How do you evaluate the cost-benefit of the government’s digital engagement?  As a result of the Improving government online review of measurement standards, several interesting discussions were started around extending the work on valuing and evaluating websites to all digital media.

Evaluation is a key priority for Government communicators. Matt Tee, Permanent Secretary and Head of Profession Government Communications, has prioritised evaluation as a key area of focus, along with skills and behaviour change. This is not surprising given the current economic climate. Government has to account for every pound spent and that means evaluating our communications activity to demonstrate cost-effectiveness.

Matt Tee has also requested that every government department develop a digital engagement strategy by March 2010, alongside the Public Accounts Committee recommendation that every department has a channel strategy, – a sign that digital engagement is being taken seriously.

So, how do we measure it?

As with any marketing communications activity, that depends on what the communications objectives are. However, there are commonalities across different campaigns and across different digital engagement tools and it’s those that I want to explore.

Recently, I’ve been working with colleagues in COI on this problem and we’ve come up with three common measures that appear to work across all digital engagement or social media tools:

1. Number of relationships
2. Number of user-generated content items
3. Number of referrals/recommendations

1. Number of relationships

The number of relationships or connections within a network is a measure of power or influence. For example, it could be the number of followers on Twitter, number of friends in Facebook or the number of subscribers to a blog. In social network analysis, this is the basic measure of centrality within a network, which is called degree centrality.

There are other interesting measures of power within a network. For example betweenness centrality measures the degree to which a member lies between other members of a network. In the Facebook analogy, a person may have 1000 friends but have less influence than a person with 50 friends, each of whom have 1000 friends.

Betweenness centrality from red (lowest) to blue (highest)

2. Number of user-generated content items

The number of user-generated content items measures participation within the network. For example, it could be the number of comments on a blog or the number of videos uploaded to a Youtube channel. It measures the level of engagement of an audience, suggestive of active participation not simply passive interest.

3. Number of referrals/recommendations

The number of recommendations is what many seek. This measures virality, advocacy, recommendability. For example, it could be the number of retweets, the number of  ‘share this’ actions or the number of pingbacks. It goes beyond mere participation; it means your content or message is valued enough to be recommended to others inside and outside the network.

We would be very interested to hear any thoughts on this. Many people are starting to think through return on investment in this area and it would be useful to have some level of consensus before applying to the government’s use of digital media for engagement.     Let us know what you think.

Read the full article (12 Comments)

It’s been a busy week.  I have often talked about the way users now behave on the Web, in response to organisations’ approaches to branding.  Mostly people use search, not only for finding things they don’t know about, but also locating the sites for things they do, but either cannot or don’t see the need to recall the URL.  Recognising this changes the way that we promote and advertise services. 

It’s not that Web search engines are the new government portals (see Persuasive Content’s nice comments on my keynote address to public sector IT Directors) but, because that is the way that the majority behave, they have become portals to all information.  What we are doing in government is working with people’s habits to make it easy for them to get to the public sector information and services they need and that we want to encourage them to use. 

There are implications for branding – and that has been the focus of most conversations this week.  Many still think of branding as the logo or, more generally, the visual image.  But that is a tiny aspect of branding.  Thinking of John Lewis or Waitrose or Morrisons, conjures up notions of quality, reliability and cost with associated personal and internalised reactions, such as ‘my kind of place’ or ‘good value for money’ – which their straplines reflect and differentiate from each other.

The key is what people know and think about the organisation and what you want them to know and think about you. That is done through language and positioning relative to other services – hence the importance of the Title and Description fields in the Web page metadata along with Search Engine Optimisation, and straplines that people use to describe the site.  The kinds of words we are working to have associated with many government digital services are authoritative, trustworthy, up to date, and comprehensive. 

Search engines are not the only routes; link text on other websites, blogs and tweets are also widely used word-based access points and create an aroma about your site.  What people say in the blogosphere is influential – hence the guidance to civil servants that moves beyond mere permission towards a positive encouragement to get involved and participate on the issues for which they are responsible (see Engaging using social media).  

Web presence is about building reputation and credibility for a particular audience through what is offered and their experience of interactions with the organisation.  A branded Web presence has to think through SEO, SEM, linking and social media strategies to do this.   It is what you and others say, and what you choose to use, that currently create brands on the Web rather than a logo.

Read the full article (2 Comments)