Rely on Activity Measures from your Enterprise Social Network at your Peril

For most people “Social Analytics” means understanding consumer online behaviours. For businesses it is about understanding how best to take advantage of the social networking channels of Facebook, Twitter, Linkedin and the like, to enhance your brand perception. Enterprise Social Networks (ESN) have essentially evolved from public social networks, with the activity based metrics migrating as well, with little thought to their effectiveness. Activity measures are typically available “out of the box” with ESNs. This is a BIG problem though! It doesn’t take too much research to find out that the Enterprise objectives for ESNs and those for external brand building are substantially different.

We are not the first to send this alert. In 2013 Carrie Basham Young, a former member of the Socialcast team, wrote an article entitled “The Enterprise Social Networking Data Party is Over” , identifying many of the common misuses. What we do now have to reinforce this message is quantitative evidence, gained from our extensive Yammer benchmarking initiative.

Before we go there let’s first review the commonly stated objectives for ESN’s. All the major management research firms have conducted their surveys of executives and come to similar conclusions. From McKinsey: “According to executives, the social tools that enable employee collaboration—through real-time, group-based interactions that can be accessed across platforms—are most valuable.” From HBR: “The bottom line: the most important impact of social media technologies comes from who — and what — they empower, not just the information they exchange...not just the “better communications” business?” From IDC: “The critical point of success is ultimately how well relationships with customers, partners and suppliers are managed and maintained”. Collaboration, Empowerment, Relationships; nothing about brand messages, social media views, likes, follows or other common activity measures associated with consumer social media. Little wonder this mismatch in objectives is becoming problematic.

Now we do hear community managers often talk about usage statistics and the need to demonstrate growth in the use of the ESN platform, in order to justify the investment. And this is a valid use of such measures for early stage adopters, when the primary concern is usage, rather than value. It doesn’t take long however, to outgrow this need, in favour of the more important question of value achieved.Now to the evidence we have generated to support our argument for smarter enterprise social analytics. Over the past 6 months or so we have been able to study the Microsoft Yammer collaboration patterns of over 135,000 individuals across more than 20 enterprises around the world. In our last blog post on social cohesion across enterprises, we reported that this measure was the one that demonstrated the greatest variance in performance across our full sample of companies.  In essence, we measure social cohesion as the number of reciprocated interactions experienced by individuals, which are then aggregated to group, business unit and/or Enterprise levels.

Following on from this we took a deeper dive into a single enterprise to explore the social cohesion amongst the Yammer Groups. Of the 132 most active groups we found the same pattern of high variability in social cohesion between groups. However, when we overlaid the traditional activity measures like number of posts, number of replies, number of likes, it became obvious that there was no association between these activity measures and social cohesion within the groups. Additionally, we were able to gain qualitative feedback from the community managers that the social cohesion scores mapped well to their own perceptions of how the groups were performing.

In this chart, the groups are sorted by their social cohesion score. We can see from the graph that the total activity measure (the sum of all activity measures i.e. posts, replies, likes, mentions, notifications) has no association with cohesion. In fact, there is even a slight negative correlation. When we added an accepted measure of value, the response rate to posts, it’s the same story. In essence, if you are relying on activity measures to guide your collaboration and relationship building efforts, then you might as well not bother. In fact, it’s possible that you will do more damage than good.

Perhaps the most worrying part is that the ESN vendors commonly promote these activity measures as “measures of engagement”. Well perhaps this is the case if “engagement with the platform” is more important to you than “engagement with each other”.

SWOOP is a leading online social network analytics tool aimed at increasing organisational collaborative performance. SWOOP analyses collaboration patterns on enterprise social platforms and provides insights to every single user, community managers and executives. Contact us today for a free 2-week trial and benchmark report.

Image citation:  http://www.therightplanet.com/2012/11/talking-past-each-other/