Category Archives: web analytics

Big data, big opportunity

I visited Adtec today and attended various presentations and seminars. One of these was delivered by Fabric/Infosys and talked about Big data and to kick it off presented some interesting data facts I want to share. You might be surprised to know that:

  • Amex is 85% certain of who will divorce based on transaction data
  • Twitter is 97.3% accurate in reporting box office takings and more accurate than the film organisation responsible for making these forecasts
  • Google’s data predicting flu (influenza) trends is more accurate than The Centre for Disease Control

There was another statistic about Facebook being more accurate than Gallup weekly polls in predicting a political outcome (I think from mid-terms) but the pace was so fast I missed it. After a quick Google search I only found less accurate claims such as this: click link.

Another amazing stat., although not from this session is that 90% of all the data in the world was created in the last two years.

So “Big Data” is hot and a term that will grow in use.

Measurement: everyone’s talking about it

In the past few weeks there have been numerous items in the on and off-line press about measurement. It seems that when the going gets tough the tough really do start measuring and for so called “New Media” it has never been tougher in its life time. Expect therefore to see a lot more on the subject in coming months as the recession worsens and ahead of that I thought it might be useful to look at one area that commentators are finding most interesting.

Social Media measurement has received some significant attention and in particular recently because of the elevation of Twitter as the new application of the hour. Before Twitter became mainstream organisations were already looking at how social media could be measured and just this week Matthew Yeomans of Custom Communications was interviewed by Econsultancy on the topic.

On measurement, Yeomans identifies three areas that combine to provide ROI data about social media effectiveness in a campaign context. These are: Reach, captured from how many people are talking about a brand post campaign; Sentiment, from people during the campaign; and Competitor Analysis to identify how the brand compared to peers. It would seem that sentiment presents the most measurement issues when talking with clients.

That measuring sentiment creates the most debate is not a huge surprise.  Methods range from software solutions that track words such as ‘like’ and ‘hate’, alongside mentions of brands, to detailed manual evaluation of comments made by identified influencers. The differences between qualitative versus quantitative methods and hybrid solutions leave marketeers no clearer about what is the right approach. Even within the social media industry there is widespread concern that when talking about social media measurment, too many people are talking different languages and that no concrete answers exist – yet.

Adding to the conversation and to the education is some research carried out late last year by Marketing Sherpa. The US study looked at social media measurement, what worked and where problems existed. It concluded that too many Marketers were hung up on quantitative measures when in fact qualitative measures added the most value. The survey found that the easiest things to measure (advertising for example) were the least effective and the hardest things to measure (user reviews & ratings and relationships with bloggers) were the most effective.

Working in a research organisation such as Foviance, it is easy to understand why qualitative measurement is so important. 60% of the work we do is qualitative but a great deal of that  qualitative work is supported by quantitative findings from parallel research in complimentary areas (i.e. using web analytic traffic data to enrich findings from lab observation studies). In our view both are needed to provide a complete picture but there are cost benefit arguments with every research project.

One thing is certain; that there is a lot more to learn and marketeers are going to have to work through the noise to develop a clear understanding of how measurement should work for them. Establishing a measurement framework and strategy is work that can and should be done before even getting involved with evaluating solutions and should start far higher than at just the social media level. Providing your organisation with a measurement strategy is the gift that keeps on giving as it makes sure any decision making is firmly grounded in the business.

Returning toTwitter offers a good example of where a strategy should be established as measuring the effectiveness of Twitter is extremely challenging. Trawling through pages of “Tweets” to establish which referred to your campaign or website is not only time consuming but also very difficult to get right. Twitter’s use of Tiny URL means that you cannot quickly see your campaign url and have to actually clicking through to see which one works. Knowing what to look for first can only help.

Today, organisations are trying Twitter out and the investment can be written off as innovation. Fairly soon someone senior is going to ask about ROI and when they do expect a range of measurement solutions to hit that market very soon afterwards. By which time of course, Twitter will have been superceded by the next big thing. Do try and keep up!

Coremetrics redundancy news

Word on the street is that Coremetrics has made a bunch or people redundant. Redundancies include sales, sales management and senior management. The reasons given relate to the Board’s desire to stop the company burning the cash (it raised $60m last year) and achieve run rate profitability.

As a web analytics software provider that relies heavily on the retail sector this is perhaps no big surprise. They have been trying to break in to other sectors but success has been limited. Enterprise software sales is a long game and the recession could not have come at a worse time for them I would think. Combine that with Omniture’s move in to the retail space and they could be facing more tough times ahead.

Talking recently to colleagues in the US they don’t seem to have the impact from the recession that the UK has but they are waiting and watching. This is certainly reflected in Coremetrics press release yesterday that the market is unpredictable. They had huge growth in Europe last year and a couple of non-retail wins including BBC Worldwide and Alliance & Leicester. Not sure it will be enough. I feel some M&A activity is ahead!