US firms struggle to keep up with software testing

Research published by Original Software suggests US enterprise IT teams are struggling to test updates to business-critical software quickly enough. The study found that 75% of senior IT decision-makers face this problem.

The survey of 500 senior IT decision-makers identifies cloud software as a major factor behind rising pressure on testing teams. Some 92% said cloud adoption has increased the amount of testing they need to do, while 54% reported a significant increase.

The findings highlight a gap between the pace of software change and the methods many organisations still use to check whether systems work properly after an update. More than a third of respondents said they still rely largely on manual testing using spreadsheets, documents or email. Only 6% said testing is highly automated across most software systems.

That gap matters because many core business processes now run across several platforms at once. A single workflow can span enterprise resource planning, warehouse, payment, finance and customer systems, making it harder for IT teams to assess the effect of a change in one application before it affects operations elsewhere.

The study found that 59% of respondents believe it is becoming harder to identify issues before they affect business operations. Fewer than half, 48%, said they are fully confident that their current testing approach would catch critical issues before they have an operational impact.

Cloud pressure

Software suppliers can test their own releases, but enterprise customers still need to validate updates in their own environments. That is especially important where businesses have their own configurations, integrations, customisations and data flows, which may produce outcomes a supplier cannot replicate in advance.

Original Software said the shift from on-premises software to cloud services has changed the testing timetable for customers. Where major releases were once planned well in advance, updates now arrive more frequently and often on the vendor's schedule rather than the customer's.

That shift has increased the burden on internal IT teams, particularly in sectors with complex operational processes. The organisations surveyed came from finance, insurance, pharmaceuticals, food and beverage, manufacturing, distribution, supply chain, retail and fashion.

Carl Andrews, Chief Executive Officer of Original Software, said: "Enterprise software has changed enormously, but the way many organizations test their software hasn't changed at the same pace. You cannot keep adding testing workload without changing how it is handled. Core business processes can run across ERP, warehouse, payment, finance and customer systems. A seemingly minor change in one application can have consequences somewhere completely different."

Testing methods

The research also points to the continued use of manual processes in environments that are becoming more interconnected. In practice, that means business users and IT staff may be repeating the same checks each time a supplier pushes an update to a critical system.

A fault that reaches live operations can turn a software update into a business continuity issue. The research presents the problem not simply as an IT workflow challenge but as a broader operational risk for enterprises that depend on continuous access to finance, customer, warehouse and payment systems.

Andrews said: "A vendor can test its software, but it cannot test your business. It doesn't have your exact configurations, integrations, data or end-to-end processes. I'm not saying the answer is to slow down software updates - the answer is that IT leaders need to change how they are currently approaching testing so that they can keep pace and not leave the business, and themselves, exposed. That means planning earlier, automating repeatable processes and treating testing as part of business continuity rather than simply another stage of an IT project."

Original Software argues that one response is to build reusable tests around core processes so repeated checks can be run more quickly when applications change. That approach would reduce the amount of repetitive validation work carried out manually by business users, leaving them to focus on new processes, exceptions and exploratory work.

The company also sought to distinguish automation from removing people from the process altogether. In complex enterprise environments, business users still play a central role because they understand how operational workflows are meant to function across different systems.

Andrews said: "Automation doesn't mean handing everything over to AI. Business users understand how these processes work and remain essential to testing. The opportunity is to stop asking them to manually repeat the same tests every time an update arrives. Start with the processes the business cannot afford to have fail and progressively automate the repeatable testing around them. As software change accelerates, testing needs to become a strategic capability for resilience."