This post was written by Thomas Schonenberger from Mediavision who was our guest at SES London last week.
Thursdays keynote at SES London was given by Jim Sterne, founder of the eMetrics marketing Optimisation Summit and current chairman of Web Analytics Association. Jim presented us some thoughts on what and why we should be measuring to gauge our effectiveness on Social Media. When we consider that everything we do these days involves other people and the community in some form or another, it is really important to consider the context in which we define “Social Media”. Essentially all channels are social media! Any platform that allows content to be easily distributed (not necessarily online) – this is how connected humans communicate with each other. It is going to be around for a while to come! Embrace it.
Jim opened with a bit of background on marketing, considering the good old days of trading: How many chickens for that sheep? We would negotiate till we had a deal – it was done face to face and one to one. That was till they invented a price tag! Then came factories and shipping. Now we do commerce with people we don’t know. So we need to make noise to sell product and we call that advertising. Repetition of your message builds trust and you create desire. In the hay day of advertising you needed a Logo, a slogan and a jingle and you were set! That was the only channel. Now with the interweb, it is about making noise – not just about how loud, but where to focus that and how best to tune it.It is about measuring, especially that which you don’t control: Blogs, wiki’s, ratings, Facebook, etc.
In Digital everything is measureable – that’s the beauty of it. So we set about collecting data, and more data, and then some more. We have data coming out of our ears: Analytics reports, CPC, CPA, eye-tracking studies, mouse hover data. We think we are creeping up on knowing what people are thinking at every point. The idea being that if we can just get enough data then we will have the secret sauce to the next big thing…. Is a racket!! – Forget it! You need insight – not simply data and tools. And so far the best tools we have cannot compete with a human on this front.
Measurement doesn’t matter unless there is a purpose. Jim has a quote: “In a room full of 5 year olds I am tall, in a room of baseball payers I am short” – so what. What is the goal?
These typically are about raising revenue and lower costs.
And then you have some sub goals:
We can measure some key metrics:
And they all impact the bottom line:
12% reach * 3 exposures at $X = Y responses
12% reach * 6 exposures at $XX = YY responses
But how do we know if they see it, whether they read it or actually consume it?
Measuring the blogosphere
We can measure the noise on the blogosphere. There are many third party tools available to us. But how do we know whether they were human visitors or robots? Did they click by mistake? Did they actually consume it?
Measuring the twittersphere
We have things like RetweetRank – this fun but they are all ego numbers. Just because something is retweeted, doesn’t mean it is consumed.
This would be a measure of people’s familiarity with your brand. Do people:
You just need to go ask people! Do surveys. Even better – if they search for your brand they know you and you can correlate it directly to spend.
It is critical that you listen. Jim cited the example of “United breaks guitars” – a YouTube video. This news had to reach NY times before United Airlines realised they had a problem. By then they had a massive crisis. What would it have taken for United to have achieved 5 Million normal impressions? Then consider how much will it will take for United to clean up those NEGATIVE impressions?
How many people remember the above example? How long will it last and how do people feel about it? It is critical to understand how people feel? What do they think – will they recommend your product or service to friends and family? Do they like you or not?
We can get some online metrics such as positive/negative mentions. You can utilise tools to calculate this Net Promoter Score. Once again it is critical to add the human element. When people cannot even successfully decide if something is a positive or a negative mention, how can a computer possibly get it right! As an example, Jim ran a search term: “Obama”. The results claim that 10% are positive 83% neutral, 5% negative. Under closer inspection, we see some major fails. Much of the linguistic analysis is wrong – Computers just can’t do it yet. However somebody will succeed at some point but until then we need to apply ourselves. Measuring attitude is really tough!
Measure Rank and Authority. What is the extent of the influence? Is it in the thousands, hundreds of thousands, or are you Opera with influence over millions! We can measure the size of the network.
More importantly we can use this information and we can choose the people we want to spread the influence. We can get hold of bloggers. We can pay them, give them gifts, whatever it takes, but be smart about it.
It is also important to understand the different influences of topics. Some people will like the power of a new car while others like the style and yet others want to know how green it is. Suddenly people are talking about other things that you were not anticipating.
Finally though – it is important to remember that influence is momentary. It changes over time
Measuring doesn’t matter unless people do something about it. You need to have a response!!
Did visitors find your site before or after they saw your product in the blogosphere? This ties back into click attribution. At this point this is a tricky metric to measure.
There must be a business outcome – did the visitor do something? Did they take out a subscription, a membership, registration, complete a survey or were they just captured as a lead, did they make a purchase or are they a repeat customer or did they simply recommend you? All these actions are quantifiable and allow you to measure against your business goals and take action.
How much is a social media participant worth? You can compute it:
Eg. 10,000 people talking about lets say a shampoo product…
and if the lifetime value is $29…
and joining the conversation causes 5% to try it….
and 5% stay loyal,
then a participant is worth $7.25 and we have something to measure against.
It is about relationships, we need to listen carefully and we need to use that information to sell more of the marketing goals. Remember, raise revenue, lower cost.