James Devon, planning director, MBA
James Devon, planning director, MBA
A view from James Devon

Think BR: Measuring social - it can be done

Measuring social media might be difficult but it's certainly not impossible, writes James Devon, planning director, MBA.

A study this year by Econsultancy and NetBooster UK found that among over 300 UK in-house marketers and over 200 ad agency employees, nearly two-thirds of companies plan to increase their social media spending within the next twelve months. 

However, measurement puts fear into many marketers as self-professed gurus promise bumper numbers while there are no legitimately robust measurement tools in sight.

I am not the first to confirm that social media is certainly not easy to measure, but it is possible and it can be rigorous, even beyond the ever-vacuous ‘friend’ counting.

What needs to be established from the outset is what you’re setting out to measure and then you can develop activity that has relevant controls and tests to prove the value.

To be honest, I don’t think social media is helpful as a term for measurement because it is seen as a singular mass, both in terms of the way people use it and the way brands use it to further their objectives.

If we draw an analogy with TV, it is more than just brand promotions in their various guises, it's documentaries, news, films, party political broadcasts, etc.

And like the plurality of TV we don’t look for a single methodology to measure social media but evaluate accordingly - content broadcasters will, by and large, use different evaluation techniques to DRTV advertisers.

Rather than bumper numbers you’re looking for metrics that prove that your social activity has worked against your objectives.

And social media’s regularly reported metrics aren’t that helpful either. A ‘like’ can mean many different things, so assigning a single value to it is pretty useless.

To explain, you may ‘like’ a brand because you’ve loved them forever, but have no intention to buy (a Ferrari). It could be that you may like something in order to enter into a prize draw; or that you like something because you’ve admired an ad, or because it’s a cause you support.

There is no single motivation for a like, so there can be no accurate single value of one. Are likes causal or correlated to success? It’s likely that the answer is both and dependent on context.

So it is clear that all likes are not alike, which limits the value of it as a measure.

There is no doubt that social media measurement must be specific to the brand’s business and marketing objectives.

Take BT’s Twitter account, which is used specifically to achieve a business objective of providing cost-effective customer service. Recently at a conference, BT reported that traditionally it costs the brand £12 to solve a customer service issue but only 12p to do it through Twitter.

At MBA we believe there are four levels of metrics from which to extract data to move a brand closer to effective measurement, whatever the social activity.

  • Brand exposure

Brand exposure means the basics a marketer can calculate such as fans, searches and visits to views and mentions. Unfortunately, as with the 'like' example, such elements can be lost.

  • Brand engagement

This level measures how interested people are in the brand through retweets, comments on posts, shares and links. It also gauges how much people have amplified the effect by exposing the brand further by sharing it with friends and others in their networks.

  • Brand influence

The effect that the brand has on the consumer, in terms of things such as warmth of feeling, can be measured by gauging the share of voice, as well as using such measurements as sentiment analysis. After all a 75% share of mentions that are positive is a very different story to a 75% share of mentions that are negative.

  • Brand action

Brand action determines solid business metrics like cost to serve, leads, sales and cost of sale, to provide tangible results.

Even though there is no silver bullet to measure such a multi-faceted beast as social media beast it is quite straightforward to get indicative results.

The only thing is that more robust business proof is trickier. Brands must use established methodologies so that they can report against their specific business objectives effectively.

James Devon, planning director, MBA