Following on from the post last month on understanding data acquisition, this post is going to address the fact that there is so much data out there – how do you know what you should be looking at, what good data looks like and which bits are genuinely useful. You are probably thinking this already, and you’re are right to do so. What constitutes good data depends on your objectives. For some it will be high numbers in social sharing or comments, for others it may be traffic to a specific web page or time spent on site.
What is good data?
It’s any data that allows you to better understand and deliver quality, contextual content to users.
What could this look like?
Data sample points that help you to understand your consumers/users wants, dislikes and behaviours:
Additional data-points to consider:
- Time on site
- Un-opened newsletters
- Platform performance
- Unfollows/unlikes
Too many look for high numbers across the board – particularly when using these figures to address KPIs. The issue here is that to achieve success in a certain space, to hit a certain objective, may mean that you have to sacrifice increasing figures in another area – and that’s okay. It’s important to set clear objectives and KPIs to reach those. Don’t just measure for measurements sake and cause strain to increase everything at once.
When looking at how people use your platforms or channels, for example, people may engage more with a post on Twitter than Facebook, even though the content was the same/similar, it’s key to take the learnings into account to ensure that you are delivering the right content types, topics and formats that the audience want to engage with.
The article you see shares for above (top left of image) shows that 2x more people were likely to share the post on Twitter than Facebook, particularly in the first 24 hours of the post being live.
Looking at the those figures (first 24 hours live) one might claim that the article was something people would associate themselves with and share out, so they must believe in/agree with the article, however it’s not something they want to share with their closer knit circle (Facebook) and not with their pure business circle (LinkedIn) which sits at zero. At the time, that may be a logical assumption, however when the same article was looked at a couple weeks later (see the sharing figures below), you’ll notice that Facebook has increased slightly and Twitter stayed relatively the same, however LinkedIn increased significantly, suggesting that the article might actually resonate even further with the more professional side of readers interests.
The key takeaway: Don’t look at your data to pull insights until you absolutely have to. Ensure you’ve given plenty of time for consumers to behave naturally with how and when they prefer to engage with your content.
But Twitter went so high, so fast, what’s up with that?
My hypothesis with 98% of the Twitter shares happening in the first 24 hours is that many people saw the article heading and shared/tweeted/retweeted without actually reading. They may have saved it to a read it later type app, however many people don’t actually take the time to read full articles anymore, and often take the risk of sharing any content that they think sounds like them from the title (yes, this can be dangerous, but that’s another post!).
Key takeaway: make sure when reviewing if your content is successful, don’t look purely at social shares alone. Make sure you overlay this information with data such as time spent on the page and bounce rates.
There is so much more that goes into understanding what good data is for you and how to use it to ensure it’s useful for your needs. In the next piece, we’ll look at the application of data and how to make it work for your needs.
