Engagement vs. measurement
I went to a talk on measuring social media earlier this week, and was thoroughly disappointed. One of the things mentioned was that measurement should be built into the strategy of a social media campaign, to which my immediate thought was: you can have success metrics, but will people come to your site once it’s built? And then what happens to your forecast? I’m reading Groundswell at the moment so it may be a sort of ‘Groundswell hangover’, but the truth is that a project or campaign that is built for the client and not the people is looking at failure even before it launches. It may be the most beautiful Flash application ever created, but unless it brings value to the user, will they come the second time around, after looking at it once and murmuring ‘Oh, beautiful site’? So considering the metrics and having them as a guideline is good, but it is best not to have that as a benchmark of success.
Another issue that was mentioned was measuring interaction with a brand if it was not on the brand’s social media landing page. I’m not sure that’s even relevant. You may upload a video on your site, but you need to face the fact that thousands of people will find it on YouTube, and they may not visit your home site. If keeping it off YouTube is your solution, then think of all the lost opportunity. You cannot control what people do on the web. If they want to watch your video on YouTube or your presentation on Slideshare, then that’s where they will look for it.
Two thought-provoking things were mentioned though: one, that we need to be careful that measurement (or our obsession with it) does not kill innovation, and the second, that the language of digital measurement is quickly becoming the language of all measurement. I’m not quite sure about that last bit but I’ll leave it to bubble in my mind a bit.
I was interested when the panel discussed the 2 criteria that are important to consider while measuring the success of a social media campaign – the overall health of the brand it was created for (which is a long-term issue), and the effectiveness of the campaign itself (which is short-term). The overall health bit ties into what I said about benchmarking success earlier – overall health to a large extent is about engagement. I just heard about the Altimeter Group’s EngagementDB report, which effectively helps to measure the health of a brand (edit): via its engagement with consumers (thanks Mike!). It takes into consideration the number of digital touch-points with a consumer and the respective level of engagement within each channel. Bonus points were awarded for corporate/executive involvement because it meant that conversation with customers was a priority for the company.
Google Analytics will get us only so far. (Edit): Engagement will take over from there – and I’ll wager engagement is way more important an indicator of brand health than Google Analytics, which is probably a decent indicator of existence.
5 comments
There’s is such a disconnect in a lot of large companies between low-level brand managers and procurement staff. On the one hand the brand managers often cite ‘engagement’ as an objective, whatever that may mean, and on the other hand the procurement teams are using web analytics to measure pathetic meaningless things like ‘dwell time’ and repeat visits. This year I have seen on countless occasions “Increase dwell time from 2m 3s to 2m 5s” as a campaign metric. Now think about the quality of positive one-way ‘engagement’ achieved by Jonathan Glazer’s in an (admittedly very long) 60 second ad spot for Guinness.
It’s actually possible to cross-referencing registered email addresses, custom site stats and Dunhumby data to work out if cocking about on a website for 2 minutes effects your purchase decision, but ultimately you would need to be made of gold to be able to afford to do that.
Or, radical as it might be for most, you could whack up a tiny questionnaire and ask them if their experience has lead them to purchase. But for some reason, most agencies see loathed to do that.
Great stuff to think about. Engagement to me is all about commitment. It’s not enough just to have a presence on Facebook or Twitter, you have to commit to continuing the engagement. From a measurement standpoint, I think we’re all going to have to start thinking not just about the brand’s ROI, but the consumer’s as well. As they put more time into interacting with a brand, what are they getting in return? What additional value do they receive by following, friending or blogging about a brand?
Cool post, Anjali.
Quick thought: If we believe that all media is social, and engagement is the key to social media, does engagement because the key measurement by which we measure campaign success?
My knee-jerk reaction is “no”. Some brands ought to entertain and engage, others are best off playing it cool and not spinning their wheels (i.e., I might also buy Hanes socks, but I don’t want content, 360-degree sock tours, or a twitter personality from them.)
Of course, engagement can mean very different things for different brands. But ultimately, I’m not sure that all brands require deep online experiences.
And perhaps you’re saying none of that at all! But figured I’d throw in my two cents. Again, great post.
Fascinating topic and one that has been reverberating in “traditional” agencies for a while-how do you measure “engagement”? There are a number of equations floating aroud, usually involving something along the lines of the number of individuals reached times the amount of time spent with a piece of content divided by the cost of the activity. Effectively treating engagment as a measure of breadth vs. depth.
One might start by asking ourselves though not how we measure engagement but how, for this piece of activity, are we defining engagement? What does engagement look like for us? I think this is where the point about building measurement into strategy does have a role-not that something should be designed around hitting specific pre-existing metrics but that we should think, at the outset of any activity what it is we want people to think and do as a result. What behaviours do we expect to see and do we have measurements in place to track them? How will we know if something is working? What is the path to purchase and how are we monitoring how far people are getting along that path?
We can then construct bespoke measurement plans by campaign or activity, plans that acknowledge the complexity and diversity of how communication workds today and draw on a range of different metrics, hard and soft, to understand whether things are working as hard as they might.
Mike – I’m still unsure about what exactly is the best measure of engagement, certainly not, as you say, cocking about on a website for a bit longer than usual. But increasingly i feel most measures of engagement are likely to be better than measuring the number of hits on a site. Re: questionnaires, I seriously doubt how many people actually have the patience to fill them in – I do myself, when I’ve had a particularly good experience with a website or brand – but I can honestly say that 80% of those questionnaires go straight to the trash bin or if it’s on a site (‘fill in our survey and be entered into a prize draw’), i just close the window.
Rick – 100% with you on working on what the return to the consumer is. The days of the selfish brand (‘what’s in it for me’) are over.
Jared – Thanks! Yup, I don’t want to see a sock-brand on Twitter either and am increasingly getting tired of seeing irrelevant brands pop up on Twitter and Facebook. What you said about not all brand needing to be online is true – in the rush to jump on the bandwagon, people sometimes forget to think whether another mode of transport may be much better suited instead. :)
Pats – great point. This is something that agencies & brands need to think about before saying ‘this is a great idea, let’s do it’. I think targeting consumer behaviour is key to a project, rather than saying ‘if my site gets 100,000 hits I consider it a success’. There’s some fodder here for constructing a new model altogether. Must explore that!