How ASOS uses AI to personalize fashion for 20 million customers
When you're matching Brad Pitt's Fight Club jacket to millions of shoppers worldwide, personalization gets complicated fast. We sat down with Shaghig Babikian, CRM Lead at ASOS, and Colleen Kerr, Lead Product Manager at Braze, to see how they're using Simon Data and AI to deliver the right product to the right customer at exactly the right moment.
Note: This interview has been edited and paraphrased for clarity and length.
ASOS started as "As Seen On Screen," where it literally sold celebrity outfits from movies and TV. Now you're approaching 3 billion pounds in revenue. How has personalization evolved?
Shaghig Babikian: Our goal is simple: be the number one fashion destination by presenting the right product at the right time. But when you're dealing with 20 million customers across 850 brands, “simple” becomes complicated fast.
Customers expect personalization everywhere now. The campaigns that work are focused, product-led, and hit customers exactly where they are in their journey.
Take our abandon suite. If someone browses jeans but doesn't buy, we don't send a generic "come back" email. We follow up with personalized messages based on exactly what they viewed, tailored by product type, and what's actually available.
We also have trend campaigns. When we want to push denim, we dig into browse behavior and purchase history to find people actively hunting for jeans. Then we build curated edits with the right price points, brands, and styles for each segment. Simon and Braze let us trigger all of this in real time across millions of customers so that this kind of relevance builds trust.
What's the data foundation behind these experiences?
Shaghig: Three key areas: First, Simon gives us a unified customer view with behavioral, transactional, and demographic data in one customer profile. This enables us to precisely target everyone, from first-time shoppers to high-value customers.
Second, we can activate these audiences across both CRM and paid channels for targeting and performance measurement. Third, we have direct integration between Simon and our web and app platforms, enabling real-time onsite personalization like loyalty messaging, targeted incentives, and exclusive features without requiring heavy product development.
Colleen, how does Braze make data activation simple for ASOS?
Colleen Kerr: We focus on reducing friction. There are two patterns: profile-level activation, where Simon data syncs directly via APIs, and campaign-level activation, where CDP logic triggers Braze messages. Since Simon runs on Snowflake, both methods work seamlessly.
I'm curious about the more complex stuff you're building. How does your architecture handle those really intricate segments?
Shaghig: Good question. Our composable setup is genuinely collaboration-friendly. The SQL interface means we can build complex segments, such as, "High-value, recently lapsed Premier customers who previously shopped Topshop Petite and browsed back-in-stock items last week" — now say that 10 times fast! — and do it quickly.
Once they are built in Simon, they pass straight into Braze, where we have pre-configured campaigns ready to activate. The biggest benefit is the combo of speed and precision.
Now this is where it gets really interesting…let's talk AI. I know you're doing some fascinating work with weather data.
Shaghig: This is so exciting. We’re finally getting to see real ways to use AI. We're building our first AI-generated segment using real-time weather data. The model looks at a seven-day global forecast and links customers to local conditions via geolocation.
When AI flags extreme weather like heavy rain, heatwaves, or early snowfall, for example, it dynamically segments customers in affected areas. The matching campaign triggers in Braze, promoting rainwear, sunglasses, or knitwear with content tailored to local stock and expected temperature.
Swimwear edits when it hits 25 degrees, and rainwear before storms. The best part is it all happens dynamically.
That's incredible — you're literally predicting what customers need based on the weather. Colleen, once these AI-powered campaigns launch, how does Braze optimize them?
Colleen: Our Catalyst product uses reinforcement learning. If you already know something about a customer, say, regionalized fashion trends, Braze starts there and refines targeting and messaging. Campaigns trigger journeys that auto-optimize, learning what works best for each individual.
This reinforcement learning approach is fascinating. Shaghig, I'm curious: what's on the horizon for AI at ASOS? Where do you see the tech taking you next?
Shaghig: The $64 million dollar question… but actually, we’re looking at micro campaigns that promote hundreds of different style trends daily, matched to the right customer through the right channel at the right time. Fashion moves fast. With AI, we'll automatically identify who's likely to engage with balletcore, oversized tailoring, or Y2K denim, and serve personal, timely content.
We're also looking at cultural moments. Think festival season, concerts, sports events. These have distinct style codes, and with the right signals, we can target the right audience with the right inspiration exactly when they're looking for it.
This is hyper-relevant, data-driven storytelling that reacts in real time to what's trending, what's in stock, and what each customer wants to see.
Before we wrap up, I always like to ask this. If there's one piece of advice for brands looking to scale personalization like ASOS, what would it be?
Colleen: You know, data's value is expanding way beyond just analytics. Braze has dozens of petabytes of customer engagement data, and layering AI and reinforcement learning on top unlocks incredible potential.
Shaghig: I'd say combine great data, the right tech, and focus on relevance. But make it granular and meaningful. The more tailored and timely, the bigger the impact. It's not AI for AI's sake — it's AI that makes CRM smarter, faster, and more human in connecting with customers.
Ready to scale personalization with AI? Discover Simon AI and see how leading brands turn data into revenue.
