Historically, customer preferences have largely been built around what customers say they want, captured through surveys, forms, and CRM records.
But customer behaviour often tells a very different story, and the way customers move between channels, repeat contacts, abandon self-service, or escalate issues all reveal valuable preference signals that traditional methods often miss.
To find out more, we asked Chris Mounce, Product Training & Enablement Specialist at evaluagent, to explain how AI and conversation intelligence are changing what’s possible with customer preferences and personalization.
Video: Customer Preferences: Stop Building Personalization Just Around What Customers Say
Watch the video below to hear Chris explores what’s now possible with customer preferences and how contact centres need to stop building personalization just around what customers say:
With thanks to Chris Mounce, Product Training & Enablement Specialist at evaluagent, for contributing to this video
This video was originally published in our article ‘12 Amazing Things You Can Now Do With Customer Preferences’
How to Use AI to Transform Customer Preferences and Personalization
AI now makes it possible to analyse every customer interaction, as Chris explains:
“AI makes it possible to analyse every conversation systematically at a scale that simply wasn’t feasible before. And that changes what’s possible with customer preferences significantly.”
Instead of relying on surveys or static customer records, contact centres can now uncover preference signals directly from customer behaviour and conversations.
To help contact centres rethink personalization, here are three major shifts they should focus on:
1. Shift From Asking Customers to Observing Their Behaviour
Many contact centres still rely heavily on surveys and forms to understand customer preferences.
But customers are already demonstrating their preferences every day through their actions, including:
- Channel choice
- Contact frequency
- Self-service abandonment
- Escalation behaviour
- The way they phrase requests
This means contact centres no longer need to rely solely on what customers explicitly state. Instead, they can identify behavioural signals across every interaction.
“The first move is shifting from asking to observing. Channel choice, contact frequency, self-service abandonment, what customers escalate on, how they phrase the requests – all of that is preference data.
Customers are demonstrating what they want consistently, without being prompted. What used to require sampling a fraction of interactions can now cover everything.
And the organizations getting ahead of this aren’t collecting more preference data, they’re finally reading the preference data they’ve always had.”
What once required reviewing a small sample of conversations can now be done across every channel and every interaction.
The contact centres getting ahead are not necessarily collecting more customer preference data, they are finally making better use of the data they already have.
2. Understand That Stated Preferences and Behaviour Are Not Always the Same
One of the biggest challenges with traditional preference capture is that what customers say they prefer does not always align with how they actually behave, Chris continued:
“The second shift is understanding that stated preferences and demonstrated preferences, they aren’t always the same thing.
A customer might indicate they prefer digital channels, but then consistently call when something actually needs resolving.
Like a customer who calls back the next day on a supposedly resolved issue, they’re telling you something that a preference form never would.
What’s changed is the ability to surface those signals across every interaction, not just a sample.
And behavioural signals, they don’t lie. Stated preferences sometimes do. Personalization built on what customers say rather than what they do, that risks optimizing for the wrong thing entirely.”
These behaviours reveal far more about customer expectations than a survey response ever could.
As a result, AI and conversation intelligence now make it possible to surface these behavioural signals at scale across every interaction, not just a small sample.
And behavioural signals matter because they are often more reliable than stated preferences, as personalization built only around what customers say risks optimizing the wrong experience entirely.
3. Use Preference Insights to Shape the Entire Interaction
Most contact centres currently use customer preferences primarily for routing purposes, such as selecting a channel, assigning a queue, or prioritizing service tiers.
While useful, this only scratches the surface of what’s possible.
The bigger opportunity is using preference insights to influence how interactions themselves are handled, which could include adapting:
- Tone and communication style
- The level of detail or guidance provided
- The pace of resolution
- The amount of reassurance a customer may need
AI now makes it possible to surface these signals fast enough to influence the next customer interaction in real time, rather than simply reporting on them afterwards.
“And the third shift, that’s about where that insight ends up. Most preference data gets used for routing, which channel, which queue, which tier. Now, that’s useful, but it’s only part of the picture.
The bigger opportunity is using preference insight to change how the interaction itself is handled. The tone, the level of guidance, whether a customer needs a quick resolution or reassurance, AI makes it possible to surface those signals fast enough to be useful before the next conversation, not just in a post-call report.”
That transforms personalization from a static profile into a dynamic customer experience capability.
From Static Preferences to Behaviour-Driven Personalization
Customer preferences are no longer fixed data points stored in a CRM system.
With AI and conversation intelligence, contact centres can continuously observe how customers behave, understand what those behaviours actually mean, and adapt experiences accordingly.
The future of personalization is no longer built purely around what customers say they want but around what their behaviour consistently reveals they need.
Author: Robyn Coppell
Reviewed by: Jo Robinson
Published On: 17th Jul 2026
Read more about - Video, Artificial Intelligence (AI), Chris Mounce, Customer Experience (CX), evaluagent, Personalization, Service Strategy, Technology Enablement Strategy, Videos
