AI CX Data Analysis

More than a pretty face – how WUA’s Digital Experience Platform uses AI to improve your CX insights

The WUA Digital Experience Platform is a launchpad for optimizing your website’s digital customer experience (CX). But getting the most value from complex CX research data isn’t easy.

There’s a lot of different data types and sources involved, so we use machine learning and AI extensively – to turn it into reliable insights you can use.

Instead of needing to trawl through massive quantities of heterogeneous data yourself, our AI tools enable you to immediately extract meaningful insights. As a result, you can make more impactful optimizations, faster – guided by the power of advanced machine learning algorithms and AI.

We’re very excited about the impact this has already had for our clients. But, as this technology is developing very rapidly, we think the best is yet to come.

Using AI to select the good data from the bad

Data selection is the first stage where AI is used, to ensure data quality. Each of our studies uses data from more than 400 people, but the initial pool of respondents is even larger than this.

For us, it’s essential to only use the best quality data, combined with the largest sample size.

 

For this reason, we analyze each dataset with a customized machine learning algorithm. It’s not ‘AI’ as most people imagine it to be, but a highly complex algorithm we’ve painstakingly developed to determine the quality of each respondent with high accuracy. This smart algorithm detects when participants drop out, cheat, or give poor quality data.

As a result, poor respondents are disqualified, and only high-quality data from 400+ qualified respondents is included in the analysis itself, so you get insights with real value.

Connecting qualitative data with quantifiable insights

To get a clear vision of how customers experience your website, we capture their entire experience. The result is a lot of data, including qualitative data from open-answer responses.

While these responses can tell you a lot about the rationale behind customer behavior, they’re also an ‘island’ of heterogeneous data that needs work to read, analyze, and understand.

So, the potential value is very high, but extracting it isn’t easy.

 

Because of the massive amount of data involved, we put specially developed machine learning algorithms to work on your behalf. These automatically connect hard, quantitative data with topics and match these topics with the open-answer qualitative responses from each study.

As a result, complex qualitative data is turned into something easy to understand, by automatically labeling and categorizing hundreds of open-answer responses, in seconds.

How AI gives you clearer priorities

Next, the WUA platform uses two remarkable AI-powered features that turn your complex data into clear priorities. These are sentiment analysis and impact analysis. Both of these utilize a complex machine learning algorithm that analyzes responses.

Using AI for this task saves many hours of manually trawling through individual answers. It ‘connects the dots’ between individual responses and the bigger trends from the entire market.

To understand which optimization has priority, an advanced algorithm calculates the real impact on conversion for each of these sentiments by looking at the entire market and measuring how each topic affects the eventual result. When competitor websites are doing better, the algorithm analyzes which topics they’re doing better on, and calculates how each topic affects eventual conversion.

This complex task isn’t something you can easily do manually, but AI and machine learning can do it with perfect accuracy in the blink of an eye.

The result: instead of reading hundreds of answers, making detailed notes, compiling them into topics, sentiments, and finally mapping their connection to conversion, you just get a summary of the most important ones.

Now, your priorities are clear. Thanks to AI, hundreds of responses are automatically read, digested, summarized, and labeled, so you can make sense of massive, complex data sets straightaway.

A clear vision across the whole market

Thanks to the power of machine learning, complex qualitative data is now linked to quantitative hard data. Each answer is mapped to relevant topics (making them easier to analyze), and you have an instant view of the average sentiment for each topic, and every website.

It’s a gargantuan task, so a machine learning model is the only way to achieve this depth of vision.

Across all sectors, we often see that the same topics are almost always in the top 10, but the relative importance can be quite different for each brand and sector. So, you really need a clear picture that reflects your situation.

To ensure we’re using the best data model, machine learning is also used to determine which model has the best fit, because this varies according to the sector and topics. This can be thought of as an ‘eye test’ for our AI, to ensure it’s properly calibrated to the market.

Unlocking the incredible potential of AI and machine learning

The data doesn’t lie, but it can be hard to get to the facts or turn them into a simple list of priorities. And this is where our advanced machine learning algorithms and AI can help you.

Our algorithms rapidly extract ‘meaning’ from hundreds of responses, putting complex and varied data into common schemas that illustrate the connections between sentiment, topics, and conversion, and give you clear insights and priorities.

This is just the beginning: as we continue to experiment with new algorithms and methods we’ll keep developing new time-saving enhancements and better insights. Some of these are already tantalizingly close at hand, like mapping performance trend data over time, integrating OpenAI and other language models to derive better analyses and redraw how topics are clustered, and developing new ways of tracking and comparing customer preferences over time. All these will help you give customers better experiences.

And it doesn’t stop there. Our future innovations will build on the vast amounts of market data we have collected over the last 12 years, from every major sector. This is a treasure trove for training algorithms and extracting trends.

Humans could spend a lifetime trying to unpick these patterns, but ‘big data’ like this is perfect for machine learning and AI, which can generate incredible insights in a few seconds.

Ultimately, our focus is always making CX optimization easier and saving time. By continuing to develop powerful AI-powered insights and tools, we can keep making this process even faster and more effective than ever before.

Start Benchmarking Today

See precisely how your website measures-up to other market leaders and direct competition within your sector.

Get in touch

Read our cases

MIE-Talk-scaled-1-1136×748-c

How ING improves customer journeys in an agile and global organization, using an optimized digital CX.

Read more
t-mobile-thuis-16×9-1

How T-Mobile Home Doubled their Conversion with More Successful A/B Tests

Read more
VGZ Screenshot

How Dutch health insurer VGZ is aiming for friction-free online CX

Read more