On most days, traffic on our WordPress landing pages follows a familiar rhythm. Users arrive steadily, browse articles, run searches, read medai releases, and move on. Our systems are designed for this kind of gradual, predictable growth.
But every so often, that rhythm changes.
A campaign launches, a major announcement goes live, or a time-bound event like Secondhand Day kicks off and suddenly thousands of users arrive at the same time, often with a clear goal and a sense of urgency. These are what we call high traffic moments. They’re short, intense, and technically demanding: and they’re exactly when platform reliability matters most.
So what actually happens during these peaks, how do we prepare for them, and how do we make sure the experience stays fast and stable even under pressure?
What we mean by a “high traffic moment”
A high traffic moment is a short period when far more users than usual arrive simultaneously, typically driven by marketing campaigns, seasonal events, or major launches.
On a normal day, traffic grows gradually and requests are spread out over time. During a peak, many users hit the same pages and perform the same actions at once: opening a campaign landing page, running searches, applying filters, or submitting forms.
That sudden concurrency is the key challenge. Even well-performing systems can slow down if servers, databases, caches, or third-party services reach their limits at the same time. Without preparation, this can lead to longer load times, timeouts, or error spikes, exactly when user expectations are highest.
Predicting peaks before they happen
The good news is that most high traffic moments are not a surprise.
We predict them by combining historical data, seasonality, and planned activities across teams. Past events are particularly valuable: if we’ve run the same campaign before, or observed similar behaviour around the same date last year, we can see how big the spike was, when it started, and which parts of the platform were under the most pressure.
Seasonality also plays a big role. Traffic tends to follow predictable patterns: weekdays versus weekends, holidays, end-of-month effects, back-to-school periods, or major retail moments like Black Friday. Events such as Secondhand Day have already created traffic spikes in the past, so we treat them as known peaks and plan for them early.
For new campaigns, we estimate impact based on expected reach: newsletter size, ad spend, social media exposure, and the structure of the landing page flow on our WordPress instances. If marketing has a fixed launch time and a large audience touchpoint, we assume a spike will happen and prepare capacity and monitoring in advance. This planning usually involves close collaboration between engineering, marketing, and product teams.
How we prepare the platform ahead of time
Before a high-traffic event, we run an additional checklist on top of our daily monitoring routines to make sure the platform stays fast and stable.
First, we estimate the expected load and identify critical flows. We confirm which pages and features will be hit the hardest and where potential bottlenecks might appear. These are typically landing pages, search, login, or forms.
Next, we scale server capacity ahead of time if needed. That can mean increasing CPU or RAM and reviewing key limits such as web server settings or PHP worker pools to handle higher concurrency. Caching and content delivery are also crucial. We make sure that “hot” pages benefit from caching wherever possible and that static assets like images, JavaScript, and CSS are served via a Content Delivery Network (CDN) to protect the backend systems.
Finally, we prepare tight monitoring and alerting for the event itself, with clear thresholds and owners on standby. The goal is to add headroom, validate assumptions with load tests, and be ready to react fast.
What we monitor during the peak
Once the peak is live, attention shifts fully to real-time signals.
From a user perspective, we closely watch page load times on key flows such as landing pages, search, and form submissions. From a system perspective, we monitor error rates, especially significant spikes in responses, timeouts, or failed requests.
On the infrastructure side, we track CPU, RAM, disk, bandwidth, and worker saturation, as well as database health indicators like slow queries, response times, and connection pressure. Cache hit rates and the performance of third-party services are also critical, since external dependencies can become bottlenecks under load.
Collaboration is just as important as metrics. During high traffic moments, all relevant stakeholders stay in a shared communication channel. Teams coordinate across tracking, ads, product, and engineering, and if intervention is needed, developers can apply the fastest safe fix while keeping everyone informed in real time.
When things don’t go as planned
Even with careful preparation, not everything always goes perfectly. That’s why we follow a simple incident approach designed to protect the user experience first.
The first step is to confirm the impact quickly by checking dashboards and logs to see what’s failing and which user flows are affected. Then we identify the bottleneck: whether it’s server capacity, worker limits, database pressure, cache misses, or a third-party dependency.
From there, we apply the quickest safe mitigation. This might mean scaling up resources, adjusting caching or CDN settings to reduce load on the origin, or temporarily disabling heavy or non-essential features.
If search and filtering start to slow the platform down, we may temporarily reduce their impact by limiting complex filter combinations, lowering the number of results per page, disabling extra sorting options, or applying short rate limits to repeated requests. When a recent deployment or configuration change is the likely cause, we roll it back.
Once the system stabilises, we continue monitoring to validate recovery and capture key data for follow-up fixes and improvements. The principle is simple: keep the core experience — browsing, searching, submitting — fast and reliable, even if some non-essential features are temporarily reduced until the peak passes.
Built for the moments that matter
High traffic moments are a sign that our landing pages are doing something right! Handling them well requires preparation, cross-team collaboration, and the ability to react quickly under pressure.
By planning ahead, monitoring the right signals, and focusing relentlessly on the user experience, we make sure our pages stay fast and stable when it matters most.
Author
Milos Bogdanovic
Fullstack Web Engineer
Group Marketing