future, however, will be that when it’s possible to authenticate someone with 100 percent certainty, it won’t just allow people to withdraw money from their banks, it could also be used to pay for things in shops or to vote in an election and it would mean the end of the dreaded password.

 

There are dangers, however, because if the authentication was hacked, it would make it easy to steal someone’s entire identity.

 

If you are a victim of bank fraud, it’s a serious matter, but if this form of identity verification could be circumvented it would prove a considerably greater risk. It’s dangerous because it represents a single point of failure, a bit like putting all your eggs in one basket.

 

One new technological development that may go some way to solving the identity verification problem is the use of biochips and biometric data. In theory, biochips or other forms of biometric authentication, such as retina or fingerprint scanning, will make our lives easier.

 

We will no longer need to struggle to remember all of the passwords for our various accounts, instead of relying on something unique to us, such as our fingerprint or a microchip embedded under our skin.

 

There have already been trials in Sweden where employees of a company experimenting with biometric authentication offered their employees the chance to get a biochip implanted under their skin.

 

This meant that when performing simple authentication tasks, such as getting into the office or logging into a computer, they didn’t need a password or a physical ID card, instead, they just needed to swipe their wrist (the place where the biochip had been implanted).

 

Hannes Sjoblad, Chief Disruptive Officer of the Swedish bio-hacking group BioNyfiken, which implanted the chips into the workers told The Times: ‘We already interact with technology all the time. Today it’s a bit messy – we need pin codes and passwords – wouldn’t it be easy to just touch with your hand?

 

We want to be able to understand this technology before big corporates and big government come to us and say everyone should get chipped – the tax authority chip, the Google or Facebook chip.’

 

There will always be people who try to crack security systems and unfortunately, there will always be criminals who won’t think twice about killing someone to get at their biochip in order to steal all of their money.

 

If a criminal is capable of killing someone in order to steal their bank cards today, are we to expect that if bank machines start to allow fingerprint scanning that criminals would start cutting off peoples’ fingers?

 

There’s no doubt that technology will continue to evolve at a rapid pace but whether or not it will change our lives will depend on the public’s readiness to adopt new technologies, such as whether we will allow biochips to be implanted in our bodies. It will also depend on the technology companies’ abilities to build security into their systems, products, and services from the ground up.

Recommend