FreshNetworks Blog

Old Trafford, Manchester (Photo credit: Sean MacEntee)

At 09:20 this morning Sir Alex Ferguson retired after 26 years in charge of Manchester United. The club, and the manager, are respected and supported far from the city of Manchester, and reaction was quick to spread on Twitter. In many analyses of event and how Twitter reacts to them, the focus is on volume – just how many people are talking about an issue. But more interesting than this is what people are saying.

There is a hypothesis that when there is ‘breaking news’ (at least on Twitter), most of the discussions convey the same information – people either retweeting the original message or people conveying the same information to their followers that lots of others are doing at the same time. So in this case immediately after the announcement, whilst they may use different words, we would expect people to be conveying the simple message: Sir Alex Ferguson has retired.

But is this true – what did people actually discuss on Twitter in the first hour after his retirement was announced?

What we did

We captured every Tweet that clearly discussed Sir Alex Ferguson during the first hour after his retirement was announced shortly before 09:20 this morning. Using Datasift, we captured all Tweets that included the terms “Alex Ferguson” or “#fergie’ or ‘#mufc’.

In total we captured 95,312 Tweets in the first hour of discussion on Twitter – or about 26 Tweets every second.

What we found

First some basic stats about the discussions on Twitter in the first hour after the announcement:

• 68% of people discussing the retirement were male (16% were female and the remaining 16% had genders that could not be determined from Twitter)
• With 4.3% of all discussions, the news was actually discussed most in Manchester; London came second (3.8%). The global impact of the club is reflected with Indonesia, Singapore, Malaysia and South Africa being in the top 10 locations for discussions
• 22% of Tweets were people retweeting other people’s content; the remaining 78% were original Tweets
• The most retweeted account was the club themselves. This was followed by a number of accounts in Indonesia (UtdIndonesia and detiksport). The most mentioned UK news provider during the first hour was SkySportsNews.

With only 22% of Tweets as clear retweets, there was a lot of original Twitter content being produced. So what were people actually discussing:

• Just over one third of Tweets (34%) were simple statements that Sir Alex Ferguson had retired
• The next largest group (26% of Tweets) were reflecting on their own experiences or thoughts – memories of the club and what Sir Alex’s time there meant for them
• A further 14% of Tweets were thanking Sir Alex for what he had done for the club or indeed for their own experiences (a trend started by the club themselves in their announcement)

Some topics were less popular but noteworthy:

• 360 people (0.7% of all Tweets) were wishing Sir Alex luck in or sending their best wishes for his future
• 53 people (0.01%) were worried that Sir Alex might have died

So the first hour on Twitter was an interesting place, and the discussions were more varied than just retweeting or repeating the simple fact of Sir Alex Ferguson’s retirement. In fact a significant proportion of Tweets were reflecting on what his role as manager had meant to them and the memories they had of his time with the club. This kind of reflection and content is altogether more interesting than mere retweets and statements of fact and shows Twitter at its best – connecting personal experiences and opinions to larger events.

Listen, Understand, Act (Photo credit: highersights)

According to Altimeter, 42% of businesses in the US are prioritising Social Media Listening in 2013 – putting real focus on how they sift through and learn from the conversations in social media. But a recent study of US consumers found that 51% of them do not want brands to be listening to what they say online. As a greater emphasis is placed on social media listening and big data, the tensions with consumer privacy will also rise.

The report, by Netbase, is based on a survey of 1,062 US consumers and highlights the challenges brands will face as they increasingly listen to and act on conversations in social media.

• Most consumers (68%) realise that brands are listening to what they say online
• Just over half (51%) want to be able to talk to their friends and contact in social media without being listened to in this way; 43% would go further, saying that being listened to is an invasion of their privacy
• Finally, 64% of consumers only want companies to respond when they are directly spoken to

These numbers are confusing and difficult to interpret, and when you add in the data about what consumers do think brands should do in social media they become more so:

• 48% of consumers say that brands should listen in order to improve products
• 58% say that brands should respond to negative comments online

The numbers are all over the place so what can we learn from this?

The data is confusing which may be a result of poor research, or indeed may (also?) reflect the fact that understanding of social media listening is confused for consumers. That brands can listen in to conversations they are having with friends and contacts online can feel intrusive, but when the potential benefits of this are explained, more consumers are willing to accept this.

This is probably the best way to understand this data and to begin to understand how consumers will react to social media listening: they do not like it, but they like the benefits that they may get from brands listening to them. So for listening to be really effective, brands will have to make sure they have worked out the consumer value proposition before people make it more difficult for them to access their conversations online.

Nobody gets me Bitcoins! (Photo credit: zcopley)

In the current economic climate, many people have been looking into alternative investment opportunities, or safe places to put their money outside of the banking system. Options such as precious gems, or fine wine, are regularly discussed. One of the rising stars in this space is the Bitcoin market.

Bitcoins are a decentralised digital currency which uses open source technology to trade coins between users all over the world. Bitcoins are obtained by completing increasingly difficult mathematical calculations, this system ensuring only that a limited number of Bitcoins are available and that no one organisation controls the flow of coins.

Using social data to predict consumer behaviour, or even the value of goods, is nothing new, and many traders have been looking to include social metrics into their trading algorithms. Various academic studies have also highlighted predictive qualities of social data in the equity markets. However, because there are so many factors involved in pricing most financial instruments, it can be extremely difficult to accurately predict how markets will change.

Bitcoin however has several characteristics which make it an ideal market for social data prediction:

• The value of Bitcoins is determined almost solely on market demand, because the number of coins on the market is predictable and are not tied to any physical goods

• Bitcoin traders tend to be in the same demographic as social media users, and so their attitudes, opinions and sentiment towards Bitcoin are well documented

• Bitcoin is predominately traded by individuals rather than large institutions

• Events that affect Bitcoin value are disseminated first and foremost on social media

We intend to test the hypothesis that Bitcoins could prove an exciting testbed for social prediction, and give us a greater understanding of how publicly expressed sentiment and behaviour actually impacts the value of a commodity.

What do you think? Can social data be used to predict Bitcoin values?

Padlock & Chain (Photo credit: spodzone)

People are sharing increasing amounts of personal information online and via social media; some of this information is shared knowingly but other is information we leave behind us without necessarily realising.

This information from consumers is fueling many of the big data conversations that brands are having. But with this growing ‘data exhaust’ come challenges. Notably challenges of privacy and trust – how do brands work with this data in a way that doesn’t alienate their consumers, and what how might consumers react if they do.

We see two main areas of challenge for brands working with this data:

1. People are becoming more wary of intrusive online behaviour by brands

• There’s a powerful “personal data economy” out there, which has the potential to grow to nearly €1tn annually in 2020 according to the Financial Times, quoting BCG research.
• This is helping create a new breed of “invisible cyberazzi” – individuals and organisations tracking, amassing and selling people’s data – making consumers increasingly uneasy (e.g. TRUSTe’s 2011 study results).

2. Consumers are becoming more aware of the value of their personal information

• As highlighted by Facebook’s IPO last year, and there is now a new market for “data locker / vault” services (e.g. Personal, Qiy, etc., and explored in last week’s Financial Times).
• Subscribers can store their data in their locker / vault and only grant access to specifically chosen and approved organisations and companies.
• Access may be granted for free (e.g. to friends, family, or trusted organisations such as someone’s own bank), or for a fee (e.g. to marketers and advertisers).
• The UK government’s Midata Programme is also highlighting the value of people’s data. Through it consumer groups and businesses (including Visa, MasterCard, Three, Lloyds TSB, RBS, British Gas, EDF Energy) are enabling consumers access to the data created through their banking, internet purchases, utility use etc., in order to help them make more informed purchasing decisions.

So what could happen if people start locking their data away en masse? Companies may be forced to start paying people directly for their data, which for advertisers and marketers is potentially more likely to be accurate when acquired this way.

There’s a lot of buzz about big data in the retail banking sector right now as all the major banks work out how best to bring new unstructured data sets (such as social data and mobile data) together with transactional data in order to improve customer experience, become more competitive and drive growth.

I recently discovered this great debate from September 2012 that provides a clear understanding of where banks are today in their use of big data. The video includes panelists from HSBC, Barclays and RBS; the full debate lasts for more than an hour.

For those of you without the time to spare to watch it all, here is my summary of the main points:

1. What are the pain points that banks are grappling with?

• Customer retention, cross-selling, up-selling, developing new products that customers actually want, and minimising fraud

2. Where are the biggest opportunities with Big Data?

• Improving insight and understanding of the customer in order to deliver a better customer experience through highly personalised communications (‘the segment of one’)
• Using social media analytics to find out what your customers think of your competitors and their products
• Identifying and reducing fraud. Part of this is showing fraudsters that you are looking for them. Most banks are doing real-time detection already and this is where Big Data, combined with social data, can come into its own

3. What are the challenges with Big Data?

• Gaining permission to use and process some of the new data sets such as mobile and social media data. The panel all admit that financial services is behind the curve in this because of compliance issues, and that a lot could be learned from some of the new technologies and techniques that companies like Google, Facebook and Amazon have developed
• The ultimate goal of Big Data should be about delivering a better customer experience for customers. Not easy when the user journey is now dynamic when it used to be confined to in-branch interactions
• Finding the right balance between giving the right access to data across the company, and making sure adequate controls are in place. This is because the further away from source it gets, the harder it is to ensure compliance is maintained

4. Where should retail banks start in Big Data?

• Think about who owns the customer and therefore the data relating to the customer. This will require a rethink in organisational and governance structures, and a real need to get the C-Suite bought in
• Focus on your strategy in order to frame the right questions and therefore data that you need. There are infinite possibilities with Big Data. That said, the business and the data analysts need to work collaboratively. Once you start to visualise data, it can raise new questions or reveal that the original question wasn’t right in the first place
• The Holy Grail is to get the single view of the customer first, and then enrich this later with  newer data sets such as social data. Take things step-by-step – unlike Facebook, banks cannot afford to get their communications to customers wrong! They are already governed by a set of regulations to use data responsibly
• The emphasis should be on quality and not necessarily speed of communications. The next best action for the customer may not be a cross-sell – that won’t drive loyalty or build trust