- Buyer Personas for eCommerce Customer Research
- How to Create USEFUL Customer Personas Using Your Google Analytics Data
- HOW TO: Create & Automate Ecommerce Customer Personas Using Google Analytics & Tag Manager
We’ve discussed how to conduct customer research and its importance. We’ve gone over the importance of personalization and how ecommerce customer personas are defined. And we’ve dipped our toes into a theoretical introduction to using analytics tools for persona creation.
Now let’s get into the nitty-gritty.
We will now, as a practical exercise, present the step-by-step process of creating ecommerce customer personas by using Google Analytics and Google Tag Manager as the main tools.
Why those tools? Because they are an industry standard, used by over 70% of marketers to gather quantitative data. They’re also free, and for an SMB ecommerce website, they’re a great starting base to get the personas you need. More advanced tools can come later.
Tools that gather quantitative data can rarely be used to determine a visitor’s specific intent. Often, we have to resort to asking that visitor a direct question to understand their intent.
But, using a structure of goals, segments, event tracking, and pattern identification based on the same quantitative tools, we can deduce the intents of specific groups of visitors with reasonable certainty.
First, segment visitors in Google Analytics
Creating ecommerce customer personas begins by identifying common customer characteristics that are unique between groups of customers. The basic idea is to select a relatively small number of distinguishing characteristics (such as age group, frequently bought products, average monthly order value, gender, or location).
Google Analytics contains an invaluable asset straight out of the box: Advanced Segments. Segments rely on the data available to Google through some of their products, third-party cookies, and other sources.
When you enable the Interests and Demographics report, you gain access to this database and can use it to create segments of your visitors:
By playing with segments, you can create a simple persona, based on gender, age, geolocation, and/or device. You could also add interests to it. Pretty soon, you’d be able to create a few simple personas.
Here’s that process in action:
In the example above, we’ve created a person we call “John”.
John represents a male, aged between 18 and 44, from the U.S., who likes technology. Over 10% of users on the site correspond to this criteria, so it is not too wide (encompassing many different types of customers), nor too narrow (creating a glut of persona types) a classification.
Next, we’ll see how users like John behave, compared to all other users:
We can see that this group has significantly more value than other groups, with both a higher conversion ratio and higher AOV. Knowing this, it is obvious we want to attract more users like John!
So how do we go about that? We could use Google Analytics reports to figure out what interests John by viewing the landing page(s) he looked at:
As we can see, a significant number of users belonging to our “John” persona browsed the website looking for apparel, while a large number was also interested in electronics. We can use this information to present users in the “John” group relevant products as soon as they reach the site.
Another possible use of GA information is to view the Acquisition Report, and try to figure out which channel brings our website the most “John” users:
Although Direct is the channel that brings in most conversions, we can see that new visitors belonging to this persona mostly reach the site through Organic Search and Direct channels. This way, we can uncover the most profitable channel to bring more new “Johns” to the site.
Of course, this is only the fraction of information that can be uncovered by using a combination of personas and Google Analytics. And this is at a very basic level.
Next, let’s look at how to use Google Tag Manager and Analytics combined to define and create a persona.
Next, use Google Tag Manager to segment automatically
So, we know how to manually establish a persona and start comparing the behavior of that group to other groups or all the visitors to a website.
Now let’s make that happen automatically. That way we don’t have to spend the effort to define personas and use segments (as easy as it is). Instead, we’ll have the data right there, ready to be used.
To begin, we need to design a custom dimension in Google Analytics. This custom dimension will be our persona.
To define a custom dimension, open your admin tab in Google Analytics. There you’ll find a guide to implementing a custom dimension for your Google Analytics property.
Follow the guide and define the dimensions for each of the ecommerce customer personas you may want to use, depending on the nature of your business.
If, for example, you sell computers, you may want to create personas for office users, power users, home users, students, IT experts, and graphic designers.
That way, you can assign attributes to each of those ecommerce customer personas as they visit your website.
To assign personas, you’ll first need a cookie to be placed on your visitor’s browser. It will just perform one function — to count the instances a user accesses content that allows you to classify them as belonging to a distinct persona. The content of this cookie will be readable by Google Tag Manager.
To be able to rely on these results, you will need a persistent (or user) cookie that contains reports on how many times user accessed the site, starting from the first time they accessed the site. This allows you to track their interests far back into the past.
Combined with solidly defined personas, this data helps you create an automatic way of tracking and reporting the behavior of different personas. Plus, it will reveal the influence and performance that your personalization efforts and persona-specific content have on your users.
When you define common characteristics of customers, combining them into personas should give you a discrete number of groups, each corresponding to those defined behaviors or demographics.
Here’s how to create and deploy this type of cookie. Make sure your cookie tracks the specific indicators that allow you to classify ecommerce customer personas. For example, frequent access to content meant for power users, to use our computer-store example above, allows you to identify an individual as belonging to the “power user” persona.
Next, open Google Tag Manager and create a new variable. This variable will be a “1st party cookie type”. This will allow you to read the value of the cookie and assign this value to a variable.
As a final step, you need to configure a trigger dependent on this variable, and assign it to the custom dimension that you established earlier.
Once you’ve defined a trigger that is activated when the cookie value reaches a certain threshold, you will be able to create a new tag using the trigger to set a custom dimension value (the dimension with your named persona).
With this setup, every visitor to your website will, within some time, be identified as belonging to a certain persona group. The mechanism will work dynamically, and can even change the persona type of the visitor as time goes by and their preferences become clearer.
How to use ecommerce customer personas to improve your store performance
When you have defined and refined personas, it opens up a host of new opportunities to target each of the persona with appropriate messages, ads, emails or other types of content. To do that you need to create surveys tied to your personae in order to get the answers to why particular persona make their purchase decisions.
To achieve this, you cannot just ask them why the customers bought the product, because frequently their answer will be superficial and will not uncover their true motivation.
For example, you may ask a customer why they bought a power drill from your website (assuming you sell power tools). They may say they like the product, the company or any number of such reasons. Or their answer to such a direct question could be ‘To put a painting on the wall.’ Your customer is nothing but honest here, but nobody buys a drill just for this reason, as they can easily borrow one from friends and family.
While correct, these factual answer do not lead to true customer motivation. To uncover it, you need to get acquainted with their story, so you need to ask them how they solved their problem in the past, how frequently did they need the drill and similar, you may uncover the real answer. The customer does not want to depend on other people any more and wants to have a tool they could depend on in their own free time.
According to job-to-be-done theory, the true reason most of your prospects buy any product is to complete some specific task that will make their life better.
By identifying this task and job-to-be-done you can devise a marketing message that will resonate clearly with your audience and create content – from top of the funnel to the bottom of it that will consistently propagate this message according to each persona type you identify. By identifying the pain points (which in effect is a job to be done) in your personas and create a content that will cater to their specific needs, you will be able to present each persona type with a appropriate content.
Personality theories: A brief introduction
There are several frameworks for establishing personas, which I’ll touch on below.
One thing to remember is that all of these theories are based upon psychological personality types. As such, there is no “right” or “wrong” model. You should use the one that you feel best suits your customers and best explains their behavior.
Jung’s theory of personality
Swiss psychologist Carl Gustav Jung based his personality type theory upon two basic human traits: extrovert and introvert. By using four additional modifiers (sensation, thinking, intuition, and feeling), Jung defined eight personality types. Mapping groups of customers to these personality types will enable you to form groups of customers who correspond to certain patterns of behavior.
For example, Jung’s theory holds that extroverted people seek broader knowledge and information, and tend to rely on a great deal of sources to gather information. Introverted people will likely be oriented to prefer more specific sources of information, and seek narrower but deeper knowledge. You can take these tendencies into account when you present your product information.
Jung also posited that intuitive people tend to rely on hunches and knowledge that seems to come out of nowhere. They might appreciate copy written in the future tense, and more easily imagine the possibilities that your product or service can open up.
Sensation-type personalities, on the other hand, tend to seek firm data and independent confirmation in Jung’s view. They distrust hunches, and prefer that data are referred to in present tense with concrete and specific language — since this is information that they can use their five senses to perceive and understand.
Each of these combinations of personality can be used as a foundation on which to build content and user experience. Of course, you should use testing to back up your assumptions about how each persona will respond.
Other personality theories
Each outlines the common preferences of individual personality types, and allows you to identify the type of actions, products, and information each personality type is more likely to seek. All you need to do is to identify the action that corresponds with the personality type(s) of your customers, and you will be able to create accurate personas based on your customers’ on-site behavior.
Personality theory-based ecommerce customer persona creation in action
Let’s return to our computer-selling example for a practical application of personality theory-based persona creation. First, you’ll pick the personality theory or framework that most easily maps to your business model or what you already know about your customers.
After you conduct a customer research study and a few surveys, you should be able to identify your customers’ personality types according to the model you selected.
When you have identified the individuals who conform to each personality types, you’ll next go through your traffic and conversion reports in Analytics. The point and idea is to find specific patterns of behavior that differ each type from each other and define persona types accordingly.
Pitfalls to watch out for as you create your ecommerce customer personas
Using Google Analytics and Google Tag Manager in the ways I’ve described above allows you to automate the persona creation process. It’s not a perfect process, however, and you must be aware of a few potential drawbacks.
First, the process can only be as accurate as the number of visitors who allow cookies. You can probably assume that this will be a vast majority, and that visitors who do not allow cookies will not have a significant influence on the final result. You should be aware of this, and set a default persona value for visitors who do not allow cookies.
Second, the accuracy of ecommerce customer personas will also largely depend on the way you define the actions that classify individual visitors into personas. If you do this loosely, or set variables that only differ slightly from each other, your mechanism is likely to identify multiple personas — since most visitors will tend to examine every part of your website eventually.
You need to do diligent qualitative research here, which can help you structure your website in a fashion that lends itself to easy visitor classification. For example, setting up tripwires in the form of content likely to draw attention of one type of visitors and not the other types can help you establish some basic persona groups.
By using basic, data-based personas, you can structure your ecommerce store content, offer, product category pages, and product descriptions around your customers’ perceptions, needs, and concerns.
However, by using personality theories and types (such as Jung’s or the Big Five), you can take personalization a step further to create discrete, finite content types and user interfaces corresponding to the personality types you define. Personality types, if applied properly and for long enough to create as-accurate-as-possible results, can be a much better personalization mechanism than purely data-based personas.
By identifying each personality type’s preferences, and basing their experience of your site around those preferences, you will increase your website’s credibility of your website and boost your conversion rates.