Retailers are under immense pressure to deliver seamless shopping experiences during peak events like Black Friday and Cyber Monday. Yet, many still underestimate how fragile site performance can be under heavy traffic.
Shawn O’Neill, VP of Performance Consulting at Yottaa, explains to Digital Journal why slow load times can erode conversions and revenue during peak periods and how third-party tools can silently degrade performance under traffic surges.
Digital Journal: What are the most common performance failures retailers experience during high-traffic events like Black Friday and Cyber Monday?
Shawn O’Neill: The most damaging performance failures aren’t always technical breakdowns – they’re rooted in false confidence and outdated assumptions. Many retailers head into peak shopping season assuming that they’ve already optimized for speed, but performance isn’t “set-it-and-forget-it”. It’s constantly shifting with every new campaign, vendor, update, or script addition. Another common misstep is assuming third-party tools, such as personalization engines, will remain stable and predictable during traffic surges. But these tools are maintained by external vendors and can be updated without warning, leading to a silently degrading site performance and more fragile storefront.
Additionally, many teams believe they’ll catch issues as they happen, but by the time a consumer flags a bug or an alerting system flags it, the damage is already done. Perhaps most critically, some retailers believe performance to be a purely technical issue, when in reality, it’s also a strategic business issue tied directly to conversion and revenue. If left unchallenged, these assumptions can lead to costly losses when traffic is highest.
DJ: How can issues like slow load times impact conversion rates and customer experience during these peak shopping times?
O’Neill: The impact of slow site speed during the holiday season is both immediate and measurable. According to Yotta’s 2025 Web Performance Index, 63% of shoppers will abandon a page that takes longer than 4 seconds to load. Reducing the page load time by just one second can increase mobile conversions by 3%. When pages lag or fail to load quickly, retailers risk having customers leave before engaging with the brand at all. When you take into consideration that 70% of pageviews happen on product or category pages rather than the homepage, it makes sense that even minor friction in those entry points can significantly impact revenue. Site speed directly influences conversion, bounce rates, search visibility, and even customer lifetime value. During high-traffic periods, every second counts.
DJ: What infrastructure strategies do you recommend for scaling efficiently without compromising performance?
O’Neill: Under heavy load, bottlenecks typically spring up on the back-end: shopping carts, authentication services, inventory synchronization, and order processing. Load testing these critical failure points well in advance of major traffic events can help teams remediate potential performance challenges, overselling, or worse, outages. Aim to minimize database access for the pages higher in the shopping funnel: e.g. landing pages, blog posts, categories, and similar relatively static content. As the user journey moves deeper in the checkout funnel, the need for dynamic data increases.
As an example: a previous client was writing a new cart entry in their database for every visitor that landed on their site, regardless of whether the user added any items to their cart. This resulted in massive load to the back-end application servers and databases, all of which was avoided by instead only creating cart objects when an item was added by the user.
DJ: How do third-party plugins and technologies – such as personalization engines or analytics tools – impact site speed during high-traffic events?
O’Neill: Third-party technologies are a significant, yet often overlooked, source of performance drag. In fact, our data shows these tools account for 44% of total page load times. As digital storefronts grow in complexity, so does the technical stack behind them, and each component – while essential for personalization, analytics, and customer experience – competes with the primary purpose a given page exists in the first place: to meet the customer’s expectation.
Even when vetted and QA-tested, third-party scripts can change without notice, leading to potential issues or outright failures during surges in traffic. They can block important content, decay speed, or even crash entirely. To avoid this, it is critical to actively orchestrate these tools, continuously monitor their performance, and establish fallback protocols to ensure a seamless shopper experience when it matters most.
Ideally, brands have run their experiments in the off-season, and can put their best UX-foot forward on the big day. Deactivating A/B testing, heat-mapping, and other CPU-intensive data collection applications under peak load can ensure your customers aren’t paying a performance penalty when it matters most.
DJ: Where does Yottaa see retailers missing the mark most often when it comes to digital performance planning? What should they be doing instead?
O’Neill: Time and time again, retailers overestimate their level of readiness going into the holiday season. One of the most common pitfalls is the idea that performance is a one-time-project, rather than a program that requires continuous attention and adaptation. Instead of assuming their stack will hold up under pressure, the most effective retailers are closely monitoring user behaviour and remediating performance bottlenecks long before a problem arises. They also establish a clear connection between performance metrics and revenue outcomes, tying it back to larger business goals for executive team visibility and accountability.
Real readiness isn’t just about having a fast site today. It’s about having a site that stays fast tomorrow and every day after regardless of any traffic, content, or technology that comes its way. This requires continuous optimization, orchestration across the tech stack, and a proactive approach to identifying and resolving friction before it costs revenue.
