Tag Archives: Analytics

Pointless babble on Twitter: The Social Data goldmine

We tend to think of Twitter as trivia and pointless random outbursts.  The people we follow flood our Twitter stream with useless phrases filled with their seemingly pointless activities.  But these random Tweets can have great value to data analysts.  For example, in 2008, bacon was more popular than sausages, Well people were tweeting about bacon much more.  We also tweet about feeling better in the morning than we do for the rest of the day.   Don’t worry if you’re an evening person, our moods seem to rally before we go to bed too.  We’re happier when it’s sunny, and when it’s the weekend.  Obviously..

These flippant top line results are the results of massive amounts of data gathering.  Cornelll University looked at over 500 million Tweets during research carried out over 2 years. (You can read the full report and supporting information here)

But what if you’re just curious and would like to take a pulse of what’s going on and you’re not interested in sleep or moods?

Have a look at Time Use which lets you play about with topics and find out how popular they are at different times of the day.  You might have wondered what people do all day, and now here’s your chance to find out what they tweet about.  Now I know that many more kebabs are eaten in the early hours of Saturday and Sunday morning than during the rest of the week.  This, just happens to correlate exactly with my experience of eating kebabs when I was at college


And now I also know that chips are far more popular than either curry or kebabs – especially on Fridays and Saturdays.


So, next time you’re wondering why your colleague is Tweeting about the trivia of their daily lives, they might be adding to some vital bit of data analysis – which could affect kebab production rates across the Twitterverse at weekends.

So keep tweeting that trivia.  Someone thinks its of value… Open-mouthed smile

PS – Huge shout out to Allister, The Digital Marketer who reminded me of Time Use…

Eileen is a social business strategist and author of Working The Crowd: Social Media Marketing for Business. Contact her to find out how she can help your business extend its reach.


Why should you follow me?

…Or why should I follow you for that matter? Amber posted an interesting read about why Reciprocity is a fallacy in Social Media. And so it should be.

Social media thrives on reciprocity.  Wikipedia notes that:

In public good experiments, behavioral economists have demonstrated that the potential for reciprocal actions by players increases the rate of contribution to the public good, providing evidence for the importance of reciprocity in social situations

If you follow someone who follows you, then you’ll get a great dialogue going, you can track what’s happening in your closed circle.  But there’s a problem…

Lots of us use social media to discover what’s happening in our circle of friends and their circle of friends.  Often friends and colleagues follow each other to find out what’s going on.  But this can lead to insularity – especially within organisations who are making their first forays into social media.  See the image below:


In the diagram on the left, A,B and C follow each other closely.  They know each others friends and they have a huge overlap in their friend circles.  Due to the high level of interaction between A, B and C, there is little time for any of them to reach out and discover new people, news or trends that might be happening outside of their world.  Only C has a link to D, but with the interaction between his other colleagues, messages from D, who doesn’t know any of the others, might be missed in the stream.

Contrast this with the example on the right.  A, B, and C have minimal interaction with each other.   A might follow B and C but not necessarily be followed back.  Their interaction is focused externally so messages from D have a better chance of being noticed and propagated through the group.

I’ve noticed this a lot in large organisations and clients I work with.  Often, they spend significant amounts of time following colleagues in the same team as them, and certainly in the same larger group.  They don’t seem to watch what is going on around them and often miss items of significance.  This has been an issue in some organisations who focus their attentions internally much more than looking externally.  Often, breaking some of the connections and replacing them with external connections has brought a new perspective and new interaction onto the group and its outreach.

So it really doesn’t matter who you follow.  you don’t have to follow anyone back because they followed you. Follow the people who you find interesting, not necessarily other members of your team or everyone in your organisation.  Follow those who interest you – and when they stop giving useful information – all you have to do is unfollow them.

Reach out to new people and stop the network imploding under the weight of the same pieces of information going round and round. 

Follow someone new today Smile

Digg This

2.2% Hard-core Tweeters generate 58.3% of traffic

I’m fascinated with growth graphs – especially graphs that show when something has reached it’s tipping point and is about to get ‘BIG’ Sysomos have done some great analysis of this and have published some stats. Here’s what’s happened to Twitter last year:

Twitter Statistics 2010 Growth in Accounts

The tipping point for Twitter obviously happened at the end of 2008.  Before then, only those in the know were using Twitter, it hadn’t gone mainstream yet.  I know the end of 2008 was when the BBC started to talk about Twitter – always a sign that a product has hit the mainstream when the BBC start to talk about it.  (The same think happened with podcasts a year or so before them).

But look at the explosion in new accounts created in 2009.  Lots of these are multiple bot accounts.  Applications like Hootsuite, CoTweet and TweetDeck allow you to switch accounts and publish the same Tweet from multiple accounts.  This shows up as buzz in sentiment monitoring tools and gives the client a warm feeling that the money he’s investing in Twitter implementation is worthwhile (<wry smile/>)

But others have started to create Twitter accounts.  As small businesses realised that Twitter (and especially Twitter search) can be Very Good for Getting You Noticed.  Hundreds of small businesses now use Twitter to connect (search for cupcakes, gourmet food and pork pies to see what I mean)  This has caused the explosive growth as millions more Twitter accounts appear. 

The growth in meaningful profiles, web URL’s and descriptions have also grown amongst genuine users (not the bots)

Twitter Statistics 2010

and has also dramatically increased in 2010.  if you’ve had a profile for a while, you’re one of 4.7% as people who created a profile before January 2009 only accounted for 4.7% of the total Twitter population.  You’re obviously thought leaders if you’ve used Twitter for some time (and nabbed the shorter account names too!)

And just over a fifth of users post 90% of Tweets



But don’t worry if you’re not prolific.  There are thousands of accounts out there that have been grabbed (usually for big brands who didn’t get there first) These Tweet squatters are waiting to sell on the accounts and are inactive.

The average number of Tweets per account is still one – so don’t worry if you Tweet once a week and wonder if you need to do more… There are millions of other accounts doing exactly the same. Smile

Full report and more images on the Sysomos blog


Reflections on 2010

I was going to make this post a whiz through the social media tools and give a few predictions for 2011.  I’m also going to reflect on the social media features that stuck out for me through my own journey through the year, the good and the bad and why these topics will need to be a focus for 2011 if you want to succeed in your social media marketing activities.  Bruce wrote a great post about what went right and wrong for him in 2010, but I think I’ll save my personal reflections for another post in 2011 and build on my post in 2009.  So this one is about 5 social media themes that I think will matter in 2011…

  • Social Web sites

More and more companies have included a social component to their websites.  Whether that’s an embedded Twitter feed or social buttons on a web page or blog, web sites have become more engaging.  And with search including social status messages and fees, real time search helps your SEO and rankings.  Social searching will find your site without any extra SEO  additions to the web page.

  • Crisis management

Large and small companies alike don’t know how to handle PR Crises often resorting to ‘delay. deflect, defend’ tactics that they’ve done in the past.  Yahoo!, who should know better had a wobble when news leaked out that it’s delicious bookmarking services was going to be sunsetted.

image (Image from the Techcrunch blog)

Without a viable alternative (Microsoft played around with a tagging service a couple of years ago), what would happen?  Delicous inicated on its blog that there may be a home for it outside the company but the search engine land blog had confirmation that this was unlikely and techcrunch reported that Yahoo had devalued delicious significantly if they were actually going to sell it!  not a good move

Having an effective crisis management plan in place would have significantly diluted the impact of this news – and having a good set of employee guidelines would minimise the impact of the leaks… Fail…

  • Lead Generation

There are a couple of interesting companies that have used Facebook to generate leads outside of the B2C space.  Forrester Research, for example has a fan page on Facebook which seems to be at odds with it’s B2B corporate image.  However, events and videos are linked to from Facebook, there are links to YouTube videos and there are reports you can download for free without a subscription. This is a really effective way of demonstrating that Facebook can actually be used for B2B successfully and acknowledges the blending of our work life and social life.  Nice one Forrester.  A great way of turning page visits into leads with a content rich, compelling and engaging Facebook Page.   Other B2B companies that use Facebook pages for leads are Cisco (who also initiate their graduate recruitment via Facebook as do Ernst & Young), Gartner, Hubspot and Dell

  • Location based applications

Facebook Places, Gowalla, Foursquare have all been talked about this year due oto privacy concerns, but they are a great way for B2C advertisers to tempt folks into the building with the promise of a coupon, discount or special offer.  Adding geolocation services to your application means that you’ll build a much richer application, and bring the community together at events and gatherings – only if the application is well written of course. I think that social commerce applications will come to the fore in 2011 – perhaps a deal with Groupon and Facebook places perhaps.

Do I use them?  Hardly at all.  I’m busy interacting with folks on another medium like Twitter or Facebook to collect another foursquare badge…

  • The rise of mobile

This is best explained by this infographic from the Flowtown blog

If you’re focusing on B2C, you need to be on Facebook, you need to have a mobile app and you need to take advantage of geo-location services.  There are over 4.5 billion mobile devices out there, many with a data plan.  Smartphones are the way to interact with your customers and allow them to connect with each other.  If you’re running a social commerce site, you need to be here…

Roll on 2011.  There are some exciting times ahead Smile  Happy New Year!

Digg This

How do you measure a hunch?

imageThere  are loads of social media monitoring tools out there on the market that claim to give you deep insight to your customers and how they feel about you and your market offering.  They drill deep into social media platforms and search for keywords to measure sentiment and emotion.  They range from £30 per month to £3000 per month (when associate with a brand agency).  But are thy accurate in gauging the real feeling of your audience.

I’m convinced that the only real way you can find out what they’re thinking is to actually talk to them and I’m not sure that these tools can ever replace the feeling you get when you’re face to face in front of your customers.


Darrell at Advertising age has a couple of great examples where these tools can get things wrong…

Here’s one it thought was "negative" about the ubiquitous coffee chain: "im tired of this cheap coffee crap. need starbucks!"

And one it thought was positive: "Pumpkin chai. Mmm. Very good. Not from Starbucks. From Argo Tea."

That’s where you need to have the personal relationship with your customers.  Surveys and sentiment don’t really connect t with the people that you want to.  i received a survey the other day from a large storage company asking me about my perceptions about the company.  i know them well.  They provide SAN solutions.  I have no idea what else they do, apart from SAN solutions. So asking me to comment on their other brand offerings was a waste of time.  And my answers, like many others I suspect, might skew their results about what folks think of them.  They do storage not cloud in my opinion!

so I think that the main point of this is that no matter how sophisticated your monitoring solution is, or how much it costs, or how good the dashboard is, it can never replace the human hunch

You need a team of people to monitor your social media traffic and conversations, and their hunches and insight will tell you how folks are feeling about you.

Sure, you’ll get the hard metrics your CFO needs, but you need to manually analyse the comments themselves to make sure you’re on the right lines with sentiment. 

It’s the human touch that buys your products and talks about them.  Make sure that the human touch analyses what your customers are saying too – and provides the extra insight that a computer can never do…

Digg This

Facebook Liking or Digging Digg?

I was listening to an argument the other day amongst my network and it concerned the importance of Digg versus Facebook likes.

Digg, the social news website has been around for a while.  Users can vote, or ‘digg’ articles that bubble up to the top of the top news trends 


Hmm.  It’s interesting to note that this screenshot, that I took today, isn’t dominated by news of the Wikileaks revelations that have been dominating the news for the last 7 days.  The more folks that vote to Digg an article, then the higher it appears in the results.  But this approach seems to be open to abuse, with mob mentality activities altering the results according to wikipedia…

So Digg is linear, and gives linear results based on the numbers of votes that it received. So… is that better than Facebook likes?

I don’t think so.. especially if you’re a marketer who wants to find out your reach in the B2C market.

Facebook’s Like feature allows you to broadcast recommendations to your friends and to your friends friends.  And as 75%of people who post reviews on products or services push this review out to their friends who are then 4  times more likely to buy.  Your friends respect your ‘social worthiness’ and will reinforce your decision in a really effective way by purchasing the product you’ve recommended. Digg is linear, Facebook Likes are exponential (viral).  Offers delivered via Scoutmob or Groupon can more easily be delivered over Facebook too, so there’s a good opportunity for B2C social commerce

Simply put. Digg articles are recommended by strangers, whereas Facebook recommendations are made by your friends.  you trust your friends, so you trust their recommendations.

So is Digg on it’s way out as a social bookmarking site?  Does the new algorithm allow for flash mob type voting on articles?  Will Facebook continue its rise for B2C market dominance or will another new mechanism appear to take Digg’s place?

Technorati Tags: ,,
Digg This

Predicting the Stock Market–on Twitter

Now I know why I spend so much time on Twitter.  Researchers at MIT published this graph in their Technology Review which has an accuracy of 87.6% in predicting whether the Dow Jones Industrial Average will go up or down.

GPOMS = Google-Profile of Mood algorithm which records the level of happiness, kindness, alertness, sureness, vitality and calmness.  The researchers looked for correlations between the GPOMS and the Dow Jones average and found that the calmness indicator index is correlated to the rise and fall and they can predict this change up to 6 days in advance.

Amazing stats for an Economist, but it might just blow your mind if you’re a mere mortal who uses Twitter for fun and information.

Full story in Technology Review

Digg This