×

Developing Rapid Audience Personas Using Web Analytics

Topics: mHealth Monitoring and Evaluation
“Am I reaching the right people?” is a question that anyone who has ever been a part of a public health campaign has asked themselves at one time or another. It is also a question that is extremely hard to answer. 
 
One of the greatest challenges that communicators face is understanding and reaching their intended target audiences. This can be particularly challenging for digitally-delivered programs and interventions available to anyone with an internet connection. Since digital interventions have the advantage of being easily scaled to reach large numbers of people, it becomes especially important to use effective communication strategies to ensure digital interventions reach the right people.
 
One way to guard against reaching an unintended audience is to ensure that your target audiences are well defined and that your content, outreach, and products align with their needs. This is where audience personas come into play. Audience personas are basically a snapshot of a target audience member, using their current behaviors, demographics, psychographics (e.g., attitudes, interests), and lifestyle interests (e.g., love of movies, interest in fitness) to understand their information needs, preferences, and the context surrounding their interactions with your program. Audience personas can be built for multiple target audiences, but they generally represent your “average” targeted consumer. 
 
Traditional audience personas are usually fairly in-depth and can be built using a combination of different data sources, including subject matter expertise, media “listening” tools (e.g., pulling data from social media feeds), primary audience research (e.g., surveys), market research data, or data from your own digital platforms (e.g., social media, websites, mobile applications). Unfortunately, leveraging these data sources can also be time-consuming, expensive, and difficult to achieve without marketing expertise. There is always a catch, right? 
 
However, fear not, because there is another way to approach persona building! Using just a single source of data, such as website analytics, can allow you to rapidly create audience personas. While not as robust as traditional audience personas, these rapid personas do provide valuable data that can guide your outreach, content creation, product design, and can help you to identify gaps in your current audience reach. While you can use any number of website analytic programs to build rapid personas, Google Analytics, one of the most popular website analytics programs in the world, may be a good starting point. Google Analytics is a free tool, which is simple to implement on websites or mobile applications. It is ready to use immediately after implementation, provides a large number of robust metrics, is easy to customize, and can be used in combination with other free Google tools such as Google Search Console (provides data on how users are getting to your website via Google Search) and Google Data Studio (Google’s data visualization program). 
 
For the Smokefree program, a National Cancer Institute US based smoking cessation program, Google Analytics was used to create a number of different rapid audience personas for the Smokefree.gov website. These personas were built around various factors, including how users entered the website (examining differences in user types based on the origination of the visit) and if users completed core program actions (examining user types based on the download of a mobile application). Additionally, the filtering tools within Google Analytics allow you to examine specific segments of your audience. For example, you can filter so as to only look at users from specific geographic regions, users utilizing a mobile device to visit your site, or users visiting a particular page on your website. There are hundreds of different factors that can be used to segment the audience, and the factors you choose to segment and filter on are dependent on your overall goals for your website or program. Examples of Smokefree rapid personas based on audience segments that were built recently include general website users (considering all visitors to the Smokefree.gov website), website users entering through social media, website users who downloaded Smokefree mobile applications, and website users who entered the site by clicking on a link in a message in the SmokefreeTXT text messaging program. 
 
So, up to this point, there has been a lot of talk about what personas are, but you are probably wondering…how the heck do I build them? In the case of rapid personas using website analytics, it honestly depends on what you are hoping users do in your program, and the type of audience that you want to be reaching (total cop-out answer, I know). However, there are a core set of metrics that should be included in every rapid persona (see table below) and your time period for analysis should likely be six months to a year in length (to encompass seasonal changes or changes due to time-bound promotional efforts). The report should be visual in nature and simple to examine, so that anyone on your team could pick it up and use it. Each metric should have a visual counterpoint. A PowerPoint or very sparse Word document is usually sufficient. Additionally, the goals for your rapid persona should be specified at the outset of the process and align with your overall website goals. So, if you are looking to identify the type of users who are most likely to download a mobile application to inform promotional outreach, then that should be a specified and measurable goal.
 

Metric Definition Additional Notes
Demographics (Gender, Age) Data on age ranges and gender collected by Google from third-party DoubleClick cookies, Android Advertising ID, and iOS Identifier for Advertisers.

You will need to turn Demographics on in Google Analytics to begin collecting data. 

Google Analytics Demographic information is not based on 100% of website sessions and is Google’s best guess based on a number of different data points they collect. It is not recommended that the number of users (sample size) be reported for this metric, due to these limitations.

User Location The geographic location of a user based on the hit coming from their IP address.

Locations collected by Google are technically approximate, but still extremely accurate down to the city/town level. 

Acquisition Source/Medium Source/Medium describes where a website’s traffic comes from. The source is the place users are before navigating to your website, like a search engine or another website. The medium describes how users arrived at your content. Values for Medium include "organic" for unpaid search traffic and "referral" for website referrals.  
Mobile Device Usage The number and percent of users who used a mobile device to visit a website.  
Google Search Console: Search Analytics A series of metrics describing the interaction between Google search and your website. Metrics include search queries, device (mobile or desktop), click-thru rates (the percent of users who click on a search result after viewing it), and ranking in Google search results.   
Top Visited Pages The highest visited pages on your website, organized by number of page views.  
Affinity Categories Pre-defined Google categories that identify users in terms of lifestyle; for example, technophiles, sports fans, and cooking enthusiasts. These categories are defined to be similar to TV audiences.  
In-Market Segments Pre-defined Google categories that identify users in terms of their product-purchase interests.  
Average Pages Per Visit The average number of pages a user visits in a session.  
Average Session Duration The average amount of time a user spends on a website within a session.  

 

Using the metrics above, the Smokefree.gov general user persona for the time period of October 2016- March 2017 was the following:

  • Female
  • Aged 18-24
  • From California, Texas, or New York
  • Entering through Google Organic Search
  • Using the query “nicotine withdrawal” or “quit smoking”
  • Visiting on mobile device
  • Entering the website on the Homepage and visiting content on Steps to Take on Your Quit Day and Withdrawal
  • Visiting an average of 3.67 pages per session and having a 2:51 average session duration
 
 
Translating User Persona Data into Action
One of the key benefits of the rapid audience persona is the ability to immediately translate the findings into tangible actions. For example, using the persona described above, here are a few of the tactics that were undertaken:
 
  • Due to the interest in nicotine withdrawal from Google Search and the high number of visits to the Withdrawal page, withdrawal content was highlighted on the website and added to more social media outreach. 
  • Due to the large number of mobile visitors to the Smokefree.gov website over time, mobile users were targeted more heavily than desktop users. 
  • Examining results from Google Search Console, additional search engine optimization work was completed to improve visits to core cessation content and improve overall organic search numbers. 
  • Due to the younger audiences visiting the Smokefree.gov website, the tone of images found on the Smokefree.gov website was adjusted to be more in line with younger audiences.
  • Due to the geographic location of visitors on Smokefree.gov, social media posting times were adjusted to reach users from different geographic regions across the United States. 
 
These are just some examples of how rapid audience personas – derived from publicly available web analytics software – can lead to a quick understanding of your target audiences and can help to inform ongoing program work. 
 
First Name: 
Brian
Last Name: 
Keefe