Here are all of the posts tagged ‘MeasurementCamp’.
Last Wednesday we hosted our second MeasurementCamp. It was very much a last-minute affair – we stepped in to host it after an appeal on Twitter the day before. Given the late arrangement it was a smaller crowd than usual, but at the same time it was intimate and very much like the first few MeasurementCamps – with fewer people we were able to hold it as a single discussion session.
I presented a case study on our recent Dunlop campaign, with a measurement-focused angle. The key learning was what we ended up measuring was different from the KPIs we had agreed at the start, owing to a change in circumstances – and that raw numbers don’t tell the whole story. For example, the audience for our Twitter activity in setting the record straight was in the tens of thousands, far less than the total audience for the blogs, but it was important to target them as they were in a chatty, lively community where misinformation has the potential to spread quickly.
We then had a breakout session where we talked about specific metrics, and how best to classify them. There was a consensus that different campaigns and clients need different metrics, but the question was raised of how to select them.
So we thought publishing this framework might be useful. The first classification – ‘traditional’ v. ‘social’ is relatively easy to make, but even then a ‘social’ metric varies from viewing a YouTube video to blogging about it. We then rate the metrics in terms of both engagement (how much effort a user puts in to an activity) and longevity (how long the effect of that activity it lasts):
Out of this you can start seeing how one might go about selecting the right metrics to best reflect the difference your work can make. If you are working on instant incidental awareness or viral spread, you can focus towards the bottom left, and if you’d push for a longer relationship-focused then you’d go for the top right where the numbers are smaller but the time and dedication greater. Of course, there is a lot of extra context that fits around this – sentiment, enthusiasm, trust, and existing relationships, which numbers alone cannot account for – but still we hope it helps frame better the different metrics out there and their relevance to your work.
It’s a question we get asked a lot, and despite the temptation to reply with Scott Monty’s (the head of social media for Ford) famed response – “What’s the ROI of putting your pants on in the morning?”, we usually say something more considered. We talk about how, despite the fact that we can measure the outcomes of the work that we do (and that we’re getting pretty good at it) and that we work with our clients to set meaningful KPIs at the beginning of engagements, it is still really hard to map those back to business metrics like ROI.
This is why we’re working closely with the rest of the IAB’s Social Media Council to get our research into the effectiveness of social media at a campaign level off the ground (you would not believe how hard it is to devise appropriate and affordable research methodologies to do so), and also one of the reasons we’re active participants at MeasurementCamp.
However, those that tell you can’t measure anything are wrong. If you’d like to know more, Jon ‘yongfook’ Cockle has a great presentation and accompanying blog post that outlines an approach to measurement that pretty much mirrors our own:
Billed as ‘an open source movement to make sense of social media measurement’, it sprung out of a panel discussion on measuring social media where Will McInnes and myself crossed swords on the subject, but unfortunately without bringing much enlightenment to the audience. Will, as he often does, decided something needed to be done, and MeasurementCamp was born.
As you can see from the photos above, we’ve come a long way since the original Measurementcamp last April, hosted in room above the Coach and Horses pub in Soho. This time, we invited our friend Josh Hallett of Voce Communications who was over from the US, and as result it looks like MeasurementCamp has now turned into a global movement. We think that’s a pretty fitting birthday present…
Earlier today, I popped along to MeasurementCamp (MMC) for a meeting with some of London’s top social media brains. This was my third or fourth MMC and it was good to see the progress made in the past year – there is a growing sense of what we’re capable of measuring, and now the focus has turned to how to structure that most effectively. Here are some of my thoughts springing from the discussion – all things we at We Are Social do and have in mind already, but MeasurementCamp was very good at crystallising them into a best practice-friendly format:
- Framing process clearly: Different channels produce different results – e.g. a customer support blog is going to generate fewer direct sales than a viral for a special offer. Dismissing those who misunderstand what a blog can or cannot do is not particularly constructive – it should be up to us to explain the process of a social media campaign and the likely effects it will have, to get realistic and significant goals agreed.
- Measure early, measure often: Measurement should not be just done at the end of campaign but as part of the process throughout to judge the difference a campaign makes. And wherever possible we should benchmark it against competitor brands.
- Not getting seduced by data: Because there are so many metrics available, grand plans for some ultra losing the wood for the trees is easy to do. Going overboard with too many numbers bamboozles us as well as the client and can make you lose sight of what works and what needs improvement. Pick & choose what you measure at the start and stick to it.
Further to the last one, there were some good ideas from the group, the idea that we shouldn’t be afraid to resort to the forms of measurement that the traditional advertising and marketing industries use, such as surveys or phone polls. These of course come with their own flaws but they still have many practical uses. And as with all these things, numbers can never tell the whole story, but they’re a good start in getting there.