How to improve customer perception with A/B testing

A 2020 consumer culture report found that 71% of consumers prefer buying from brands that align with their values.

But despite investing millions on customer surveys, social listening and third-party research to understand users, businesses have another way to validate what customers think really about them in real time.

A/B testing, combined with data analytics, shows how customers react to on-site changes, thereby removing gut feeling from measuring customer perception.

While qualitative methods like surveys and panels will always have a seat at the table, organizations need to run tests to see how customers actually react to changes in copy, branding, and messaging.

Without A/B testing and experimentation, brands risk campaigns having a negative impact on customer perception, with damage flying under the radar until it’s too late. There is also the risk of false positives, where some good results lead to you missing the things that actually bother your customers.

To find out how you can maximize the value of your qualitative user research with a powerful, AI-driven A/B testing tool, book a live demo with Kameleoon.

1 What is customer perception?

Customer perception refers to the sum of feelings towards and beliefs about a company or product.

Customer perception can be further broken down into objective and subjective perception.

Objective perception is what the customer sees, hears, feels, smells and tastes. Researching this would mean a clothing brand asking a user how a product literally feels on their skin.

Subjective perception is the customer’s opinion or judgement of the object. Consumers put an outsized importance on subjective perceptions, leading companies to invest heavily in promoting a particular image and conduction research into how to cultivate it.

There are many factors that can affect customer perception, such as culture, personal experiences and advertising—and a company’s “brand” can reap rewards well after it accurately reflects reality.

E.g. Safe car? Volvo. Even though Volvo isn’t much safer than average according to recent crash tests.

While a company can carefully shape its brand through strategists working with copywriters and graphic designers, upholding that brand through your website and communications is a continual task.

2 How is customer perception measured?

Large brands usually hire specialty firms to conduct different kinds of research to help them understand customer perception. Methods include:

• Likert scales
• Semantic differential scales
• Net promoter scores

 

This data is almost always collected by asking panels and survey respondents for feedback. There are many ways to do this, from online focus groups to on-the-street surveys.

 

Likert scales show you participants' agreement or disagreement with the asked statements.

Semantic differential scales illustrate where your participants' views lie on a continuum between two contrasting adjectives.

Net promotor score has become a foundational business metrics since its inception nearly 2 decades ago. It asks respondents one single question: “On a scale of 0 to 10, how likely are you to recommend our company?”

3 Why NPS and other qualitative metrics are not enough to understand customer perception

One of the most common ways to measure customer perception, mentioned above, is net promoter score (NPS).

Developed by Fred Reichheld, Bain & Company, and Satmetrix in the early 2000s, NPS relies on a single question to categorize respondents as “promotors”, “passives”, or “detractors” or your brand: “How likely is it that you would recommend our company/product/service to a friend or colleague?“

The problem with most satisfaction surveys is that the data is all over the place. There are way too many questions with ambiguous implications that the company will probably not even end up using.

NPS provides a single point of reference, correlated with growth, which a a company can benchmark against and focus on improving.

However, as CXL writes, NPS is dangerous because it only gives you one data point. “And you think you know your audience because you have a data point, but don’t realize how nuanced their interactions with your company actually are.”

For example, if you run an e-commerce store, you might get vastly difference scores depending on when your customer filled out the NPS survey you sent. If your site has easy-to-use, intuitive UX that makes purchasing friction-free, your customer might be very likely to reccommend you to a friend immediately after completing their purchase. However, if they are not satisfied with the product once it arrives, that quickly changes.

Without a way to rigorously test each part of the customer journey, you risk making wrong assumptions about where your company should focus on improvement.

The CXL article also goes on to say that subsequent studies on the NPS’s predicative validity—i.e. How accurate it is in predicting a growth—where underwhelming, indicating that it even underperformed other survey methods.

“It’s pertinent to figure out correlative metrics that are specific to your business,” wrote Alex Birkett. “Maybe it’s NPS, maybe it’s something else. But simply relying on NPS as a crystal ball is no way to forecast future growth.”

4 How A/B testing and experimentation can improve customer perception

Customer perception research can give you powerful qualitative data about your users. Experimentation is about turning that data into insights, which you can apply to your website or product with a range of techniques.

Example 1: Experiment with pricing to validate your product

Price is often the main barrier to purchase and therefore the likelihood of a customer recommending your business to others.

Contrary to popular belief, price experiments aren’t a cheap way to milk money ut of your customers, but a powerful optimization technique in your skillset as a CRO, marketer, or developer that can help you discover and solve business pain points.

• Are you failing to build customer loyalty by improving the product alone? You can try giving discounts to existing customers.
• Is mobile adoption lagging for your platform? Experiment by offering discounts to mobile shoppers.
• Does your software have low upsell rates? Remove and add features in the pricing table to discover what your positioning should focus on.

 

Pay special attention to the last point. Price experiments are not just a way to nudge customers to perform a desired action, but can also give you insights into whether your product development and marketing is focusing on the features your customers actually want.

Example 2: Use AI to find segments of customers most likely to convert

Your branding will never be everything to everyone. That’s why you need to narrow in on your ideal customer persona when assessing brand perception.

However, how do you know whether your ideal persona in in fact ideal?

Some A/B testing tools like Kameleoon come with AI capabilities which process data in real time and show businesses which of their website’s users are most likely to convert based on given triggers.

The Kameleoon Conversion Score detects opportunities and allows you to choose who you want to target in a particular A/B test or variation. You’ll see propensity scores that not only change in real-time, but also tell you who you should target and where the opportunity lies.

Example 3: Test simple on-page elements to see how different customer segments perceive changes

Uplift and growth from experimentation programs is most closely correlated with the complexity of tests and number of variants involved. However, if a company is still in the early stages of building a culture of experimentation, they should empower more teams to run experiments in their respective areas. Here is a list of some basic elements to test:

• Website headlines
• Call-to-actions
• Hero images
• Pop-ups
• Lead capture methods
• Input fields
• Alignment
• Field labels
• Design
• Accessibility
• Device type (mobile vs. tablet vs. desktop)
• Font sizes

 

Once an organization or team is comfortable running basic A/B tests, they can progress up the customer experience pyramid to more advanced techniques.

5 Conclusion: Complement qualitative research with quantitative

A/B tests are a staple of the digital experience/customer perception mix that lets you drive perception shifts that result in more sales, repeat sales, and increased customer loyalty.

A/B testing forces you to look for the strongest factors that seem to influence your customer perception in your data and hypothesize ways to drive changes. They let you deliver digital experiences that can change your customer perception.

Moreover, the experiences you deliver via your A/B tests, personalizations, and other forms of experimentation show you what your users do instead of what they think or feel or say in their surveys.

To find out how you can maximize the value of your qualitative user research with a powerful, AI-driven A/B testing tool, book a live demo with Kameleoon.