Social media companies are attempting to solve one of many digital planning dilemmas with new social media metrics that measure “how social” a brand, company or topic is on the web.
A great way to connect with consumers online is to talk about topics that already interest them. Brands do this all the time…
BlendTec is constantly blending up things like iPhones, iPods, and other popular (but blend-hardy) consumer products that people are already interested in.
Other brands take a more subtle approach. Earlier this year we helped Volkswagen identify people and conversations around “hot hatchbacks”, and then use that information to invite people to a track day of their new Golf GTi.
Here’s the thing: if instead BlendTec blended CD-ROMs, and VolksWagen targeted “hot rods”, these brands would probably not have had the same success at reaching a mass audience. Their approaches worked because they entered into conversations that were already taking place. In other words, they took topics that were already very “social” and got in on the conversations.
The challenge for brands is to find those “happening” conversations and avoid wasting a whole heap of effort entering into a conversation that doesn’t even exist (after all, who would listen?).
But how do you know if the conversation is there and whether it’s big enough to pursue?
We’ve been thinking about ways to measure the size of these conversations mathematically, a social media metric we call “Sociability”. This metric would assign a value to a brand, company or issue that reflects how conversational that topic is.
As with many social media metrics, the trick is to differentiate between “conversations” and the “general web”. After all, Sociability is about conversations, not Google hits. So what’s the difference?
A conversation happens over time in a “flow” of comments, whereas the general web is just a bunch of documents, or a “stock” of comments. To figure out whether a topic is conversational, you have to look at the trending “flow” of comments around that topic, versus the entire “stock” of comments that exist on the web (after all, just because loads of people were talking about Atkins in 2002 doesn’t mean there’s still loads of conversation in 2010).
An example comes from a few recent projects we’ve been working on, where we tried to find topics of conversation that are large enough to pursue and analyse. A simple Sociability score could be based purely around the number of comments received on a topic in a recent time period.
For example, here’s some data we collected last month on a few topics and their respective Sociability scores:
| Topic | Comment bucket | Sociability Score |
| “Favourite airport” | 1-1000 comments | Tiny |
| “Doctors without borders” | 1001 – 10,000 | Small |
| “Amazon Kindle” | 10,001 – 100,000 | Medium |
| “Rome” | 100,101 – 1 Million | Large |
| “iPhone” | > 1 Million | Enormous |
This is Sociability at its simplest, and we’re looking at ways to make the measure more robust and meaningful, for example, by factoring in measures of engagement such as comments, downloads, time spent with content, retweets, social bookmarks, and so on.
As we research Sociability, it’s been interesting looking at how other companies answer the “how social” question.
For example, Infegy uses Social Radar, their proprietary crawler, to track blog posts, news feeds, forums, social networks and Twitter posts and then uses that data to find words and brands frequently mentioned on the Web. Their approach to “how social” is to look at the number of unique sources of content that posted content about a brand or word rather than the overall number of mentions (so that certain brands wouldn’t be heavily influenced by active fans who frequently post about the brand).
Their 2009 report found that that (somewhat unsurprisingly) social media sites, including Twitter, Facebook, and YouTube, topped the list of most social brands.
Another company is Vitrue, which looks at social media channels specifically – social networks, video sharing, status updates, photo sharing and blogs – to compute a topic’s Social Media Index (SMI). Their website has a nifty realtime app that lets you calculate and compare the SMIs for two brand names, phrases or keywords.
Earlier this year, Vitrue came out with their 100 top social brands of 2009. Vitrue’s list paints a slightly different picture than Infegy’s, with a stronger slant on television and media brands. Both, however, reflect Apple’s dominance online.
The prevalence of measures such as Market Sentinel’s Sociability Index, Infegy’s Social Radar metrics and the Vitrue 100 illustrate the growing need to understand “how social” a topic is, particularly in relation to other topics. Such measures could be the missing link between idea and execution for many brands. After all, when you can answer the “how social is it” question, you’re better equipped to know which conversations are worth entering, and which ones to leave in the dust.





I love the approach, but what worries me is that people sometimes don’t realise that something is interesting until we point it out.
For instance, we’d all agree that business cards are fundamentally dull. And associating business cards with a Samsung video camera seems an utterly lame idea. And yet the resulting campaign is just brilliant…
http://www.youtube.com/watch?v=FVq0HdiM-Ok
Hey Monica
I like the concept, a lot. Great work Market Sentinel!
Good work. My only reservation is that it seems to be weighted not to just online people but online people who spend a lot of time online. There’s not a lot that can be done about that at present but if we consider user data portability as the goal then that could easily be solved.
@Mike – A very good point. Like so many metrics of its time, Sociability really doesn’t mean much unless there’s someone there to tell the story, to explain why it’s interesting. Good news for digital marketers (at least we’ll have jobs!), but hard work too. The goal with metrics like sociability is to find something measurable to help support great ideas, and possibly offer some ideas as to how to help those great ideas spread (great Samsung spot, btw).
@Robin – Cheers!
@Charles – That’s why I liked Infegy’s approach to look at unique mentions rather than total mentions, but I agree, sociability needs to take into account that one person talking a LOT is NOT a conversation. Stay tuned for Sociability 2.0… =)
Monica, could you please specify the difference between sociability, engagement and social action metrics?
@Dmitry – Excellent question.
When we use “Sociability”, we use it in reference to particular contexts. For example “how social is conversation about favorite airports?” and “how social is conversation about e-book readers?” Sociability is a way to measure that.
Engagement, on the other hand, is about interaction with a specific brand, company, blogger, pretty much any web property. It’s more about clicks and time spent on a page by individual users, rather than interactions and conversations between users across different parts of the web (as is the case with sociability).
I hope this makes sense. Can you tell me exactly what you mean by “social action metrics”?
Monica, thank you
As far as I can see now, “social media actions’ refers to clicks, likes, adding to bookmars and favorites etc, so we can consider them the part of engagement metric.