Last week Twitter gave advertisers and users of its official analytics access to a new layer of data, wrapped up in a package they’re calling Audience Insights. If you’ve ever used Twitter advertising or their official analytics in the past, you’ve probably like me, been underwhelmed by the data on offer. I’ve always used data from Sprout Social and SocialBro to augment my understanding of client’s Twitter audiences, however, Twitter clearly holds the keys to the deepest data. Is the data helpful? Lets see, but first, where do you find the data?

Screen Shot 2015-06-02 at 09.41.56

Screen Shot 2015-06-02 at 09.47.31

That’s you in! So what data is available? Well, there’s two options here. The first one is ‘your followers’ which will give you more data, but is based on your whole audience. If you click the drop-down in the top left, it will also give you ‘your organic audience’ – these are the people you actually reach. That data is very important of course as you’re never going to reach your whole audience organically. For the examples below, I’ve used data from my ‘your followers’ section as it gives me more data to show you. The advice I give is valid for both ‘your organic audience’ and ‘your followers’, however I would say that the organic audience option should take precedence.

Pay attention to the ‘Match Rate’, from what I can find, this means what percentage of your audience could be matched within those categories. In my case, it’s usually very low, and that is a big caveat for all of this data!

Overview section > Interests

This section starts by giving you a bar chart of your followers and the growth of those over the past couple of months. Nothing that exciting really. The first thing that caught my eye is the ‘Interests’ section…

Screen Shot 2015-06-02 at 09.51.44

This clearly isn’t going to get that granular, however, it’s handy…

How to use this

If you’re a brand, then you need to know that your content sits within the realms of interest (did I just invent something?) of your audience/following. Of course, if you’re running ads, you can reach far out-with your followers (you’re likely reaching a staggering small % of them), however you need to make sure your content outside of ads is relevant to your ‘loyal’ posse. So for me, I’ve clearly got a number of followers interested in tech, business and marketing. I’m happy with that as those are key areas of interest for me and the vast majority of my content (apart from the occasional rant or misguided selfie) is centred around those things. If those top interest categories were cats, dogs and frogs, I’d be barking up the wrong tree.

Key takeaway – Make sure your organic and paid activity matches the interests of your followers. If your followers interest categories aren’t in-line with what you expect them to be, then you should be looking to shift your focus.

Screen Shot 2015-06-02 at 11.24.51

Overview section > Consumer buying styles

This is an interesting section as it brings out the consumer classifications of your followers. It would appear my audience likes to spend on premium brands. Must just be the way I put myself out there. Or not. When I dig a bit deeper into this, somethings really stand out. For example ‘Tex-mex/hispanic’ – I know why this is in there – I follow @illegaljacks and a lot of other people I’m connected to on Twitter also follow Jack, which makes me think that is where this comes from. Interesting data but fairly broad.

How to use this

If you’re a brand within a certain sector, then you need to ensure your audience is falling within that, this data from Twitter isn’t gospel, however if you class yourself as a ‘premium brand’ then you need that to be up there as number one or very close. If this isn’t the case, you should be looking to tweak your content but also your Twitter outreach work.

Key takeaway – Make sure your audience matches your overall brand placement

I had to break this up with something, right!

I had to break this up with something, right!

Demographics section > Gender

Screen Shot 2015-06-02 at 11.41.37

This nature of gender (in most cases) means that this is only going to have two data points.

How to use this

Fairy straight-forward this one! If you’re a brand that expects to appeal to or have a main customer base within a certain gender, and this data tells you otherwise, then you need to switch your approach. For example, you’re a product designed for babies, your target is parents, but predominantly Mums, if this company’s data matched mine above, I’d be worried as in the majority of cases, it will be the female in the parental relationship that makes the decision what to buy. This of course may not apply in same-sex parent families.

Key takeaway – Make sure your audience gender-type matches your expectations

Demographics section > Region and country

Screen Shot 2015-06-02 at 11.57.06

Screen Shot 2015-06-02 at 11.57.06

Here we’re seeing Twitter’s best guess at where your followers are based. As a Twitter user, there’s no prerequisite that you tell them where you live (unlike Facebook), so this is based on geo-location and is the reason that the geo-targeting isn’t as tight within Twitter ads as Facebook.

How to use this

If you’re a business working at a ‘local’ level, then you clearly want to see your immediate area of coverage within the Region data. However, this isn’t nearly granular enough. Scotland is a ‘region’ apparently. Take this as an indication of course, and if you know your business area is say, Greater London but that’s not featuring high up on this list, then you need to start realigning your audience. This of course applies to the country data too.

Key takeaway – This is an indication of geo-location but isn’t as accurate as we would hope, however its useful as an indicator.

Within the Demographic section there’s a lot of data around income and worth. Net worth, household income categories, home value etc. I don’t know how they are surmising this data, but I’m really not sure of its validity. Handy as a guide for sure, but I wouldn’t take it as read.

Lifestyle section > TV Genres

Screen Shot 2015-06-02 at 12.08.14

I covered the ‘Interests’ section in the Overview section of this post. This TV genres data interests me though!

How to use this

My followers are in to drama and sports in a big way. This data could be used to form your Twitter content. For example, my brand makes popcorn for the home, Twitter tells me that drama is high on my follower’s list of TV genres. Game of Thrones is on tonight, the buzz is huge, my followers are in to drama so lets create a nice Twitter image relating to GoT and our popcorn. The right approach here could really hit home with your followers, but also extend your reach if others pick up on it.

Key takeaway – Use this to mould certain areas of your content approach, but don’t rely on it!

Consumer Behaviour section > Consumer goods purchases

Screen Shot 2015-06-02 at 12.23.23

Ice cream & novelties? What a strange category!

How to use this

Selling a product? Does it fall within one of the top categories here for your followers? No? Either your messaging is all wrong, or your ad targeting is way off, but this needs fixed!

Key takeaway – Handy for making sure you audience sits within the right good categories, but the categories are a bit strange!

Mobile footprint section > Device categories

Screen Shot 2015-06-02 at 12.29.16

Mobile, mobile and more mobile. You know how important mobile is!

How to use this

Couple of things here – my followers are heavily-weighted towards iOS devices or in other words, iPhones/Pads. I could use the YouGov profiler to take a look at what else users of those devices are in to…

Screen Shot 2015-06-02 at 12.33.18

The sample size from YouGov for this is over 7000, so not a bad data-set but as ever, it’s not gospel. I could look at this insight and put two and two together and start to shift parts of my Twitter content to fall inline with those. For example, talk a lot about hummus (Waitrose) and bad driving practices (BMW). I jest. Sort of. You get the idea though. If I flip this and use the profiler for Android, we see very different data and interests (sample of only 190).

Screen Shot 2015-06-02 at 12.38.44

Big differences and more info that can be used to help mould the right approach for your content.

The other key thing here is that you should just generally be thinking mobile-first with your content – images are key, what time of day you are posting and so on. That’s a whole other article!

Key takeaway – Think about what the device tells you about the person and test certain approaches based on that.

There we go, a taste of some of the data within Twitter Audience Insights.

What else does it do?

The drop-down menu in the top left gives you access to ‘all Twitter users – this supplies data across all of the categories for the whole of Twitter, currently only using state-side users. You can compare this to your audience by selecting ‘Add comparison audience’, but again only for the USA.


There’s no doubt this data is helpful, however you should never base all of your activity on this. You need to test and learn from your actual interactions and results from your day-to-day Twitter activity – this data is unrelated to response. It has limited functionality as it stands, however it is early days!

If you’d like to hear more of my thoughts on Twitter Audience Insights, or need support with your social media advertising, please do contact me!

Non Twitter etc Images via Shutterstock.

blog comments powered by Disqus

Get in touch