The volume of online sales on Christmas Day 2009 increased by 29% to £132m compared with last year according to a recent article from the Retail Bulletin, with Boxing Day drawing the highest sales volumes at £281m.

It seems as though many stores, both online and offline, started their promotions earlier in December, therefore you could still pick up a bargain before Christmas. This was particularly seen on Christmas Eve, which is traditionally a quieter day for online sales, but in 2009 sales rose by approx 68% to £181m compared with 2008.

David Smith, Director of Operations at IMRG (Interactive Media in Retail Group) said: “It appears that this Christmas we saw internet activity becoming part of most people’s routine, whether e-mailing friends, catching up on social network sites or surfing for bargains. Online retailers were able to benefit from this changing consumer behaviour by achieving record sales volumes with many sales and offers starting on Christmas Eve”.
It could be argued that the weather had an impact on internet activity, with many people preferring to shop online at home rather than venture out in the snow, or not have much choice if they were snowed in!

An online research specialist, eDigitalResearch, supported the findings and commented “Our research shows that 62 per cent of respondents were online on Christmas and Boxing Day searching for bargains. What’s really encouraging is the fact that Christmas 2010 looks like it will be even more prosperous for the online retailer.”

Therefore it seems as though online activity around Christmas is changing, and it will be interesting to see how 2010 performs in comparison.

They concede that social techno-graphics must also come into play and may be factor in determining the level of engagement adopted by a brand. For example, Toyota (the report is based on US data) is highly engaged over many channels especially around the Prius, while luxury auto brands such as Mercedes and Porsche have a low level of engagement.

This report is one approach to answering the question posed by many a Sales and Marketing Director:
“If I take up this social media strategy how will it effect my bottom line?”
So how can we measure the comparative effectiveness or value of social media engagement over other more traditional forms of consumer engagement? Some argue that is it a lost cause. The aim should be just to be part of the conversation, to be seen to be engaged. This raises the perceived value of a brand in the minds of consumers. It is about being connected and the benefits may occur way downstream of the initial engagement. This is a valid and is arguably the right approach, but at the end of the day the ROI question will always come up. We can look at applying different metrics and ways of assessing ROI, and we can say the the rules have changes, social media is not like traditional media. This argument is also valid, but the bean counters will never be satisfied and will always seek a monetary answer.

However there may be another way forward. I discussed this problem with some people at Glasgow University. Professor Muffy Calder is interested in mathematical modelling and automated reasoning for concurrent, communicating systems. She can see a potential relationship between the approaches and tools she uses in molecular modeling and the problem of social media ROI. There is a form of mathematical modeling called Markov chains. Markov chains have been used in chemistry to model the likelihood of a chemical reactions taking place, and to model the spread of epidemics. The Google page ranking algorithm is a form of Markov chain. They allow you to model the probability of a change from one state to another. Markov chains are a random process in which the future is independent of the past, and can be determined from information about the present state.
The prime problem of applying this to human behavioral modeling is the need for accurate data, but given accurate data on the types and percentages of individuals in different states of engagement, then one may be able to predict the percentage of a given population that may ultimately change from a stage of low awareness to high awareness, or from impassive to active engagement with the brand, or to an eventual state of making a purchase. If one combines this with a measure of cost, or rather comparative cost – i.e. the cost of one form of marketing activity against another – then you can equate the value of one form of engagement over another.

Such a model would be very useful as it would not only be applied after the event to measure effectiveness but could be a predictive modeling tool. The tool would be parametrized so that one could see the potential effect of changing elements of the engagement and the types of social engagement, the numbers of people interacted with etc.

The idea is that when you have to make an estimate of a potential value, such as a date, a numeric value a percentage etc, you follow this procedure:

1. Make your first estimate. This should be done in the way you would normally make an estimate; with as much thought and consideration you would normally apply.

2. Assume that your first estimate is off the mark.

3. Think about a few reasons why that could be. Which assumptions and considerations could have been wrong? What do these new considerations imply? Was the first estimate rather too high or too low?

4. Based on this new perspective, make a second, alternative estimate.

Then average the two estimates to get a new dialectic estimate – based on the internal dialog one has had about the potential value. Their research shows that the average of the two estimates produces more accurate results than can be achieved by either of the two independent estimates.

Although they describe this, any technique that prompts people to generate the dialectical estimate using knowledge that is at least partly different from the knowledge they used to generate the first estimate can suffice.

As they say: “Dialectical bootstrapping is a simple mental tool that fosters accuracy by leveraging people’s capacity to construct conflicting realities.”

So in effect there is a potential benefit of being in two minds about a subject. It is interesting that an internal dialogue will help us to make better decisions. We know from other work that an ability to view the world through different eyes and to take this into account when forming opinions or courses of action can improve the final outcome. Now we have a technique to apply this to a numeric process.

Only a few do all the tweeting…

A new book "Visualized, the Information Atlas" by David McCandless reveals that

• 5% of users make 75% of the noise
• 20% are ‘dead’ accounts
• 50% haven’t posted in the last week
• Only 5% have >100 followers



..And they’re not teenagers…

As previously mentioned on this blog, an intern at Morgan Stanley has recently produced a report on teenage media preferences and Twitter in particular. This reveals it’s not hot in high school because "it’s not safe". Teens prefer Facebook and MySpace because they have the ability to select who they want to connect to.

Matthew Robson (15) believes his report represents the collective wisdom of about 300 teenagers and comes to the following conclusions: they don’t listen to radio, don’t go to the cinema after 15, don’t use Twitter, never read newspapers, don’t pay for music, and don’t like web ads (although they do like viral marketing).

… But then, only 9% of Tweets have enough value to warrant forwarding anyway…

Pear Analytics dipped into the Tweet stream every 30 minutes for a fortnight and put each message into one of six categories; news, spam, self-promotion, pointless babble, conversational and those with pass-along value.

It found that 40.5% could be classified as pointless babble, 37.5% as conversational and 8.7% as having pass-along value. Self promotion and spam stood at 5.85% and 3.75% respectively. Conversational tweets were those that bounced back and forth between two users, and those dubbed "pointless babble" were of the "I’m eating a sandwich" type.

…and those are not news

Twitter also came dead last in the Pear study as a useful source of news, contrary to the company’s position of being the premier source of real-time news and events.

Tweet patterns

• Tweets are passed on most frequently at 11:30 am and on Mondays
• Spam is pretty consistent all day
• News is heaviest at 2pm
• Conversational tweets are highest between 2 and 4pm, and heaviest on Tuesdays
]]> http://amazeconnections.com/2009/08/25/twitter-stats-and-studies/feed/ 2 http://amazeconnections.com/2009/08/19/how-teenagers-consume-media-the-report-that-shook-the-city/ http://amazeconnections.com/2009/08/19/how-teenagers-consume-media-the-report-that-shook-the-city/#comments Wed, 19 Aug 2009 12:37:12 +0000 Paul Carysforth http://blog.amazeingideas.com/2009/08/19/how-teenagers-consume-media-the-report-that-shook-the-city/ This is the full copy of the research note written by Matthew Robson (aged 15 years and seven months), an intern at Morgan Stanley, which caused a stir after it was published by the bank.. http://www.guardian.co.uk/business/2009/jul/13/teenage-media-habits-morgan-stanley

Paul C

]]> This is the full copy of the research note written by Matthew Robson (aged 15 years and seven months), an intern at Morgan Stanley, which caused a stir after it was published by the bank.. http://www.guardian.co.uk/business/2009/jul/13/teenage-media-habits-morgan-stanley

Paul C

]]> http://amazeconnections.com/2009/08/19/how-teenagers-consume-media-the-report-that-shook-the-city/feed/ 0