Software Upgrades Consume 28% of Software Budgets, Survey Finds

Large companies in the United States and the United Kingdom spend an average of 28% of their software budgets on vendor-driven upgrades, migrations and licensing changes they would not have actively chosen, according to a survey published by Reuters on October 20, 2026. The research, conducted by software maintenance provider Origina, surveyed more than 2,000 companies and identified recurring technology changes as a growing source of financial pressure, operational disruption and lost engineering time.

The findings highlight a challenge for businesses trying to modernise their technology and invest in artificial intelligence. Companies often accept software changes because refusing them could mean losing technical support for systems critical to their operations. Origina describes these costs as a “software tax,” arguing that businesses are spending substantial amounts to maintain existing technology rather than develop new products and services. The survey also found that many organisations had retired functioning hardware before the end of its useful life and diverted engineering resources away from innovation to accommodate vendor-driven changes.

Key Takeaways

• 28% of software budgets: Large US and UK companies spend this share on vendor-driven upgrades, migrations and licensing changes they would not have actively selected.
• 95% of respondents: Said they had paid for a vendor-driven upgrade they expected to worsen outcomes.
• 80% of respondents: Said leaving a major software vendor would significantly disrupt their business.
• 84% of respondents: Said senior engineering staff always or often worked on vendor-driven changes instead of innovation.
• 91% of respondents: Had retired fully functional hardware in the previous three years to meet vendor support deadlines.
• $12.87 million: Average annual cost attributed to the most recent vendor-driven software change among surveyed organisations.
• AI investment pressure: The survey suggests that supplier-driven technology costs can slow or displace new AI and data initiatives.

What Is the “Software Tax”?

The term “software tax” describes the recurring expenses businesses incur when software vendors introduce upgrades, change licensing terms or require migrations that customers would not otherwise have prioritised.

Unlike a government tax, this is not a legally imposed levy. It is a description of the financial burden created by commercial software relationships, particularly when organisations have limited practical alternatives.

Large companies often depend on specialised software for financial reporting, customer management, supply chains, manufacturing, cybersecurity and internal operations. Replacing these systems can require extensive planning, employee training, data migration and integration with other applications.

When a vendor ends support for an older product or changes the terms under which it is maintained, businesses may have to spend money even if the existing software still performs its job.

For example, a company might operate a stable enterprise application on an older software version. If the vendor stops providing security updates or technical support, the company must weigh the cost of upgrading against the risks of continuing to use an unsupported system.

The decision may involve a substantial migration project, new licensing fees, infrastructure changes and possible disruption to daily operations. The business may therefore spend money on a change that creates little immediate additional value for its users.

This is the central problem highlighted by the Origina survey: technology spending is not always driven by a company’s own priorities. In some cases, the vendor’s product roadmap and support policies determine when the customer must spend.

The Survey’s Main Findings

The survey paints a picture of businesses facing a difficult choice between accepting technology changes and risking disruption to critical operations.

Survey finding	Reported result
Software budget spent on vendor-driven changes	28% on average
Respondents who paid for upgrades they expected to worsen outcomes	95%
Respondents who said leaving a major vendor would significantly disrupt operations	80%
Respondents whose senior engineers often or always worked on vendor-driven changes	84%
Respondents who retired functioning hardware to meet support deadlines	91%
Average annual cost of the most recent vendor-driven change	$12.87 million

Source: Origina survey reported by Reuters on October 20, 2026. The figures reflect respondents’ answers and should not be treated as universal measurements for every business.

The findings suggest that software-related expenses can extend beyond the invoice for an upgrade. Organisations must also account for employee time, implementation work, training, testing, downtime and the possibility that a migration will introduce new problems.

The reported average cost of $12.87 million for the most recent vendor-driven change illustrates the potential scale of the burden among large organisations. It is not a typical cost for every upgrade or a figure that smaller businesses should expect to incur.

Why Companies Feel Trapped by Major Software Vendors

One of the survey’s most significant findings is that 80% of respondents said leaving a major software vendor would significantly disrupt their business.

This reflects the complexity of enterprise technology environments. Large companies may rely on software that has been integrated with systems across multiple departments over many years. The longer these systems remain in place, the more difficult it can become to replace them.

Switching vendors can require businesses to transfer large volumes of data, rebuild integrations, retrain employees and validate that the replacement system works correctly. Some applications may also contain custom configurations that cannot be replicated easily.

For businesses operating in regulated industries, migration can create additional compliance and audit requirements. A change to a core banking, healthcare or industrial system may require extensive testing before it can be deployed safely.

These factors create what technology managers call vendor lock-in. A company may technically have alternatives, but the financial and operational cost of switching can make those alternatives impractical in the short term.

When support for an existing product is approaching its end, the customer can face a difficult decision: pay for the upgrade or accept the risks associated with continuing to use the older system.

Origina chief executive Tomas O’Leary said accepting an unwanted change can be a rational decision when the alternative is losing support for systems on which the business depends.

The survey therefore highlights not just the price of software, but the limited negotiating power some customers have after becoming deeply dependent on a vendor’s technology.

How Software Changes Can Slow AI Adoption

The timing of the findings is important because companies are trying to increase spending on artificial intelligence, automation and data infrastructure.

Businesses want to use AI to improve customer service, automate repetitive work, support software development and analyse large volumes of information. These initiatives can require new computing resources, specialist staff, data preparation and integration with existing systems.

However, organisations cannot always redirect their technology budgets freely. If a large portion of software spending goes towards upgrades and licensing changes, fewer resources may be available for new projects.

Computer Weekly reported that, when respondents were asked about AI or data programmes that had been slowed, paused or cancelled, 34% said IT supplier-driven costs had made a major contribution to the decision, while another 56% said those costs had made a moderate contribution.

These figures suggest that the financial impact can extend beyond maintenance budgets. A business may postpone an AI project because it must first pay to migrate an existing application or update the infrastructure supporting it.

The opportunity cost can be significant. A delayed automation project may postpone expected efficiency gains, while an engineering team occupied with migration work has less time to develop new products or improve customer-facing services.

Nevertheless, not every software upgrade is wasteful. Updates may be necessary to address security vulnerabilities, improve reliability, maintain compatibility or enable new capabilities. The challenge is distinguishing essential investment from changes that provide limited value relative to their cost.

Engineering Teams Are Losing Time to Maintenance Work

The survey found that 84% of respondents said senior engineering staff always or often devoted time to vendor-driven changes instead of innovation or business-focused development.

This matters because experienced engineers are expensive and often difficult to replace. Their work can include designing new applications, improving system performance, strengthening security and building the integrations needed to support AI projects.

When these employees spend weeks or months handling a vendor migration, the company incurs both direct implementation costs and the opportunity cost of the work that must wait.

A migration can also involve testing, troubleshooting and adapting custom software to a new version. Even when the upgrade is technically successful, it may not deliver an immediate improvement in the company’s business performance.

For technology leaders, the problem is therefore one of resource allocation. A project can remain within its approved budget while still reducing the organisation’s ability to pursue higher-value work.

Businesses can respond by improving their software inventories, negotiating clearer upgrade terms, testing alternatives and measuring the total cost of ownership of critical applications. They can also assess whether third-party maintenance is a viable option for certain products, provided it meets their security, compliance and operational requirements.

Fully Functional Hardware Is Being Retired Early

The survey found that 91% of respondents had retired fully functional hardware during the previous three years to meet vendor support deadlines. They reported sacrificing an average of 1.65 years of remaining useful life.

The finding illustrates how software support policies can influence hardware investment decisions.

Enterprise systems are often designed around specific combinations of operating systems, databases, servers and applications. When software support changes, a company may be forced to upgrade hardware even if the existing equipment continues to operate reliably.

Early replacement can increase capital expenditure and create additional environmental costs through electronic waste. Businesses may also need to install new equipment, transfer workloads and test compatibility before the updated system can be used.

However, older hardware is not automatically safe or economical to retain. Unsupported systems can create cybersecurity risks, and newer equipment may offer better performance, energy efficiency or reliability. Companies must evaluate these considerations rather than assume that extending hardware life is always the better option.

The concern raised by the survey is that the replacement decision may be driven by software support deadlines rather than by the hardware’s actual condition or the organisation’s own investment priorities.

What Companies Can Do to Reduce the Burden

Businesses cannot eliminate every vendor-driven change, but they can improve how they plan for and manage software costs.

First, they can maintain an accurate inventory of software licences, support contracts, dependencies and renewal dates. This helps IT teams identify expensive renewals early and avoid unexpected migrations.

Second, procurement teams can negotiate contract terms that provide clearer notice periods, predictable pricing and flexibility around upgrades. Large customers may also seek service-level commitments and explicit information about how long a product will remain supported.

Third, organisations can assess the full cost of changing software before committing to a migration. The calculation should include licensing, infrastructure, implementation, employee training, downtime and future maintenance rather than just the quoted price of the new version.

Finally, businesses can review alternatives for systems that are expensive to maintain. These might include moving to a competing product, simplifying the technology stack or using independent maintenance where appropriate and legally permissible.

Any decision to delay an upgrade should include a security assessment. Continuing to use unsupported software can expose businesses to vulnerabilities, compliance failures and operational risks. The objective is to make changes according to business needs and risk, rather than avoid all upgrades.

The Bigger Picture

The Origina survey highlights a tension at the heart of enterprise technology spending. Companies need software vendors to maintain the systems that support their operations, but the costs and timing of vendor-driven changes can limit their control over IT budgets. When these changes consume engineering capacity and force early hardware retirement, their impact extends beyond maintenance into innovation and capital planning.

The findings also come as organisations look for ways to fund AI adoption. Technology leaders must balance the need to keep critical systems secure and supported against the opportunity to invest in new capabilities. The survey is a warning about the cost of dependence on established vendors, although its results should be interpreted in light of its US and UK sample and the commercial interests of the company that commissioned it.

Looking Ahead

As businesses prepare their technology budgets for 2027, vendor contracts, licensing models and support timelines are likely to receive greater scrutiny. Companies will need to determine which upgrades are essential, which offer measurable benefits and which might be handled through alternative maintenance or migration strategies. The findings suggest that improving visibility into software costs could help organisations protect funding for innovation without compromising security or reliability.

For software vendors, the results highlight the importance of demonstrating the value of upgrades rather than relying solely on support deadlines to drive adoption. Customers increasingly want predictable costs, longer support windows and flexibility to decide when a migration makes commercial sense. Whether these expectations lead to changes in enterprise software contracts will depend on customer bargaining power, competitive alternatives and how much businesses are willing to invest in reducing vendor dependence.

Frequently Asked Questions

1. What does the survey say about software upgrade costs?

The survey found that large US and UK companies spend an average of 28% of their software budgets on vendor-driven upgrades, migrations and licensing changes they would not have actively chosen.

2. What is the “software tax”?

It is a term used by Origina to describe recurring costs associated with vendor-driven software changes. It is not a government tax or a legally imposed charge.

3. How much did the most recent vendor-driven change cost surveyed organisations?

The survey reported an average annual cost of $12.87 million for the most recent vendor-driven change among the organisations surveyed. The figure should not be interpreted as the cost of every software upgrade.

4. Why do businesses continue using software vendors they are unhappy with?

Replacing a major vendor can require expensive data migration, system integration, employee training and testing. For many organisations, the risk of operational disruption makes switching difficult.

5. How can software upgrades affect AI projects?

Vendor-driven changes can absorb IT budgets and engineering time that could otherwise support AI and data initiatives. In the survey, respondents said such costs had contributed to delays, pauses or cancellations of some projects.

6. Should companies avoid software upgrades to save money?

No. Upgrades can be essential for security, reliability and compatibility. Businesses should evaluate their costs and benefits, assess risks and choose an upgrade schedule based on operational requirements rather than cost alone

Sources and further reading

• Reuters coverage of the Origina survey, 20 October 2026

Clarification, 20 October 2026: The headline has been narrowed to software budgets, the denominator used in the survey, rather than total IT budgets. The article’s reported 28% figure is unchanged.