Learn five essential safeguards to protect your brand from IT crashes and minimize fallout.
Hardly a week goes by without news of a ‘technical issue’ or outage. Sainsbury’s, Marks & Spencer and Tesco Bank are just some of the well-known brands that have experienced tech meltdowns in recent weeks.
We’ve come to expect IT crashes as a part of life, but if handled poorly, they can snowball into a major crisis, tarnishing a company’s reputation, eroding consumer trust and resulting in lost sales.
Here are five safeguards brand owners can put in place to protect their sites and minimise fallout from an IT crash.
Get your comms in order
Tech glitches are much easier to address – and fix when everyone’s in the loop, from your software developers to your marketing team to your customer base. Making your users aware of problems is paramount.
All too often, a bad situation can quickly get worse when details of an outage spread like wildfire across social media (and for some unfortunate brands, this could be the first time they are made aware of the problem). Don’t bury your head in the sand. Be proactive and acknowledge the problem; email subscribers or post a message on social media, reassuring users that you are dealing with the problem and outlining expected timeframes for resolution.
It’s also a great opportunity for your brand to establish or reinforce their ‘personality’ – using a heartfelt apology or humour, for example, if that tone is appropriate. But don’t promise what you can’t deliver – make sure that timeframes and assurances can be met. This is particularly important if there has been a glitch that could affect customer data.
Test, test and test again
We’ve come across many brands that don’t have a back-up structure in place or rollback code, where a system can revert to its previous state. While there’s no standard way to build a system, in our experience, the lack of robust software and website testing often comes down to cost, particularly with smaller brands and start-ups. It can also be due to not having regular access to a software development team or thinking an outage will never happen. Trust me, it can.
Don’t underestimate this – testing is vital to ensure your website works. Typically, this involves a build phase, a testing phase and a quality assurance (QA) check, which aims to identify any issues across the entire build. Too often, testing and QA are the first things to go when brands want to cut costs. It could be an expensive mistake.
Be alert
Do an audit of your site at least once a year with penetration (PEN) testing, where you look for any vulnerabilities in any of your systems. It’s not a guarantee that your site is absolutely secure and glitch-free, but as a brand, at the very least, you will have tried to identify any potential issues and protect your site and stored data. Being proactive with system security alongside testing and QA reduces the risk of outages drastically.
Alongside PEN and code testing, you need to know when systems go down. There is nothing worse than a customer notifying you that your website or platform doesn’t work.
Setting up monitors for your systems to notify you is the first step of your action to an outage. Depending on your user, you can even make these publicly accessible, like Slack (https://slack-status.com/), and other platforms so your users are aware of this issue as it happens.
Consider your site structure
It is possible to limit outages to specific parts of a site, but it will depend on how your website or platform is built and whether different parts of it are hosted on separate services, for example. This approach could help contain the fallout from an outage. Take X, formerly Twitter: its likes and tweets are kept separate where microservices are used for each, so if ‘likes’ were to go down, tweets would still be visible. We would advise this type of structure for brands that would benefit from such an approach. A microservices set up would benefit anyone that’s creating a platform for users need to complete things in, such as banks, ecommerce but not needed for things like marketing websites and ‘brochure’ websites.
Post-crash follow ups
The level of testing required and the number of times this is needed will depend on the size of the company, its user base and its product. It is also essential to take tech developments into account, all of which can impact even the most robust of sites. We recommend PEN tests once a year, but above all, be vigilant, take customers seriously and respect their data.
If you’ve suffered an outage, there’s nowhere to hide. Being proactive rather than reactive shows that you care, which can make a big difference to your reputation.
Explore AITechPark for the latest advancements in AI, IOT, Cybersecurity, AITech News, and insightful updates from industry experts!
