Well it’s not. Not really.
But that was the title of a talk I got roped into delivering a couple of weeks ago at Social Media Week and, although, there are no perfect solutions for measuring Social Media, there are certainly some steps that you can take (Kaushik recently published a magnus opus on the subject, with a particular focus on Facebook which I’d urge you to read).
But before diving in, we need to take a step back. The best way to identify the value of something is to measure how it delivers against its objectives. Its peripheral benefits can then be measured following this.
And herein lies my bugbear (and you CANNOT land this point strongly enough). Nobody has a clue on how to measure the value of PR, TV, radio or print.
But that’s been okay and they’ve worked for 100 years. Why should Social Media be held accountable for perfecting a measurement on ROI? It’s no different and can give you a ton of more information. Marketing instinct has gotten us this far without a foolproof safety net.
The onus isn’t on practioners to supply perfect reports – that’s never going to happen in our lifetimes. Not until somebody invents a robot who can crawl through our consumers’ brains and tell us what even they don’t know – why they act as they do. The onus is on organisations using the information to make good business decisions. Don’t like making a decision without a hard measure next to it? Well then all marketing and innovation may as well stop.
Secondly, is ROI even the main aim of most Social Media channels? Is that what we should be spending our efforts on trying to measure? Or are your Social Media properties trying to act as customer-service, CRM or PR channels?
Bearing this in mind, we can put a pause on spending gallons of energy in trying to build a perfect multi-channel attribution model, and take some very simple steps to give us a sense of how valuable our channels are.
The value of testing
If you’re not using any kind of testing structure on your channels, then shame on you. The insights you get out of Social Media can go a long way to better inform your other activities. Let’s take an example from the entertainment space, like LoveFilm. Say that they have a new film coming out but are not sure where to focus their advertising messages.
So the savvy Facebook community manager decides to set up a structured test to see what type of posts resonate best with users. Maybe she sets the test so that in one week the content is focused on a certain star, while the next week it’s focused on another. Then she can also test how other elements work – the cars in a new James Bond film for example, or the book on which the film is based (if she’s really on it, she’ll mix up her calendar across various cities through geo-targeting).
What should she be measuring? Kaushik’s four key metrics: your Appreciation Rate, Amplification Rate, Conversation Rate and Economic Value.
These learnings are undoubtedly valuable. So, what value can she show at the end of this? Well, there are a few that come to mind right away. By focusing on the right messaging in its advertising, LoveFilm can:
- Save time and money on only developing the correct advertising assets. Maybe you’ll find that that banner of James Bond in a speedo doesn’t really apply to your target audience after all.
- If LoveFilm are running Paid Media campaigns as one would assume, they have substantial testing budgets set aside. Well, Social Media has just saved on that testing budget.
- Similarly, they get to run better optimised campaigns from the get-go with lower CPAs.
The good bit? You can put a hard currency figure next to all three of those.
The value of cost-efficiency
To my earlier point, customer-service might be hard to measure as well. But if you are Lufthansa and are using your Twitter account for customer service, you can compare that to your call-centres. You should look to answer two things here:
- How much of a strain are you saving on your call-centres? Has having your channel meant that you can move resource away from your call centres?
- Proportionately, how much value does Twitter deliver for you over and above your call-centres?
You’ll have decent information on the first point, and you should have a decent opinion on the second that’s not too far from reality.
From here putting that £ sign next to your data or opinion is quite an easy task. You can obviously do the same exercise if your channel is being used for CRM or PR.
The value of sales
And finally, comes the tough one to crack. If you’re Volkswagen and are trying to identify whether your Facebook page is driving up sales, then you’ve got your work cut out for you.
The reality of the matter is that Facebook is very rarely a direct trigger to a sale. But, it is going to have some effect – just like PR or CRM. So how do we go about measuring it?
Well, I don’t have a perfect answer for this and I’m somewhat relieved to say that no one else does either. All that you can really do is to set up a number of tests, trying to identify whether your Facebook activity causes higher sales (remember: correlation does not equal causation).
The best thing I’d recommend would be to identify potential geographical test groups and compare correlations in your activity and sales across both the test and control groups. Unfortunately you’re never going to be able to exclude Germany or Spain. The local markets will rightly feel wronged if they’ve been left out.
But, will not targeting Birmingham or Leicester be as much of an issue? Unlikely. Maybe set up a test group of a couple of cities and monitor any sales differences in the cities that you aren’t targeting via Facebook.
That’s just one example, but the key here is to set up tests. The idea with all of these steps is to get a better informed opinion. Not to get perfect measures that will never exist.
When we start focusing on the former and stop obsessing about the latter, maybe then we can all move on to actually using data for something useful.