Video series: The state of personalization for 2026 and beyond

We all know what personalization is, right?

Yea, we understand the core concept: your website or app changes as users interact with it in order to provide a more relevant experience.

It's both as simple as that... and way more complicated.

Two years ago, Optimizely rolled out Optimizely Personalization. It’s designed to meet the needs of customers who want to scale their personalization programs fast. With Optimizely’s best-in-class personalization engine, customers can provide the 1:1 experiences that their customers have come to expect.

As it turns out, launching a personalization strategy is no small feat. There’s tons of work that needs to occur behind the scenes to ensure your customers aren't just seeing the right content, but the right content at the right time. And more and more, that work isn't done by stacking more rules or hiring more people. It's done by letting AI handle the complexity that human segmentation was never really built for.

I was lucky enough to sit down with Nicola Ayan, VP of Solution Strategy to talk about all things personalization, where it's at, and where it's heading in 2025.

Where do you get started with personalization when no one knows where to get started? That's what I wanted to know, and that's what many marketers want to know. Typically, you want to start with either your highest performing pages or most reliable data sets.

And if you're not sure what those are, start with a problem you want to solve; for example, a bottleneck on your checkout page.

Do companies need a ton of data or a robust CDP to execute personalization?

Sure, more data usually leads to better results, but don't let a minimal amount of data stop you from launching your personalization strategy. If you're not working with mountains of data, start with one or two segments or one problem and go from there.

Do companies need a ton of content to personalize?

Much like data, you can never have too much, but personalization is all about creating impact and influencing customer behavior. How many pieces of content are required to do that will depend on the organization and the goals.

This should come as no shock to anyone, but fragmented customer data is a massive problem for companies of mid-to-high maturity. When you have data stored in multiple systems like websites, data warehouses, social media, physical stores, wrangling that all together to build a customer profile can be daunting.

So where is personalization going?

The two things shaping personalization in 2026 are unified context and governance. Making sure data flows between every system so AI is working from the full picture, not just a slice of it. And making sure that as programs scale, the experiences staying on-brand and consistent.

Speaking of RAG and AI, how does AI influence the personalization process?

Rules-based personalization works great until it doesn't. "If a user buys a washer, show them a dryer" makes perfect sense until you have 50,000 products and a million different customer journeys. Rules cover the scenarios your team thought of. Everything else falls through the cracks.

That's where AI picks up. It reads the signals no rule was ever going to catch — which products someone pauses on, how their behavior shifts on a return visit, what they actually need versus what their segment suggests. The strategy stays with you. AI handles the volume.

Do companies need a ton of traffic to their site? You might be suprised to learn that you don't need a ton of traffic. Unlike experimentation, which relies on statistical significance, personalization is all about influence customer behavior and measuring outcomes, which doesn't require a predetermined amount of traffic.

With Optimizely's acquisition of Netspring, who specializes in data warehouse native analytics, how does that play into personalization?

It's actually a surprisingly simple concept; A lot of data that you need for experimentation and personalization simply doesn't live on a website or within easily accessible data platforms. Sometimes you need data that's buried within a data warehouse, and that's where the magic happens...

Lastly, we spoke about how customer data platforms (CDP)s tie into personalization. This is done in three key ways:

Personalization has been around forever and in many ways, it hasn't really changed. It's always been about making the customer feel special, unique, and prioritized. What has changed, and will continue to change, is how fast we can execute it, how accurate we can be, and the degree to which we can influence customer behavior.