The past few years have seen a sharp increase in the number and severity of software glitches at major companies. These newsworthy software “fails” or glitches have a negative impact on customer satisfaction, public reputation, and the bottom line. The aftermath of each new failure is frequently projected across the internet, and can follow businesses around for years. One glitch this year was so big that it caused 1.25 million pickup trucks to be recalled because it affected airbags and seatbelts during crashes!

Before we discuss a simple way to prevent these failures, let’s take a look at some of the top software glitches from 2017:

Various airports - Check-in chaos

In September, airport chaos hit all around the world when a global network failure caused disruption to the system used for checking in passengers and luggage. It affected seven airports in seven different countries. Although most airports were able to get the system back up and running in a relatively short amount of time, this failure highlights the vulnerability of critical infrastructure.

American Airlines – No pilots for the holidays

Airlines were hit with another glitch closer to the holidays when American Airlines accidentally told too many of its pilots that they could take time off the week of Christmas. The cause was a computer glitch in their scheduling platform that would have affected around 15,000 flights if not caught. The airline offered its pilots time and a half to resolve the problem.

Fiat Chrysler - Life-threatening recall

In May, Fiat Chrysler recalled over a million trucks due to a software glitch that was linked to at least one deathly crash. The issue was caused by erroneous code that temporarily disabled airbags and seat belt functionality. This isn’t the first software failure to negatively impact the automotive industry in the last few years. A growing number of recalls are being sparked by glitches as cars become increasingly dependent on software.

A lucky Christmas

A software failure affecting the South Carolina lottery made a lot of people very happy. Due to a computer glitch, thousands of winning lottery tickets were printed on Christmas day – totaling 19.6 million dollars if they were all validated. The game was shut down, but not before seven or eight people won hundreds of dollars each!

Suncorp Bank – Vanishing cash

In February last year, a malfunction during a routine upgrade caused the disappearance of money from customers’ bank accounts. Additional customer complaints included overdrawn and locked out accounts.

Bitcoin – Unlimited node crash

Bitcoin suffered from two software failures in the same month last March. The most serious glitch was linked to a software bug that caused over 100 Bitcoin Unlimited nodes (Nearly 70% of the nodes running Bitcoin Unlimited at the time) to disappear from the network completely.

Dodge Ram – 1.25 million recalls

A major software glitch that could cause the airbags and seatbelts in Ram trucks to fail during rollover collisions caused Dodge to recall more that 1.25 million trucks. To prevent the problem, the FCA must now reprogram the onboard sensor of every impacted vehicle.

Cairns Hospital – Security patch gone wrong

A catastrophic glitch affecting five Australian hospitals was introduced during the application of security patches designed to counter potential future cyber-attacks. It required more than two weeks for the hospitals to recover their electronic medical record systems.

British Airways – Disruption for 75,000 passengers

For the sixth time last year – a major IT software failure led to massive cancellations on local flights and significant delays on international flights. According to NPR.org – it took over three days of cancellation chaos to resolve the problems that plagued BA during this outage.

It’s easy to shrug and assume that glitches like this are simply a cost of doing business in an always-on, connected world. Fortunately, this just isn’t true. In fact, one of the primary reasons companies are embracing test automation is because it allows them to find defects before they enter production.

With continuous integration and delivery becoming standard practice, there are more updates and upgrades happening in today’s technology landscape than ever before. But that doesn’t mean these frequent changes should put your business at risk. It is possible for businesses to deploy change while ensuring that every business process works and every application runs as it should.

“Lights out” testing automates your business processes so that almost no human intervention is needed to conduct critical testing on a regular basis. Lights out testing combines traditional testing best practices with next generation automation technology to ensure flawless business process execution and boost test coverage – all while changes and updates are happening in real-time. Ensuring flawless execution of critical business processes is one of the strongest benefits of lights out testing when it comes to the topic of software glitches. Although it can be difficult to quantify savings directly attributable to preventing production defects, the value of ensuring uninterrupted business execution is enormous. Technology glitches create PR nightmares – as you can see from the list above.

Ensuring that all business processes are being tested and evaluated while updates are happening just can’t be done via manual testing which is why continuous or “lights out” testing is necessary. It allows businesses to test at virtually any scale and frequency to catch potential points of failure across the entire enterprise. Automated business process testing can safeguard businesses against software-related disruptions and failures, and keep your business off lists of top glitches in the years to come.

Avoid negative press attention, angry customers, and frustrated employees by embracing continuous testing. This is especially important given the increasing prevalence of Agile and DevOps methodologies. The frequency of change involved in these approaches can’t be handled using traditional methods without risking an epic software failure. Waiting to test until after changes have been deployed is no better – business processes need to be checked early and often to avoid failure. More about the SAFe bet over Agile for enterprise applications here.

Don’t wait until something breaks to make a fix. Worksoft automated testing solutions support the dynamic and collaborative nature of today’s development approaches. You can learn more here.

* This article has been updated with additional software failures on January 8, 2018.