Noibu blog

The silent killers of ecommerce performance: How Hanna Andersson modernizes without breaking the business

Matt Ezyk of Hanna Andersson

Get the full episode here

TL;DR — What you'll learn:

  • Why Hanna Andersson modernizes in pieces instead of risky all-at-once rebuilds
  • How composable commerce only works when you have the internal team to run it
  • Why AI without clean, unified data is a budget sink — and how to get “AI ready”
  • How embedding engineering into marketing protects site speed and campaign performance
  • Why optimizing product data now is critical for the AI-driven search era
  • Matt Ezyk’s 5-step playbook for modernizing ecommerce without breaking the business
“Composable isn’t a magic word. It only works if you have the internal team to run it after launch.”
Matt Ezyk, Senior Director of Engineering, Ecommerce at Hanna Andersson

In this episode of Ecommerce Toolbox: Expert Perspectives, host Kailin Noivo talks with Matt Ezyk about how a 40-year-old children’s clothing brand is modernizing its ecommerce stack without high-risk relaunches, vendor lock-in, or performance dips.

🎧 Listen to the full conversation on Apple, Spotify, or YouTube

Performance-first modernization

When Matt joined Hanna Andersson, the site was running on an older Salesforce Commerce Cloud implementation, burdened by years of patches and slowdowns.

The “standard” move would have been a full replatform. Matt chose incremental modernization instead:

  • Start with checkout performance for immediate ROI
  • Roll out new components one at a time
  • Avoid taking critical revenue-driving flows offline

Customers feel improvements sooner, and risk stays low.

“Every change needs to drive value without putting the customer experience at risk.”
Matt Ezyk, Senior Director of Engineering, Ecommerce

Headless and composable—only when you're ready

Headless commerce is powerful, but fragile without the right team.

  • Build internal capability first — don’t rely solely on vendors
  • Use API-first microservices for flexibility
  • Own your roadmap, avoid vendor dependency

Without in-house skill, composable setups can become expensive bottlenecks.

AI starts with clean data

Matt’s take: AI without foundational data hygiene is a budget sink.

AI readiness checklist:

  • Clean, unified, accessible customer data
  • Stable performance baseline
  • Clear, specific business problems to solve
“Without clean data, AI just automates bad decisions.”
Matt Ezyk, Senior Director of Engineering, Ecommerce

Marketing + engineering = speed and stability

One overlooked performance killer? Misalignment between marketing and engineering.

Matt’s fix: embed engineers into marketing meetings so teams can:

  • Spot performance risks before campaigns launch
  • Align technical upgrades with business goals
  • Maintain site speed during marketing pushes

Digital experience = trust and consistency

For Matt, site performance is only part of the equation. True digital experience excellence comes from delivering a reliable, friction-free journey — every time.

“To me, DXA is about trust and consistency. So like a quote, unquote clean site doesn't mean it's fast. It means that there's no broken flows. There's no unexpected friction, and it's just a reliable experience every time.”
Matt Ezyk, Senior Director of Engineering, Ecommerce

That means weekly “store walkthroughs” from his QA team, regression testing after every change, and an operational discipline that catches small issues before they become customer frustrations.

Preparing for AI-driven search

Keyword-based search is giving way to AI-assisted discovery.

Matt’s recommendations:

  • Invest in first-party data
  • Improve identity resolution
  • Enrich product data for AI-friendly search

Brands that act now will own discoverability in the next era of ecommerce.

Hanna Andersson 6-step modernization playbook
1
Modernize in pieces — start where revenue impact is highest
2
Build your team before going composable
3
Put data before AI
4
Embed engineering into marketing conversations
5
Optimize product data for AI-driven search
6
Everything we do is rooted in site performance
“When performance is part of your culture, those silent killers of ecommerce don’t stand a chance.”
Matt Ezyk, Senior Director of Engineering, Ecommerce

Listen to the full episode now

Frequently Asked Questions

Why doesn’t Hanna Andersson do a full replatform?
Large-scale replatforms are risky for revenue and customer experience. Instead, Hanna Andersson modernizes in smaller, high-impact increments — starting with critical areas like checkout — so customers see improvements sooner and downtime is minimized.
What does “composable-ready” mean?
Composable commerce allows brands to choose best-in-class tools for each function, but it only works if the internal engineering team can manage and evolve the system post-launch. Being composable-ready means having that team in place before making the shift.
Why is clean data so important for AI in ecommerce?
Without clean, unified customer data, AI tools amplify bad inputs and produce poor recommendations. Matt Ezyk’s team prioritizes accurate, connected data before layering AI into the stack to ensure insights are reliable and actionable.
How does Hanna Andersson align marketing and engineering?
Engineering leadership joins marketing conversations proactively, not just when there’s an issue. This collaboration prevents campaigns from hurting site performance and ensures tech changes support business goals.
What’s Hanna Andersson doing to prepare for AI-driven search?
The brand is enriching product and content data to make it AI-search friendly, investing in first-party data, and improving identity resolution so products are easily discoverable through natural language queries in AI tools.
Back to all

Identify the top 10 errors and friction points impacting conversion and revenue
Free checkout audit
Share

Don’t lose customers to site errors—protect your revenue with Noibu