The mapping tool that can help your social media get to the point

Every marketing activity is followed by two questions: what happened and what did what happened actually mean? Sometimes the two answers are the same, but often there is too much going on in the market, or the things we’re measuring are abstractions or proxies for what we really want to know (“Did sales go up?”), so that the answer to the second question becomes vital.

Digital marketing is no exception when it comes to these two questions. In fact, it’s usually much harder to answer them for a digital campaign than it is for any offline activity, mainly because so much data is generated that it’s hard to pick out what’s important.

The question of meaning is also clouded by how new a lot of the techniques being used are. Back in the dotcom boom, I interviewed someone in digital advertising, who told me that he had loads of results data, but digital advertising itself was too new to know whether those results were good or bad. The subsequent collapse in clickthrough rates for banners highlighted another aspect of the problem; a lot of what looks like good results is just due to the novelty of the medium.

Of all the digital channels, social media currently has these problems in spades. When you run a campaign with a social element, it can be really difficult to know what the effects are and what the changes mean. You can look at sentiment analysis, you can look at buzz, but it’s hard to know how things have changed out there in the social space.

So it was fascinating to come across the work being done by the Social Media Research Foundation at the Local Social Summit last November. The key thing the foundation has done is to create an open tool (called NodeXL) which works with Excel spreadsheets to allow anyone who has data on people talking about something in the social space to create a map of the relationships between them.

Marc Smith of the SMRF illustrated what the tool can do using examples from US politics, but it applies equally well to people talking about brands. Having created the tool, the foundation is now encouraging people to use it, and to share the graphs they create via a site called NodeXL Graph Gallery. Smith’s ambition is that this will eventually result in a taxonomy of social media, a set of examples allowing anyone with an interest to identify what kind of network they’re dealing with.

Of course, identifying individuals within communities is nothing new. Malcolm Gladwell’s book The Tipping Point introduced us to the different types of connected individuals and their power as influencers back in 2002, and since then social media agencies and marketers have been looking for and engaging with these people. But one of the other aspects of the SMRF’s work has been to show where these people sit on the map, and what the nature of their influence is.

It turns out, for example, that the simple number of connections isn’t a reliable guide to degree of influence. Someone with relatively few connections can wield more influence than somebody with many, if through their few connections they can link to otherwise unconnected groups.

But where the SMRF’s graphs really get interesting is when they start to show changes over time. The example Smith uses is the comparison between a brand and a community. He believes the graph of a brand in social media is typified by a lot of isolated points, representing people who have mentioned the brand name, but who aren’t connected to each other. A community is the opposite; the typical graph shows a group of people very connected to each other. By looking at the graph before and after a piece of marketing activity, he suggests, you can see whether what you did had a brand-building or community-building effect.

This means that not only will marketers have the ability to compare their own graphs against a taxonomy to see what kind of social media situations they’re dealing with, and perhaps make predictions of behaviour based on that, they’ll also be able to watch how the situation changes as they interact with it.

The other example Smith gave at the Local Social Summit was of ‘answer people’, the brand enthusiasts who act as an informal help desk online, answering people’s questions about the brand. That sort of behaviour can be hugely valuable to a company; indeed mobile phone operator Giff-Gaff has formalised it as part of its business model.

As Smith points out, the graph of that sort of community has a distinct shape, which means that if you keep mapping the community you can tell when the structure breaks down because the answer person has quit. Otherwise all you’d see would be a massive increase in the number of calls coming in to your help desk with no explanation of why.

And if all that’s required to get that answer person back on board is, typically, a few kind words and some branded items, that alone is information worth having.

Michael Nutley is a writer and consultant specialising in interactive media michael.nutley@blueyonder.co.uk