In this week’s masterclass Brandwatch’s Will McInnes explores five ways to master social media analytics. 

Everyone these days is “doing” data. Or is at least attempting to. And we all know that “doing” data is not easy.

It takes time, patience, precision, skill, logical thinking, and that special sauce that magically helps you find the right information in the ocean of social mentions.

If meaningful and successful social analytics is what you’re after, here are the top five ways to make sure you’re doing it right.

1. Get the right data

The social age has brought a wealth of data to marketers, but dissecting the parts that matter from the parts that don’t is difficult.

Beyond simple like and follower counts, the social web grants us the possibility to actually listen to what’s being said online – a data source that taps into the candid opinions of the public on a scale we’ve not witnessed before.

One inherent challenge in this is filtering out the useful parts from the rest of the noise. Different technologies will provide a variety of solutions to help you find the stuff that matters (and ideally not much else), but ensuring that you know how to take advantage of these tools is a must.

Whether it’s Boolean queries or simpler inclusion/exclusion fields, spend time on creating a search for data that pulls in the stuff relevant to you. If you have a tricky brand name, such as Next or O2, for example, it’s crucial that you tailor your search to gather the right data as there can be so much extra info that is of little to no use.

Depending on the industry you’re in, you’ll find anything from link farms to pharma advertising, counterfeit products, and adult entertainment pages. That, along with sorting through your data to make sure it is what you need, is why you can sometimes spend 60-80% of your time preparing your data set for analysis. It can be frustrating, boring and seemingly endless, but ultimately will save time and yield better results in the long term.

But whenever in doubt and on the verge of giving up, remember the golden rule – garbage in, garbage out.

2. Dig into it

It’s easy to analyse social media when you know what you’re looking for. You’ll most likely be tempted to go into the pool of data and look for things that prove your theories and beliefs.

It is far more interesting to find out about the things you don’t know are there. So explore and try to learn more, rather than just settle and pat yourself on the back for being brilliant and coming up with a solid hypothesis that checks out.

The most interesting findings will likely not sit at the top of the data pile, waving at you. You will need to put time and effort to find them. That is why it is crucial that you unleash your inner curiosity and dig, dig, dig.

A really clever way of doing that is looking for “white space” conversations where customers are not talking about your brand, your competitors or even your market or services. Look to find out what else they’re interested in, the type of language they use, the neighbourhood of products, brands and issues they do care about. 

Fair warning: you will need to ask a lot of questions and go to places you didn’t think you would go to (sometimes the data will take you there whether you want it or not).

3. Don’t try to measure everything

If you try to analyse everything, not only will you lose your mind but you’ll actually look at nothing. You need to focus on things that will help you make actionable decisions that are relevant to your business.

Instead of saying “we want to improve sentiment around our brand”, think more specifically. Pick a demographic that works for you,  think about who is key to your success and say for example, “we want to improve sentiment about brand x amongst 21-32 year old males, from the UK, who are interested in mountain-biking”.

Define your needs and objectives right at the very beginning of your analytics journey. It will guide you on what is important to look at and give you a more sophisticated way of measuring  And of course, be aware that they might change. But it’s significantly easier to tweak things once you know what you’re doing, than to just wander from one set of data to another with no clear direction.

4. Don’t take your data out of context

You’ll be surprised how easy it is to manipulate data. And you’ll probably be tempted to do it at times. After all, you just remove a few mentions from the mix. And they weren’t even that relevant anyway, right?


All you do is lie to yourself and cut corners. You can be sure that it will backlash eventually.

Imagine you organise a focus group and tell people to leave if they say something you don’t like. Obviously, that wouldn’t be representative. Neither is manipulating data.  You will get value from spending the time figuring out why the data is the way it is.

5. From “what?” to “now what?”

Long story short: findings are data, data is the “what”, insights are the “so what?” and smart insights help you decide “now what?”.

Findings themselves are not good enough. They are a great first step, but then you need to match them with your business objectives, your SWOT analysis and a variety of other components before you can transform your findings into actionable insights.

Solid analysis of clean relevant data will help you discover the “so what?” and that will help you power lead generation, respond to PR issues more effectively or simply help you make the right decisions and grow your company.