Earlier this week, Virgin Money Giving announced that they would boost donations to the London Marathon by 10% after its website crashed the night before the race.
Potential donators were left frustrated well into the morning of race day, despite the organisation’s best efforts to fix the issue. Runners and charities missed out on the traditional last-minute wave of pledges, frustrating their fundraising efforts at just the wrong time.
Virgin Money Giving was swift to apologise, and the donations boost is a nice touch. But of course, the downtime should never have happened in the first place. Especially as a ‘result of high demand’, as the official line ran.
So what can we learn from the fundraising machine behind the biggest annual fundraising event in the world? Here are five lessons we can learn to ensure sites can withstand peak volume:
1. Learn from the past – Organizations should check previous peak numbers, and gauge estimated traffic growth. The Virgin Money Giving site was able to process over £23m in donations on the day of the 2016 London Marathon – that would have been a good starting point!
The most important aspect of website and app load testing is to set and test for the maximum amount of anticipated traffic, plus and additional margin based on past data. Load testing tools can simulate peak environments quickly and easily, and, if your site fails, they will pinpoint exactly what went wrong.
2. Be realistic about use cases. People don’t queue in an orderly fashion online, and no-one’s co-ordinating donation times to give sites a break. The day before, morning of and during the race were always going to be extremely busy. During such peak times, websites must be able to handle the maximum volume of users.
3. Don’t forget external factors. Ads and any other graphics can also impact your website performance. Testing your site ahead of the event to ensure you have plenty of time to fix any issues.
4. What causes your site to crash? Modern transactional websites have multiple components which need to interact to function correctly. Under heavy loads, these components may not function as planned, and often lead to site failure. Through load testing, organizations can identify the weak links and the load levels that case the site to slow, and eventually crash.
4. Strive for resilience. If there’s one thing worse than a website being down, it’s a website coming back at minimal capacity. Understanding how to quickly get your site back up and running following an outage is a vital part of understanding the working components behind your site.
If your site does crash during peak times, make sure you follow in the footsteps of Virgin Money Giving; admit the error, apologise for the problem, and make amends. Better still, make sure it never happens in the first place.