What's The Problem With A Traffic Spike?
In a word: the inconsistency. A spike is, as it sounds, a shock to the system for your website.
It’s not prepared for that level of interest, so suddenly and all at once. This can lead to problems such as glitchiness, crashing, and slow loading times. Not great for the old User Experience (UX) unfortunately.
In fact, sites that load in 1 second have a 3x higher conversion rate than sites which load in 5 seconds. As you can see, customers are an impatient bunch: those first few seconds can make all the difference. The positive experience your users have on your site speaks to whether they stick around.
What Causes Traffic Spikes During Competitions?
Generally speaking, a few things can spike site traffic:
- Viral Social Media Posts — a sign of growth, sure. But what happens when it sparks sudden interest in your brand that your site isn’t ready for?
- Media Mentions — maybe you were featured by a news outlet. Just like with viral social posts, this causes a sudden influx of curiosity which your system might not be able to cope with.
- Heavily-Promoted Events — it’s great to get the hype going around your brand and the services on offer. But if it causes a big spike, it can be at the expense of your UX.
- Popular Communications Channels — your newsletters, email and SMS communications can all play a part in driving traffic. So, you know, congrats on the great marketing, but condolences on the potential crash.
- Bots – alas, they’re everywhere these days. Sometimes the traffic to your site is sadly not from human interest at all…
Competition sites in particular require special consideration regarding the closing dates of prize draws and the opening dates of competitions. Unlike other industries where traffic is more consistent, competition sites will see an influx of users around a certain event or anticipated prize.
Here’s what you can do to set strong foundations for your site…
1. Prepparing Your Infrastructure for High Traffic
Step One: Anticipate the Spike
You know what they say: fail to prepare, prepare to fail.
The best thing you can do to set up your site is look at your calendar and see if you’ve got any upcoming events that might spike your traffic. If so, contact your web developers and inform them in advance.
They’ll be able to scale up your website’s capacity for a brief period.
You might be wondering: if it’s possible to scale up capacity, why don’t we just have it scaled up 24/7?
You’re sharp. We like you. However, it can cost a lot to scale a website up in this way full-time.
So, you want to keep it efficient, and just use it as a dam against specific waves of high traffic.
Step Two: Optimise Site Content
You know all your big files, images, and video content? Have you thought about compressing these? At Zap, we’re careful to use plugins that automatically sort this stuff for you without reducing the quality of the content.
The truth is, while they might look fly, lots of big images can cause problems for your loading speed, even when your traffic is at a lull.
It’s good to think about optimising this kind of content—chat to your web developers about how best to do this.
Step Three: Use a Load Balancer
This is a device which distributes your traffic across multiple servers.
It’s like lanes on the motorway, spreading out cars, so that no single lane becomes overwhelmed.
Step Four: Do ‘Load Testing’
Run some tests of your site, making a note of what loads well, of the things that crash, and the things that glitch.
This helps you be better-prepared because it shows you what works and what doesn’t.
It’s like a fire drill: it’ll show you where you need to make adjustments to your infrastructure before the spike actually happens.
Step Five: Compress And Clear Your Cache
‘Caching’ is the process of storing copies of files in a ‘cache’. In human terms, it’s basically a big ol’ storage unit for your data.
Browsers download so much information every time you click on a site. Most of them cache the content to save you loading time. That just means they store it until it expires or the hard drive gets full.
You can compress your cache for periods of high traffic.
Similarly you can use a Content Delivery Network (CDN). Oooooh…
2. Using a Content Delivery Network (CDN)
CDNs store copies of your site in data centres worldwide. They’ll cache the relevant content in a server that’s nearer in proximity to site users, than it is to your server.
Basically, it’s like if you wanted to borrow a cup of sugar from someone, you’d go to your neighbour, rather than someone 600 miles away. Presumably.
It’s best to employ CDNs in your site’s infrastructure BEFORE the traffic spike, and therefore before high traffic becomes an issue. It makes for less strain on your server, giving you quicker loading times.
That means that you don’t have to worry quite so much about whether traffic spikes are going to happen.
What to Do When Things Go Wrong
Now, with a lot of this advice, it might feel a bit like we’ve asked you to crack out your crystal ball and predict a traffic spike, somewhere in the distant future.
🔮“Get your CDN before traffic spikes!” “Think about what might spike your traffic!” “Predict the future!” 🔮
But, look, we’re not näive. Sometimes, the unforeseeable might happen. Traffic might spike, BEFORE you’ve sorted out your infrastructure. Oh! Man! But don’t panic.
Here’s what to do:
First and foremost: communicate with your customers. Set up a ‘Status Page’ which will make sure they’re aware of what’s going on and how long they can expect to wait.
You can also use social media to get the message out there: ‘there’s trouble with our server. We’ll be operating as normal soon.’ It helps them to know that you’re aware of the trouble and looking to fix it.
Then, go back to your devs team at Zap. We know that high-traffic crashes happen to the best of us, and can be a chance to learn about your site’s limitations.
To fix it, we’ll have a look at:
1. Load Balancing and Traffic Distribution
We’ll think about your load balancer software and whether it’s adequately distributing traffic across servers. Then we’ll think about what we can do to distribute more evenly.
2. Auto-scaling and prioritising policies
Your website needs to be able to adapt to demand. If you’re experiencing a lot of crashes, we’ll implement policies into your coding so that the website adapts to meet spikes.
That means it’ll prioritise loading critical content, before any secondary information. User Experience is the most important part of a site, and by giving yours priorities, we’ll make sure it’s still usable, even if it’s not necessarily loading everything at maximum speed.
That limits how often you have to close it down completely due to a traffic spike.
3. Afterwards, Review Site Performance
We’ll use the experience to learn and grow.
By that we mean that we’ll analyse your server logs. These can help us see what kept working well, and what will need optimising for next time your website has to handle high-traffic.
We’ll also encourage you to collect user-feedback. This should cover their experience of the site itself, as well as how openly you communicated with them and answered their needs.
Beep Beep: Coming Through
The important takeaway from this is ‘don’t sweat it’. At the end of the day, a high traffic spike is a sign that your communications are pulling through.
It might just mean that you need to give your site infrastructure a bit of love. Which is exactly what we’re here for, by the way.