Social media and user-generated content may be the hot new thing, but from a measurement standpoint, they represent some challenges. In large part, the challenges stem from the fact that the space is immature and growing rapidly; faster, in fact, than experts can come to agreement on what to measure and how to measure it. For example, there is still active and healthy debate in social marketing circles about whether qualitative insights can tell you more than sheer numbers. Some, like noted blogger (and now Forrester Research social media analyst) Jeremiah Owyang, believe that a) quantitative metrics should be balanced in equal measure by qualitative insights, and b) measurement processes will vary depending on a campaign's goals. Yet, the reality at this stage in the evolution of social media is that we lack the equivalent of standard web metrics like the unique visitor, clickthrough and page view. Focusing on the number of friends your brand may have in a social network like Facebook or MySpace fails to capture the essence of "friending" and the interactions those friends may have with your brand.
Keep in mind that as a brand marketer, your opportunities to leverage social media exist within three basic spheres, each with decreasing proximity to your own brand (and a correspondingly decreasing degree of measurability):
- Your own website or sites that you either brand or sponsor
- Sites within your broader link network, such as partners or affiliates
- The vast range of sites on the open web that are relevant to your brand or industry niche (such as blogs, forums, communities and networks).
Any activity that takes place on your own site, for example, carries a high degree of measurability, and should be covered by whatever analytics package you have in place. The second and third tiers -- where actions linked to your brand or site and discussions referencing your brand -- are the least measurable by conventional standards, and that's where you are most likely to enter unchartered territory.
Charting the uncharted waters
Any major brand is going to be mentioned on hundreds if not thousands of sites and blogs across the web. These sites and posts need to be sorted by level of engagement. While it is the case that online word-of-mouth -- in the form of consumer-generated reviews, photos, videos and ratings -- can have a significant impact on purchasing behavior, not all of this content carries equal weight. Examining the relevancy, frequency and visibility of brand mentions is one way to reduce the list of thousands or hundreds to a far more manageable list of tens.
Colleagues of mine did this when they found themselves in unfamiliar waters with a consumer electronics client for whom they were developing a social media campaign, the cornerstone of which was engagement with online communities and forums. And by engagement I mean the highly granular process of actually entering into discussions and responding to questions about the client's products with community members on the client's behalf.
Initially, my colleagues found themselves staring at an extremely broad universe. To avoid getting lost in space, they focused on basic goals:
- Finding the networks most relevant to the client
- Determining how to engage with and be useful to the audiences
- Tying that engagement back to actions on the client's site.
They succeeded in narrowing down the larger universe by applying basic relevancy criteria, such as the extent of client-related discussions and the frequency of postings mentioning the client. The visibility of sites in search was another important consideration, as it provided a measure of potential influence for one community site versus another (in terms of the likelihood of being found by consumers interested in the topic and/or the client's products).
With a more circumscribed list of sites to monitor in hand, my colleagues were able to move on to consider the weight and social authority of links and discussion networks, contextual references, visibility scores and tonal analysis as building blocks for measurable and actionable data. Bearing in mind that their primary aim was to improve the client's online reputation, they started with the number of sites they were monitoring for brand references and then the number with which they actually engaged. This helped to show the client how often opportunities arose and the resulting need for ongoing engagement. Then, my colleagues focused more granularly on the number of threads posted in a given conversation, the number of community members with whom they engaged on behalf of the client and the overall tone of those conversations. This helped to gauge the audience's level of interest in developing and maintaining a dialogue about the client's brand and the level of credibility my colleagues were able to forge with each network of users.
Each metric gave rise to a number of derivatives, all linked by the process my colleagues developed. These metrics may not be part of any marketing textbook -- yet -- but baselines can be established for each, and the evolution of each metric can be tracked over time. Even if these metrics currently lack the authority of the clickthrough or page view, they do constitute a stake in the ground in what will be a developing conversation. And as we all know, you have to start somewhere.