A 100x traffic spike. A live national broadcast. Three years. Zero downtime.
When the Mark Hughes Foundation runs its annual Beanies for Brain Cancer campaign, millions of Australians tune in across national television. Donations flood in. Social media lights up. And in the background, our team at Clickk is watching every metric in real time to make sure the website holds.
This is the story of how we do it, and why it matters.
The Mark Hughes Foundation and Beanies for Brain Cancer
The Mark Hughes Foundation exists to fund life-changing brain cancer research across Australia. Each year, its signature Beanies for Brain Cancer campaign brings the nation together, broadcast on national television, amplified across social media, and anchored by a website that becomes the central hub for donations, merchandise, and awareness.
The Foundation’s team works incredibly hard to make this campaign what it is. The supporters, volunteers, and community who show up year after year are the reason it succeeds. This year, they raised $3.8 million against a $3 million target, and every dollar of that reflects their effort and the generosity of Australians who care about this cause.
Clickk has managed the Mark Hughes Foundation website for three years. Our job is a simple one: make sure the website is the last thing anyone has to worry about. The Foundation’s team should be focused entirely on the campaign, not on whether the site is going to hold up.
The Technical Stack
The Mark Hughes Foundation website runs on WordPress with Elementor, hosted on Kinsta, with Cloudflare as the content delivery network (CDN). It is a stack we know deeply at Clickk, and one we have optimised specifically for high-traffic event scenarios.
Each layer has a job:
- WordPress and Elementor: the content management and page building layer, kept lean and fully updated ahead of the campaign
- Kinsta: managed WordPress hosting with serious infrastructure behind it, including PHP thread management and server-level caching
- Cloudflare: the CDN layer that sits in front of everything, absorbing the bulk of traffic before it ever touches the origin server
During this year’s peak broadcast period, Cloudflare served approximately 37GB of traffic directly. The origin server handled around 370GB in total across the campaign. That ratio tells you everything: Cloudflare was doing the heavy lifting, protecting the server from the full weight of the spike.
What a 100x Traffic Spike Actually Looks Like
On a normal week, the Mark Hughes Foundation website sees around 50 visits every 30 minutes. During a televised Beanies for Brain Cancer broadcast, that number jumps to approximately 5,000 visits in a 30-minute window.
That is a 100x spike. It arrives fast, it is unpredictable in its exact timing, and it has to be absorbed without a single user hitting an error page.
This year’s campaign also behaved differently to previous years. Rather than one sharp spike, traffic was higher across all days, driven by a broader mix of television and social media activity. The infrastructure had to perform not just for a single peak moment, but across an extended period.
Video: MHF website load speed during a televised campaign with 5,000+ active visitors
How We Prepare: A Month Out
Preparation for Beanies for Brain Cancer starts approximately one month before the campaign goes live. It is methodical, not reactive.
Step 1: Page weight optimisation
In the first year we managed the site, we started by optimising the home page and calculating exactly how many megabytes it takes to serve a single page load to a single user. That number becomes the foundation of every infrastructure decision that follows. If you know the cost per user, you can model any traffic scenario with confidence.
Step 2: Load simulation and testing
We run load simulations to stress-test the site before the campaign. We take our estimate of peak concurrent users, then double it. That doubled figure becomes our planning benchmark. Over-engineering at this stage is intentional. It creates headroom for the unexpected.
Step 3: Server specification and CDN configuration
In year one, we provisioned a server with a specification generous enough to handle our doubled traffic estimate, then configured Cloudflare to serve as the primary delivery layer. By the time the first campaign launched, we had confidence the setup could handle up to 25,000 concurrent users. The actual peak that year was around 8,000, and the site handled it without any strain.
Step 4: Iterative scaling based on real data
After year one, we had real traffic data rather than estimates. That allowed us to scale the server down to a more efficient specification, tuning PHP thread limits and server thresholds based on what we actually observed. This is the difference between infrastructure built on assumptions and infrastructure built on evidence.
Step 5: Plugin and WordPress updates, followed by a lockout
All WordPress core files and plugins are updated and tested well in advance of the campaign. Once we reach the pre-campaign window, we implement an update lockout. No changes to the site are made during the live period. This eliminates the risk of an update introducing instability at the worst possible moment.
The Backup Infrastructure
Good preparation is not just about the primary setup performing well. It is about what happens if it does not.
For this campaign, we maintained a backup server on a separate hosting provider, connected via a load balancer. Had the primary infrastructure encountered any issues, traffic could be redirected automatically. The backup was sized to handle the load independently. It was a genuine fallback, not a token contingency.
In three years, we have never needed it.
Real-Time Monitoring During the Broadcast
When the broadcast goes to air, our team is watching in real time. The monitoring stack during peak periods includes:
- Google Analytics in real time: tracking live visitor numbers, page activity, and traffic sources as they happen
- Kinsta server metrics: monitoring PHP thread limits, megabytes served, and server load at the infrastructure level
When you are watching a 100x traffic spike arrive in real time, these dashboards are not background noise. They are the instruments you are flying by.
Beyond the tooling, we also had direct access to Kinsta’s senior engineering team throughout the broadcast period. As a premium Kinsta partner, part of our pre-campaign due diligence was ensuring we had an account manager and senior engineers available and genuinely reachable in the event that anything required immediate escalation. That relationship is part of the infrastructure.
Three Years. Zero Downtime.
In three years of managing the Mark Hughes Foundation website through Beanies for Brain Cancer, the site has never gone down during the campaign. It has also consistently loaded fast under peak load. We have video captured in real time during a broadcast peak to demonstrate this.
The Foundation and their team pour an enormous amount of work into this campaign every year. The result they achieved this year, $3.8 million raised against a $3 million target, is a reflection of that dedication and the incredible community that supports them.
Our part in it is simply to give them the best possible platform to succeed. A website that loads fast, handles the traffic, and stays online when it matters most. Everything else is theirs.
What This Means for Your Business
Most businesses will never face a 100x traffic spike from a national television broadcast. But most businesses do face high-stakes digital moments, such as a product launch, a media feature, a seasonal sale, or a viral social post, where their website needs to perform under pressure it was not designed for on an ordinary Tuesday.
The principles we applied here are not exclusive to large campaigns or not-for-profit organisations. They apply to any business that cannot afford for its website to be the thing that lets it down when it matters most.
Load calculation. CDN configuration. Infrastructure headroom. Real-time monitoring. A genuine fallback. These are not exotic capabilities. They are standard practice for a team that has done this before.
At Clickk, our clients have complete confidence in us when their most important moments arrive. That confidence is not assumed. It is built through preparation, real data, and three years of proof.
If your business has a high-stakes moment on the horizon, we should talk.
