Social media consists of online conversations. The reason for measuring these conversations in some way (and the reason for all web analytics) is to collect data that can be used to make informed and therefore better business decisions.

Now if the data is not used or is not useful, there is no point in doing the measurement. This is not to say that the only use for measuring social media is to calculate an ROI in terms of sales (or equivalent conversions for non-transactional websites). But I firmly believe there is value in collecting whatever data you can collect, acknowledging it is incomplete and will always be incomplete but then using it as a guide in the decisions you are making.

As with any spend within an organisation, the first step should be to define objectives, what you would like to get in return for your spend. Simply tying social media to immediate sales, while easiest, is likely to underestimate the contribution it is making to your company. But not recognising that there is a need to justify the resource allocation (spend, time, etc) on social media in terms of how it is helping a company to meet its overall business objectives, will put at risk the existence of those resources.

The compromise may be a Balanced Scorecard that contains a mix of metrics with associated targets based on what you are trying to achieve with your complete marketing campaign, of which social media is one element. These metrics could include revenue, cost of providing customer support (targeting a decline), brand awareness across the internet, visitor participation on corporate blog, website engagement, ideas generated through customer feedback, etc.

This Balanced Scorecard would show the value and impact of social media to a company beyond the basic ‘how much money did it make us’. The metrics used would be tied to business objectives so they are relevant and meaningful to management. Each metric would have a goal so that performance can be clearly evaluated. If goals are not being met, then that is a trigger to say that action needs to be taken in order to bridge the gap. And any decisions made would of course be a more informed decision using the data that has been collected through the measurement of your social media efforts.

I guess the point of all this is to say that people need to be thinking about how to evaluate the performance of their social media strategy from the initial stages, this is not an afterthought that can be tacked on at the last second.

*******

This was an article I wrote for one of the Amplified 08 newsletters, a network of network event that is occurring on the 27th Nov in London.  For more discussions around calculating the ROI of social media, I recommend having a read of a post by Jason Falls (along with all 93 comments) or the post by Clare Munn (which contains links to other relevant blog posts).  A lot of my thinking around this was influenced by the comments on Jason’s post.  From that, I agree that social media must be more than a straight forward ROI calculation but I still feel that it needs to be measured and tracked over time in some way, hence my suggestion of a Balanced Scorecard approach.

This tool offers a means for tracking how your brand is being talked about online, across blogs and social networks.  The benefits of this should be immediately clear, offering a means to measure brand awareness and an easy aggregation of what people (potential customers and influencers) are saying about your brand.

It all got very interesting when I learnt that 6Consulting had made an offer to Vero (the organiser of SocialMediaCamp) to create a report using Radian6 to track the conversations around the event.  Having seen the tool, I thought my skills in creating dashboards would complement it well.  Following discussions, it was agreed that 6Consulting would provide me with the raw data from Radian6 and that I would transform it into an easy to understand dashboard.

The idea I had when creating this dashboard was to segment the data in the key areas (date, sentiment and social media type) and from this to show when conversations were occurring, what the median being used for this conversation was and the nature of the conversation.  The end result can be seen below with a PDF version available for download by clicking on SocialMediaCamp London Dashboard Download.

Note that this dashboard was built around an event with a short period during which there were conversations.  A similar dashboard would be produced for the online discussions around a brand but using consistent time periods, either weeks or months.  This would allow for the tracking of trends over time, with it being immediately obvious if there has been a shift in the perception of the brand.  Any change to a metric within the dashboard would then be investigated in more detail using the Radian6 tool.

Like all data, it is pointless measuring and reporting on it if the information from the data is not used to make decisions.  From this data, the key insights and recommendations that I would make for the future are:

• The majority of comments were being made live via twitter, 13% of all mentions were during one half hour period (involving a game of bingo).  It would be useful to have a moderator participating in these discussions, highlighting key points using twitter and also writing a blog post that summarises twitter conversation.  This could be published at the end of the day or even every couple of hours if the conversation is very heavy.
• While most twitter comments are made immediately, blog posts are generally written post the event and can be a number of days after the event.  Therefore a method of listing/highlighting all these posts to the attendees of the event should not be distributed until a few days after the event.  It would be interesting to see if there would then be a second peak in conversation, especially around comments appearing on blogs, at that point.
• Negative comments were in a minority and should be examined individually for potential improvements for future events.

This was related to social media through the examples I used, especially with my suggestions on which visitor interactions STA Travel and Bebo might be focusing on in evaluating the performance of a particular social media element within their websites.  While the methodology of how to actually calculate and use engagement is still being argued about in the web analytics world, I suggested a process that can be followed for anyone who wishes to start using engagement in evaluating their own websites.

My presentation at SocialMediaCamp London 2008

After finding BarCamp very interesting last weekend , I managed to score a ticket at the last minute to SocialMediaCamp London which was held yesterday.  The crowd was very different to BarCamp with a wide mix of professions and interests and it was great to know people from such a wide variety of areas are really interested in discussing social media on the internet.

After the debacle of my attempted presentation at BarCamp when 2 people turned up (one was a friend), I decided to learn from my mistakes and lure more people to this presentation.  Accordingly, the title for my presentation ‘Measuring Engagement of Social Media websites in the Web 2.0 world’ was stacked full of buzzwords.  It appeared to work as the room (advertised as holding 20 people) ended up containing around 30, leaving me a little nervous.  Thankfully everything went well and everyone seemed to appreciate the points that I was making.

To download a copy of the presentation, please click here.  Let me know if there are any questions regarding what I was speaking about.  I was very excited to discover that a couple of people were actually writing a post for the SocialMediaCamp blog during my presentation on what I was talking about.  To read more about my presentation (and some of the other talks given yesterday), please check out the SocialMediaCamp London website.

Other presentations

While all of the other presentations that I listened to yesterday were interesting, I got some very useful ideas from two of them.

Julius Solaris gave a presentation on how to use LinkedIn to get a better job and to promote yourself.  I had thought I was using LinkedIn properly already with around 60 contacts but I was wrong.  I will now be accepting any invites to connect as well as joining appropriate groups and trying to promote myself as an authority on web analytics.

Tom Whitwell (assistant editor of online at The Times) talked to us about headlines and how important it is to get them right in order to attract people to reading your content.  The trick is not to be clever or funny but to try and summarise the content (article or blog post) within the heading.  Again, I will be trying to make use of these tips and will stop trying to be clever with my presentations.