Filter Metrics: analysing campaigns and topics by filtering content

Filter Metrics: analysing campaigns and topics by filtering content

In content marketing, it is very common that one channel will share many different kinds of content. Maybe a certain brand has a number of different products, or events come up, or news from different departments in the company are shared, the possibilities are endless.

To deal with it, the analysis sometimes needs to separate these different content pieces to better understand how each broader topic is relating to the community and the market in terms of acceptance and general performance.

There are essentially 2 ways to do that:

  • Pre-Publishing: using tags on a publishing platform.
  • Post-Publishing: using filters in analytics.

The option 1, of tagging content Pre-Publishing, will usually work best when the content creator or team is very organised and is also building a bank of published content to be easily accessible in the future. The task of tagging will involve creating a standard for terms being used so that the content pieces can truly match the topics they are involved with. While this is an interesting strategy to be used, it is really a lot more connected to the management of content, content creation and publishing than with analytics. The reason is that in analytics it is very likely that a more dynamic approach will be needed.

Option 2, where the content is filtered post-publishing by the analytics platform, will likely be a better way to approach filtering in analytics. It can be much easier to work with the data that way until we find the insights we need. These are a few advantages that can be pointed out when going for option 2, and filtering the content for analysis after it has already been published:

  • No need for pre-publishing standardization efforts.
  • No-stress on the content team to match each tag perfectly.
  • Dynamic categorization of content – follow project needs.
  • Match and create new categories for competitor benchmark analysis.
  • Dig deeper and find categories within categories.
  • Automate the entire process after it is done – for future analyses.

It is likely that you will find even more ways of taking advantage of filtering mechanisms in analytics.

On a practical side, how exactly can we go about filtering content? Essentially we will have 2 ways of doing that:

  • Fixed triggers – such as hashtags and “@” tags.
  • Custom triggers – any letters, keywords or combination of keywords.

Fixed triggers are great because we don’t have to do any manual work to make it happen. We simply tell the analytics platform to fetch all the hashtags, or all the tags and it will bring us everything it finds. This is a great way to also discover unexpected uses of content very quickly, and be able to react strategically to what we find out.

The following ‘Hashtag Detection’ metric by quintly analytics is one example for using fixed triggers:

You will notice on the example of the ‘Hashtag Detection’ metric that many of the hashtags in the list are non-branded. This means that anyone can make use of such hashtags and get into the trend of good performance surrounding it. So this type of filtering metric can have strategic significance to the content creation team as well. It would need to be used before the new content is created, naturally, or in-between cycles of production. It will therefore defy the common assumption that analytics is only good for a post-campaign analysis. Sometimes the greatest competitive advantage with analytics comes from a pre-campaign stage.

Taking now the second possibility, we can look into a simple keyword filtering as an example. What I am showing here is an analysis of Sports Media pages, ESPN, CBS Sports and FOX Sports, and filtering from all of their content only the posts that relate to my keyword – NBA.

I ranked the content table by ‘shares’ in this example, and you can notice that ESPN is dominating in terms of number of shares in the period compared to FOX and CBS (which doesn’t even show in the top 7 posts).

The interesting point here is that I did not have to crawl through their pages to get this result. These media pages post about all kinds of sports, and it would take a lot of time for me to filter their content only around the NBA. There is also a chance that a human being could miss a post when trying to do this filtering manually.

Revising the uses for such filters, we can think of examples such as:

  • Filter our own content by topics for our internal analysis.
  • Filter content from external media channels that is related to us.
  • Filter influencer content to see how much engagement they generate to us.
  • Filter content by our own executives to see how they generate brand engagement.
  • Filter competitor content to match only certain product categories during analysis.

Because analytics covers 100% of the content from these channels, this process is guaranteed to cover everything that happened in the period. So while many marketers rely on Listening technologies alone for such searches, analytics is really the first step we all should take in such process. Listening, in this case, would come as a second step, when we already understand how the channels themselves are performing, and want to go beyond that and into audience generated content.

The example of the NBA keyword is very simple, but you can already understand how such mechanisms can help you save time and filter anything out of the content of any social channel. This level of freedom is unheard of in traditional competitive analysis, and can really promote strong strategic insights for the managers behind the marketing plan.

In Conclusion:

Analytics will sometimes hold secrets on its capabilities if we don’t research further into what we can do with such tools. It is worth to keep an open mind and approach such a tool with a hunger for knowledge. If we are spending a lot of time in one process (such as filtering content manually) we can bring that problem to the people behind analytics platform providers, and let them share with us what such platforms can do. We will always lose when we assume that we know what can be done and we truly don’t. The more we push the tools, the better they get, and the better our campaigns and data-driven marketing can be.


Recommended review: In case you are working in social media analytics, look into quintly analytics to make use of such dynamic filtering mechanisms for your analyses.

Recommended Read: Social Media Analytics Strategy, by Apress

For more content around analytics please follow

1 Comment

Comments are closed.

  1. […] Once we understand the value of this analytics process for audience insights and how such “triggers” or metrics can point us to an understanding of our audience, we will understand why certain metrics might have superior value in our analytics process. This is the case with filtering mechanisms, for example. You can read more about that in this article:… […]


Get in touch, will be great hearing from you!


Log in with your credentials


Forgot your details?

Create Account