Poor Man’s Guide to “Persona-lized” Testing
Published: March 25, 2015
Author: Theresa Baiocco
Your site’s visitors don’t all think, behave, and most importantly, convert, the same.
Different segments of visitors respond better to different triggers. So if you’re lumping all of your visitors together in your A/B tests, you’re getting the average conversion results.
But averages lie.
Even if you don’t have the budget for personalization testing software, you can still personalize your A/B tests. Here’s how:
Create Hypotheses for Tests Based on Personas
Be honest: how many of you created (or paid a market research firm big bucks to create) customer personas…that you’ve never used? Now those personas can pay you back.
If you don’t have customer personas, follow this process for creating them quickly or read The User Is Always Right by Steve Mulder.
You can be more thorough, but at a minimum, make sure your personas have these basic elements:
-Descriptive background that paints a vivid picture of the person
-Where he is in the buying cycle
-What he’s looking for on your site
-What you want him to do
-A quote that summarizes his problem
-A real picture – not a drawing or a cartoon – that makes the persona feel real
Here is a high-level overview of a few sample personas I threw together to demonstrate how to create testing hypotheses by persona. Your personas should be more flushed out, but this gives us enough to work with.
Let’s imagine these sample personas are for your hypothetical business that provides nationwide public speaking coaching.
Your services include:
-One-on-one coaching in person
-Team virtual coaching
-Low-cost, self-paced courses with recorded video tutorials
Notice how drastically different these personas are – which is exactly what you’re aiming for.
To create your testing hypotheses, imagine selling public speaking coaching services to these people face-to-face. You’d take a very different approach with each one:
Be encouraging with Pastor Pattie and build rapport:
-Ask her questions so she feels like you understand her
-Tell stories about clients you’ve helped like her
-Show respect for her calling in life
-Help her imagine what it would be like to powerfully deliver her sermons
-Give her good resources for free. Encourage her to shop around
Appeal to Consultant Chris’s ego:
-Recognize that he’s already comfortable on stage. Commend him for that
-Name-drop famous people you’ve helped
-Show how you can help him reach his goals within a certain timeframe
-Explain how investing in your services will yield a positive ROI
-Explain why you’re better than your competition
Take a no-nonsense approach with Sales Manager Sammy:
-Get right to the point; minimize small talk
-Convey that although you’re expensive, you’re worth it
-Demonstrate professionalism, consistency, proven results
-Apply pressure to get him to buy now
Here’s the Problem with Lumping Them Together
Let’s say you want to test which tone increases phone calls:
-Appealing to the customer’s ego
So you run an A/B test for all of your site’s visitors and find that the tone of encouragement has the highest conversion rate.
But if you were to run this test on each persona, you might actually find that only Pastor Pattie actually converts better with that message. And because you sell more low-cost courses to Pastor Patties, the winning message for her outweighs the winning messages for the other personas in your results.
In fact, if you had segmented your test by persona, you’d find that Consultant Chris actually converts better by appealing to his ego, and Sales Manager Sammy converts better with a no-nonsense approach.
And the services that Consultant Chris and Sales Manager Sammy buy are more expensive, so you would actually lose revenue by implementing a message that won with your highest-volume, but lowest-dollar, service. This is why it’s critical to test for revenue, not just conversion rate.
Path for Each Persona
Let’s assume you have a landing page for each service you offer. Your ad groups are set up with appropriate keywords and ads to match.
The fastest way to test your hypothesis is through CTR of ads.
In order to test your hypothesis that Pastor Pattie needs encouragement to sign up for the self-paced courses, for example, you’d A/B test two ads: one that simply talks about the courses, and another with a more supportive tone. If your hypothesis is validated and the ad with the supportive tone gets more clicks, you implement that language on your landing page.
You’d repeat the process to test whether Sales Manager Sammy responds better to a no-nonsense approach.
And again to see whether Consultant Chris is more likely to buy private coaching with a message that appeals to his ego.
Once you determine the winning tone for each persona, continue flushing out nuances in the copywriting by A/B testing the landing page.
Test Geographical Stereotypes
To get even more personalized, test different messages for the same persona in different locations with geotargeting. For example, you hypothesize that you can be more bold and direct with Consultant Chris if he’s in New York than you can if he’s in California.
So you set up the following A/B test on your ads about private coaching services in NY, then duplicate the test and run it in CA also.
Ad #1: Control
Public Speaking Coaching
Become an Even Better Speaker Today.
6-Week Private Coaching. Call Now!
Ad #2: “In Your Face” variation
Good at Public Speaking?
You Make Mistakes You Don’t Realize
Call Now. Learn to Speak Like a Pro
Notice how both ads appeal to Consultant Chris’s ego, but Ad #1 is more positive, whereas ad #2 is more direct.
Your hypothesis is that the control will win in California, but the “In Your Face” variation will win in New York. If that happens, you need a different landing page for Consultant Chris in those cities, which matches the tone of the ad.
And for the Nay-Sayers Out There…
I know someone reading this article wants to write a comment about how I’m missing the outliers by assuming that everyone who buys private coaching for public speaking has an ego like Consultant Chris. And everyone who buys self-paced courses needs support like Pastor Pattie. And everyone who buys team virtual training responds to a no-nonsense approach like Sales Manager Sammy.
You fear that by writing to Consultant Chris’s ego, for example, you’ll be turning off other, viable prospects with different personalities and different motivational triggers.
Here’s my response to that:
Don’t worry about the outliers. Let them go.
If you try to be everything to everyone, you’ll end up being nothing to no one. Build excellent personas. Know them well and sell to them. Don’t waste your advertising budget on anyone who doesn’t fit, because you’re not targeting the average visitor; you’re writing for the best potential customers.
So go on, dust off those personas and try this testing strategy. Write a comment and let me know how it worked out!