Analyzing Personas Using Advanced Segments

Analyzing Personas Using Advanced Segments

I like getting insight from analysis. I like getting insight from analysing true new vs. true returning visitors (note the distinction from standard new and returning visitors) and from analysing visits that convert vs. those that don't convert. These are the sorts of analyses which can often give you the bigger and easier quick wins. But what I really like is analysing how the site caters for each of the different personas that visit my clients' website.

Definition Of Personas

A persona is a group of visitors that visit a website, that share a common objective, purpose or background which make them distinct from another group of visitors.

By understanding what proportion of visits belong to which personas, as well as starting to understand how well they are being catered for, you can start to gain real gems of insight which you can use as the foundation to make those really important small tweaks to your website which, cumulatively, can be a powerful, powerful piece of optimisation

Examples Of Personas

The easiest way to understand personas is through some handy examples:

  • Charity website - some visitors will come to the site to look to volunteer; others will want to understand how to fundraise. You might also get some media folk who aren't interested in doing either and just want to write about the good work the charity does.
  • E-commerce website e.g. Handbags - here nearly everyone will come to the site to look at buying a handbag. But you'll get some men looking to buy for the lady in their life; some women looking to buy for themselves; some women looking to buy but not for themselves; and you may even get a curious group of men who are looking to buy another man a bag.
  • B2B website - regardless of the industry, whether it's finance, incentives or call centres, the different personas may reflect the different levels of seniority of visitor that come to your site, e.g. finance assistant, finance director, business owner...etc.
  • Not for profit website - one of the best pieces of persona segmentation I've done recently focussed on a site like this. Visitors may come to the site from a consumer perspective, a business perspective (to sign up to the client’s scheme) or a media/fellow NGO perspective.
  • Our site - we have some visitors who come to us looking to use our awesome PPC services and others who want to work with us on their Analytics front.

Analyzing Personas

Using Advanced Segments. What dimension you use to create your persona-based segments depends on your own site, so there's no easy answer here - and often you'll use more than one method, but I'll outline some of the dimensions we've used before. I'll use Google Analytics in my example but whatever your web analytics platform, you should be able to utilise most of these dimensions (if you are our not acquainted with Google Analytics advanced segments check this video):

  • Keywords - this can be long and relatively time-consuming...but sometimes the best things in life are worth working for! If you can identify personas by keyword then often this can be combined with one of the other methods to create a more robust advanced segment. For example, in the case of our site, we could separate out those who come to the site on a search including 'ppc' vs. those that came to the site on a search including 'analytics'.
  • Pages or landing pages - if someone lands on one of our 'analytics' pages, it's fair to say they must be interested in one of our analytics services. We could separate out visitors who landed on our 'analytics' vs. 'ppc' vs. 'blog' pages to create different personas. So now you should be able to see how we can build out Advanced Segments; any visitor that searched for a keyword containing 'analytics' OR landed on an 'analytics' page, can go into the 'analytics' persona; the same for 'ppc' or 'blog'.
  • Events or Goal completions - if you have events or goals set up, incorporate these into your persona segments. For example, if someone clicks on one of our case studies on our CRO page, they could be classified as 'analytics' even if they didn't land on an 'analytics' page or come in on an 'analytics' keyword. If a visitor clicked our PPC training mailto link, they can go into our 'ppc' persona.
  • Referring sites - take a quick look at your top referring websites. It may be obvious to you what type of visitor they are sending. For example, if a visitor came to our site from the GA blog or Avinash's Occams Razor, we could conclude that these are Analytics visitors, as both of these blogs are analytics focussed.
  • Custom variables - these allow you to label a group of visitors. For example, we have custom variables in place to track whether a visit included a view of an 'Analytics' or a 'PPC' article. We can then use this to further flesh out our advanced segments.
  • Surveys - the final piece of the persona jigsaw are surveys - and if you don't already have an inkling about what persona groups are visiting your site (or what proportion), they are a great place to start! By asking a couple of key questions you can immediately understand how each persona type is being catered for on-site, plus more often than not you get some great qualitative data to add to the stats you get in your Analytics account. We've used 4Q in recent implementations, adapting the 'Purpose of Visit' question to better suit our purpose, e.g. asking visitors to identify themselves from one of the following options:

Survey Analytics for Personas

Now, as I mentioned, not only can we see this in 4Q's own interface, but due to its fantastic integration with Google Analytics, which sends every question and answer to GA as an Event, we can build these into the persona segments too!

So, hopefully I've built up a pretty good picture of how to build your personas. Now all you have to do is translate that into an advanced segment. I've given you an example of one below:

Advanced Segment for Personas

Closing Thoughts

Remember, as with everything in Analytics, absolute numbers aren't important - but confidence is. If you are unsure what sort of person would download document X then don't include that in your segments. You don't need to make sure that every landing page, event and keyword is covered in one of the segments, but you should try to cram as much in that you are confident about as possible. For the same reason you should be aware that it will be possible, using this method, to include one visit in more than one persona (e.g. visit came from a PPC keyword search but they navigated to the Analytics blog articles). However, if you are confident about your dimensions, this should not be happening often.

Once you've got your persona-based Advanced Segments you can start to analyse what each persona is doing on site and, arguably more importantly, what they are not doing. All the usual reports should be analysed, from landing page quality (look at bounce rates!) to most viewed pages; to what they are searching for on-site and what micro-conversions each group is interacting with. Remember to use the qualitative data in your chosen survey solution to back up the insight you've gleaned from Analytics and you will be well on your way to website nirvana!

Please let me know the results of your persona-driven explorations in the comments - and good luck!

Related Articles

  1. Google Analytics Custom Variables: Segmentation Power
  2. Data Segmentation with Google Analytics [video]
  3. Audience Segmentation - Giving Clicks a Personality

Subscribe To Our Newsletter For Monthly Updates

Your e-mail will be kept private

S.Hamel | July 2012

Great post Oliver - I'm always shocked when I present at a conference or in a workshop and I ask how many analysts are leveraging the concept of "persona" to do their analysis... I was even told that "personas are good in theory but not very useful for analysts"...

Oh yeah?!!! Why is it that most websites were initially built around personas and once launched, suddenly, we would end up doing analysis without looking back at them? That is ridiculous!

In the digital analytics course I teach at ULaval I refer to Andrea Wiggins article "Building a data-backed persona" where she explain how to leverage your web analytics data to enrich the personas that were initially created. Check out

O.Walker | July 2012

Thanks Stephane - much appreciated!

I'm equally surprised when we propose doing some persona-based analysis of a website and the client doesn't seem to have have considered their persona's half the time!

Hopefully through my blogs and your training courses we can start to make a small difference! :)

Nice article too - still relevant many years later. Thanks for sharing.

Post new comment
The content of this field is kept private and will not be shown publicly.
Refresh Type the characters you see in this picture.
Type the characters you see in the picture; if you can't read them, submit the form and a new image will be generated. Not case sensitive.  Switch to audio verification.
Online Behavior © 2012