Peak performance: 7 decisions enterprises can make before the next big commerce moment

The decisions behind peak performance

Every year, brands and retailers face traffic surges around major events, from Black Friday and Cyber Monday to Valentine’s Day. Major promotions or product launches may trigger similar conditions that, even with all the prep before the event happens, remain challenging to manage.

This is because handling increased volumes is only one aspect of making peak seasons a success. It’s also about managing inventory that moves across channels at near-light speed, campaigns that change quickly to accommodate demand or customer feedback, and even an empowered workforce managing every dependency in the technology stack.

The enterprises that perform under pressure have made a series of decisions long before the peak arrives. They have designed their technology and operating models so that they can scale the parts of the business that need it, test changes safely, reuse what already works, make product data AI-ready, and give teams the freedom to respond when customer demand changes.

Looking at how companies, including L.L.Bean, Chemist Warehouse, Express, Flügger, Moonpig and SPORT 24, approach commerce modernization reveals seven practical moves enterprises can make to become more peak-ready.

1. Scale selectively

When demand spikes, it can be tempting to treat the commerce platform as a single, large system. But modern commerce long ago changed that monolithic paradigm with modular systems operating as a collection of services, integrations, customizations and experiences, each with different workloads and different scaling requirements.

That’s why L.L.Bean’s decision to scale only the parts that need scaling — and not everything around them — is clever. The American outdoor apparel retailer built an architecture using more than 50 microservices and more than 400 cloud functions. Its customizations were stored in its own cloud environment, allowing them to scale independently rather than tying every piece of custom functionality to the commerce platform itself.

That distinction matters during peak. If one part of the customer journey suddenly experiences a surge in demand, an architecture that allows individual components to scale independently can respond more precisely than a monolithic system in which everything must scale together.

This approach was put to the test. During L.L.Bean’s first holiday season fully on commercetools, the company experienced zero unplanned outages, with a fast order during its Cyber Monday peak.

The takeaway: Don't make the entire platform scale because one component is under pressure. Identify your critical workloads and give them room to scale independently.

2. Make commerce machine-ready

Whether supporting AI-assisted shopping or full-blown agentic commerce — where AI agents can execute actions, such as purchases, on behalf of shoppers — brands and retailers need to structure product data so AI agents can understand products, pricing, availability, promotions and fulfillment, and interact with those capabilities reliably.

Retailers like JD Sports, Liverpool and Vision Healthcare are among the first companies to invest in AI readiness, structuring product data and making these API accessible.

The takeaway: Build your commerce foundation so AI can discover, understand and eventually act on your products and capabilities.

3. Make capacity elastic

Not every peak can be predicted. Spikes may come from a successful campaign, a viral product drop, a major sporting event or even an influencer’s recommendation. Retailers that need to provision capacity in advance to handle high volumes find it difficult to deal with such dynamic conditions.

What usually happens is that you either end up provisioning for maximum possible demand — and paying for capacity you rarely use — or provision too conservatively and risk running out of headroom precisely when demand is highest.

Moonpig encountered this challenge as its business grew. Its on-premise infrastructure was becoming increasingly difficult to scale for major occasions such as Christmas and Mother’s Day, two peak moments when demand can increase dramatically.

Moving toward a cloud-native, serverless architecture gave the business a different model. Instead of relying on infrastructure sized manually for peak periods, capacity could scale automatically with demand.

The takeaway: Demand is rarely 100% predictable, even more so in the era of agentic commerce. Design capacity to respond automatically rather than making your business permanently pay for its big day(s).

4. Reuse what works relentlessly

Every peak season brings new campaigns, promotions, product ranges, markets and/or customer experiences. But that doesn’t mean the underlying technology should be new every time. One of the biggest advantages an enterprise can build over time is a library of modular capabilities that have already been designed, tested and proven, and that can be adapted for different needs.

Flügger offers a compelling example of how to build reusable capabilities instead of rebuilding experiences from scratch. The Nordic paint manufacturer and distributor has built reusable components that can be applied across brands and channels, including the point-of-sale (POS) system. Instead of treating each new digital experience as an entirely separate technology project, teams can draw on existing capabilities.

That has a direct implication for peak seasons. When a Black Friday-like event arrives, the technology team shouldn’t have to reinvent the machinery that supports the promotion. The reusable pieces are already there. Teams can concentrate on the commercial proposition: What to sell, how to merchandise it and how to create the experience customers actually want.

Reuse can also change the economics of experimentation. Once a capability exists as a reusable building block, the cost and effort of deploying it somewhere else can fall significantly.

The takeaway: Turn successful commerce capabilities into reusable building blocks, so every new peak becomes an opportunity to assemble and improve rather than rebuild.

5. Design for omnichannel interactions

For customers, there is no such thing as a “digital customer journey” and a separate “physical customer journey.” They search online, check availability, visit a store, compare prices on their phone, order online and collect in person. The journey may be completely different from customer to customer.

That’s why it’s become crucial to support those omnichannel interactions with a unified commerce approach that consolidates all data in one place.

Chemist Warehouse, Australia’s largest discount pharmacy chain, illustrates this well. With hundreds of physical stores, its digital experience plays a role within a much larger retail ecosystem, helping customers discover products, check prices and understand availability before deciding how and where to purchase.

For instance, the company connected its commerce environment with real-time inventory data across its store network, enabling experiences such as omnichannel fulfillment. Approximately 60% of Chemist Warehouse’s online sales now come through Click & Collect.

That kind of model turns the physical network into part of the digital proposition and changes how retailers should think about peak readiness. With real-time access to data, a surge in online demand doesn’t necessarily have to mean more pressure on a central warehouse. If inventory, stores and fulfillment options are connected, the business can use the assets it already has to respond to demand.

The takeaway: Treat stores, inventory, fulfillment and digital experiences as a unified commerce system, and give customers flexibility over how they move through it.

6. Test aggressively

Peak traffic isn’t the right time to find out whether your architecture works. Yet many retailers effectively treat their biggest shopping moments as the ultimate performance test, pushing systems to their limits and discovering what breaks in production.

Peak-ready businesses reverse that equation, treating high-volume traffic as something you rehearse and test for.

Once again, Chemist Warehouse took a winning strategy here with an aggressive approach to performance testing before launching its new commerce environment. The company ran a spike test designed to push the platform beyond expected demand. The test was so demanding that it broke the retailer’s internal performance-testing engine, while the commercetools platform continued to handle the traffic.

The point, however, is to identify the real breaking points before customers do. So, testing should extend beyond infrastructure, like new experiences, checkout flows, promotions and changes to the customer journey before exposing them to everyone.

The fashion retailer Express ran tests requiring the system to process 100,000 orders per minute, equivalent to 1,667 orders per second, to mimic a potential traffic spike. Even as the commercetools platform passed the test, the point is to consistently test performance to catch potential issues that may affect the customer experience.

The takeaway: Reduce the amount of the unknowns before peak shopping moments arrive. Test beyond expected demand, validate changes incrementally and use controlled experiments to find problems before customers do.

7. Empower teams

Peak moments are full of decisions: Change a promotion, update content, adjust merchandising, test an experience, react to inventory or respond to customer behavior. If every change requires a lengthy development cycle or a ticket to another team, the business can become slower precisely when it needs to move faster.

Chemist Warehouse designed its new environment to give commercial teams greater autonomy over content, promotions and customer experiences, reducing their dependence on lengthy technology cycles.

Similarly, Flügger enables commercial teams to experiment and introduce new experiences using reusable components. The Danish sports retailer SPORT 24 has used its new commerce environment to enable experimentation and A/B testing, allowing the business to learn what works rather than relying entirely on assumptions.

Peak agility is so much more than having infrastructure that can absorb a traffic spike. Brands and retailers should also be able to respond to what customers are doing during that spike.

The takeaway: If the technology is flexible but the operating model isn’t, you’ve only solved half the problem. Put the ability to change in the hands of the people who understand the customer, and make experimentation part of the operating model.

Peak performance is built before peak arrives

Enterprise brands and retailers that succeed during peak seasons are those that have prepped their architecture well in advance and acknowledge that peak performance is a requirement that extends beyond Black Friday/Cyber Monday.

Most importantly, it demonstrates that peak readiness isn’t only about leveraging cloud-native infrastructure and calling it a day, but a series of choices about how the business wants to operate when demand is at its highest. The result is a commerce operation designed to perform under pressure and to keep improving once the pressure is gone.

Ready to see how commercetools can support your next high-traffic event? Contact our sales team or start our 60-day free trial to learn how your business can scale during BFCM and other important sales events.

FAQs

Peak-ready eCommerce is the ability to maintain performance, availability, and customer experience when demand increases suddenly. It requires more than additional infrastructure: Enterprises need flexible architecture, elastic capacity, rigorous testing, connected data and teams that can respond quickly when conditions change.

Enterprises can prepare by scaling critical workloads independently, making capacity elastic, testing beyond expected demand and building reusable commerce capabilities. Preparing teams and operating processes to respond quickly is equally important.

Not every traffic spike can be predicted. Campaigns, viral products, sporting events, influencers or unexpected changes in demand can create sudden increases in traffic. Elastic capacity allows commerce infrastructure to respond automatically rather than requiring enterprises to provision for maximum demand in advance.

Enterprises can make commerce AI-ready by structuring product data and commerce capabilities so AI systems can discover, understand, and act on them. This includes making information such as product details, pricing, availability, promotions and fulfillment accessible through APIs.

Testing helps enterprises identify performance and customer experience issues before they affect shoppers. Peak-ready businesses test beyond expected demand and validate critical experiences such as checkout, promotions, and changes to the customer journey before exposing them to everyone.

Modular commerce enables enterprises to scale and manage individual services, integrations and experiences to match their specific workloads. Instead of scaling an entire platform because one component is under pressure, businesses can give critical workloads room to scale independently.