Three in four respondents to an Original Software survey of 500 US senior IT decision makers said updates to business-critical systems were arriving faster than they could effectively test them. The finding points to a gap between the pace of software change and organizations’ ability to verify their own workflows—not proof that every US firm is falling behind or that surveyed firms experienced outages.
What the survey says—and what it does not
Original Software’s announcement for The Enterprise Software Testing Gap reports results from 500 senior IT decision makers in US sectors including finance, insurance, pharmaceuticals, food and beverage, manufacturing, distribution, supply chain, retail, and fashion. Its page does not state when fieldwork took place or provide a sampling method, response rate, or full questionnaire. Treat the figures as the company’s survey findings, not as a census of US businesses.
As an Amazon Associate I earn from qualifying purchases.
| Finding | Original Software survey result |
|---|---|
| Updates to business-critical systems arriving faster than respondents could effectively test them | 75% |
| Cloud adoption increased respondents’ testing needs | 92%; 54% said the increase was significant |
| Finding issues before they affected operations was becoming harder | 59% |
| Respondents relying largely on manual testing using spreadsheets, documents, or email | 36% |
| Respondents with highly automated testing across most systems | 6% |
| IT leaders fully confident their current approach would catch critical issues before business impact | 48% |
All figures in the table are from Original Software’s survey of 500 US senior IT decision makers; the research announcement does not state a survey year. These results describe reported testing capacity and confidence. They do not establish how many firms suffered a failure, or whether cloud adoption caused the reported testing gap.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWhy updates can outpace validation
Testing one application may not test the whole workflow
Original Software says business processes can span enterprise resource planning, warehouse, payment, finance, and customer systems. A change in one application may therefore affect a process that crosses several systems. The company also argues that a software vendor cannot fully reproduce a customer’s particular configurations, integrations, data, and end-to-end processes. Vendor testing can check its product, but it cannot substitute for validating the customer’s environment.
Original Software presents cloud adoption and interconnected workflows as part of the operating context behind rising testing demand. Its survey does not isolate the contribution of either factor or measure a causal effect.
Manual regression work competes for scarce business knowledge
When recurring checks are performed manually, each update can require people to repeat familiar steps. That consumes time and depends on staff who understand what the process should do. Those people are also valuable for interpreting exceptions, exploring unexpected behavior, and judging whether a technical change creates business risk. Automating repeatable checks can free that capacity, but it does not remove the need for human judgment.
AI does not automatically close the testing gap
Separate surveys offer context, but their results should not be combined with Original Software’s figures: they ask different questions of different populations.
SmartBear surveyed organizations already using AI in development
SmartBear says it surveyed 1,436 technology professionals at organizations with more than 500 employees and more than $50 million in annual revenue. Fieldwork took place in Q3 2026. In its US results, 69% said AI wrote or accelerated at least 41% of their code, compared with 43% in January 2026 on the same question. Across all respondents, 55% reported application quality issues in the previous 12 months that they attributed to development moving faster than testing. SmartBear also says 36% reported that testing and verification capacity was starting to fall behind or already behind AI code volume.
SmartBear reports that 46% had shipped AI-generated code that later failed in production; among that group, 69% still expressed a lot or complete confidence in AI-written code behaving as intended. While 83% believed autonomous testing would improve their ability to keep up with AI-generated code, respondents also cited trust in results (23%), integration (18%), governance or compliance (16%), skills (12%), and cost (12%) as barriers to adopting or scaling it. These are SmartBear survey findings, not evidence that a particular testing product will solve the capacity problem.
Applause’s global survey shows adoption alongside quality concerns
Applause’s September 30, 2026 release describes an August 2026 global survey of industry professionals and interviews with technology leaders. More than 92% of respondents said they used AI in testing, up from 60% in the prior year. At the same time, 29% reported an increase in the number or severity of defects, including 15% who said both had increased. Applause also reports that 86% considered human involvement extremely important to functional testing. The figures are global, not estimates specifically for US enterprises.
Rank #4
The surveys show that AI use and concerns about quality can coexist; they do not establish that AI caused defects or that adopting AI testing will improve outcomes. As Applause’s executive statement puts it, “Shipping faster only helps if you can still trust what you’re shipping. That means testing that’s grounded in real-world context and backed by human expertise,”
Free tools Windows power users keep installed
One-click scans. No signup required.
How to build testing capacity around business risk
A practical response is not simply to add more test cases. Start with the processes where failure would matter most, then make recurring validation more reusable while keeping people involved where context and judgment are essential.
Best Value
- Identify critical processes. Work with business owners to name the workflows the organization can least afford to have fail and the operational consequences of failure.
- Map dependencies. Trace the applications, integrations, and data paths involved in those workflows so a change can be assessed beyond the system where it originates.
- Plan validation before updates arrive. Prioritize coverage according to business risk rather than assuming every check can or should receive equal effort.
- Automate repeatable checks. Build reusable tests for recurring updates and routine paths, reducing dependence on people manually repeating the same steps.
- Keep people in the loop. Use business users and quality specialists for process context, exception handling, exploratory testing, and decisions about acceptable risk.
- Review what escaped or was missed. Check whether the approach catches critical issues before operational impact; adjust priorities and coverage based on incidents and gaps.
Original Software’s CEO, Carl Andrews, recommends beginning with processes the business cannot afford to have fail and progressively automating repeatable testing around them. The evidence cited here does not quantify the reduction in defects, costs, or outages that this approach would deliver.
What to assess before adopting more automation
Automation is an approach to capacity, not a guarantee of quality. When evaluating an internal program or tool, consider whether it can support the organization’s actual workflows and constraints.
- Business coverage: Do tests include your configurations, integrations, data, and end-to-end processes, not only individual application features?
- Repeatability: Which checks can run consistently after recurring changes, and which still need manual execution?
- Human oversight: Can people review exceptions, investigate unfamiliar behavior, and make risk decisions rather than merely approve automated results?
- Operational fit: How well would the approach fit existing tools and workflows, governance or compliance needs, staff skills, and available budget?
- Evidence: Can you see what was covered and whether critical issues were caught in time? Adoption rates and stated confidence are not proof of performance.
SmartBear’s reported barriers—trust, integration, governance or compliance, skills, and cost—are useful prompts for that evaluation, but its survey is not a neutral comparison of products. The cited surveys do not establish a universal winner or provide controlled evidence of a particular tool’s performance or return on investment.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
