Over the course of the last few years, online marketers everywhere have started to recognise the importance of content as a marketing tool, not just for the benefit of search engines but also for the purpose of: building and engaging with an audience; increasing loyalty & retaining customers; increasing the performance of marketing channels among other benefits. Naturally as businesses start to invest in something, measurement systems are built to report on the performance of the content that they produce. With all the analytics software that is available, it is a bit of an oversight to spend resources producing content without following up and understanding how it actually impacts KPIs. But whilst this is great, and should be encouraged, often the ways that businesses measure the performance of content are very rigid and not applicable to all goals and metrics. For example, likes and shares as a metric is one that everyone is guilty of using. It’s an obvious KPI for a content piece to show how many people read it and then enjoyed it enough to tell their friends on Facebook that they like it. But actually if you want to increase the likes, shares, tweets or retweets of all your posts then all you need to do is paste as much content about: how awesome cats are; how difficult it is to decide whether Nicholas Cage is good or bad or how crappy Mondays are… Obviously that is quite tongue in cheek, but no matter how “on brand” you think your content is, if you’re only measuring on the appeal to your audiences sense of “what they like”, then you’re not necessarily telling them what they need to hear to increase their propensity to convert with your brand, which is what you want!
Beyond Social Metrics
Now I’m not trying to say that likes & shares are a bad metric to use, obviously it’s extremely important to produce stuff that people like and engage with through social because otherwise you’re boring! What I’m saying is that if you’re focusing on social engagement as a reporting measurement then you will have probably at some stage said a sentence along the lines of: “Yeah, it didn’t do as well, but that’s because it’s more of a brand piece”. Which is fine, but then what are you measuring that performance on? Understanding the fact that different content serves different purposes is key to having a well rounded marketing strategy. Yet so many marketers are then not following this through by setting different goal posts for different marketing purposes to understand how your content performs in that field. Here’s a useful post that I wrote that goes through the process of measuring content against business goals (conversion rate, average order value etc.). The principle is that, using Google Analytics, you can make direct comparisons between segments of your users that have been exposed to your content campaigns at different points in their journey with your brand. This is such an important point when it comes to understanding the stuff that’s on your site. There are many, many different kinds of users on your site and they behave very differently based on their previous interactions, exposure, intent, etc. Thinking that a content piece serves them all equally is not proper marketing and if you think that’s an obvious point then why are they being measured equally? The process of then putting this into practice and measuring the pages on your site according to what they were made for is not actually that hard, it just takes a bit more planning during the content creation process:
1. Identify your user segments
Depending on the scope of your marketing, the size of your site and the depth to which you want to analyse, you can have anywhere from 5 to 500 different user segments to use in your analysis. These could be segmented by level of exposure to different marketing campaigns or channels, entry & behaviour in different areas of the site, previous conversion history, you can go as deep into it as you’d like – the point is to identify & segment your key customer audience groups that you are trying to influence with the content that you produce.
2. Track everything as events!
No analytics package works properly straight out of the box, even the ones that take a lot of setting up, take more setting up in order to track the things that you should be tracking in order to get the performance data that you need. Simple exposure to a page does not necessarily tell you that a user has absorbed the contents held inside, in the same way that a “like” is not the only way to tell you that a user has paid attention to the contents. Event tracking through Google Analytics can be used to track loads of interactions on the page that give you more information about how the user has behaved on your site. Take advantage of this and measure as many desired interactions as possible because this can be both your performance metrics and also your segmenting tools to understand your content performance against the goals of the business (making money!). At the very least, you should be tracking:
- Scroll depth (has the user actually read the whole article? How long did it take them?)
- User commenting
- Clicks through to other articles (they’ve seen it, liked it and so read more)
- likes/shares/tweets/other social engagements
- video plays (and stops/finishes), pdf downloads etc.
3. Apply your conversions to your user segments
Once you have created your user segments and you have a complete list of desired actions on your site then you can start applying one to the other! There are two approaches to this side that each has value to find out different things:
- How much have different user segments engaged with content pieces – what’s the percentage of users who read the whole thing (scroll depth), what’s the engagement rate, follow through rate, etc.
- What is the influence of that piece on that particular user segment – does it increase their propensity to convert? Encourage them to re-convert? Upset and increase their value?
This is a case of identifying the type of user that the particular piece of content is targeting and then understanding how that particular segment both behaved with the article and then behaved after having engaged in a certain way.
Where to take this
The whole point of this approach is that there are many different types of users who come to your site, many of them might convert, some have converted and some will. On top of that, these are broken down into many different stages of buying or searching, which means your content will not serve them all equally. What you do with this approach is identify what the purpose of your content is, for different user sets and understand whether it has met that goal. You can then start to make better decisions on the content that you place around your site, based on the knowledge that you have on your user segments and their response to what you produce.