Business Discernment: Why many are rethinking the cloud

David Zey Gerry Pitt

February 23, 2026

For much of the last decade, the message to business owners was simple and consistent: move to the cloud.

It promised lower costs, better performance, stronger security and freedom from managing servers.

Many small- and mid-sized businesses followed that advice.

Some saw immediate benefits, while others are now slowly reassessing those decisions.

This shift does not mean the cloud has failed.

It means business leaders are discovering through experience that technology decisions are not trends to follow, but tools to be applied deliberately.

This column – the sixth in our series on discernment in business technology decisions – is geared toward owners and executives who are asking more sophisticated questions about where their systems should live, what they should cost and how they impact daily operations.

From cloud-first to business-first

Early cloud adoption was driven by speed.

Organizations wanted to modernize quickly, avoid capital expenditures and offload infrastructure management.

Vendors and advisors promoted “cloud-first” strategies, often framing them as the inevitable future.

What many business owners are realizing today is that cloud-first is not always business-first.

The goal isn’t “all cloud” or “no cloud.”

It is the right workload, in the right place, for the right reasons.

This change in thinking is reshaping infrastructure decisions across the SMB market.

The cost conversation has become real

One of the most common reasons businesses are rethinking cloud infrastructure is cost.

Cloud pricing is consumption-based.

For systems that scale up and down, such as customer portals, seasonal workloads or development environments, this model can be highly effective.

But many core business systems do not scale.

Accounting platforms, ERP systems, file servers and databases typically run around the clock. 

When these systems move to the cloud, business owners often encounter rising monthly costs and difficulty in forecasting long-term spend.

This challenge is especially pronounced in testing environments, pre-production systems and emerging AI workloads.

These workloads are often experimental, iterative and resource-intensive.

Though they may only be used intermittently, their cloud consumption can spike quickly and unpredictably.

In contrast, private or on-premises infrastructure provides cost certainty and predictable utilization, which often aligns better with how financial leaders evaluate investment and return.

As a result, financial leaders are now deeply involved in infrastructure decisions that were once considered purely technical, pushing organizations to reconsider where variable cost models make sense and where predictable ownership is the better business fit.

Performance is no longer just an IT issue

Another factor driving reassessment is performance.

When applications were hosted locally, performance problems were typically isolated.

In cloud environments, performance depends heavily on internet connectivity, network routing, geographic distance and provider availability.

For many SMBs, this has led to:

• Slower application response times
• Delays when accessing large files or databases
• Increased reliance on internet uptime
• User complaints that systems feel less responsive

In practice, cloud solutions do not always simplify performance concerns.

In some cases, they shift or even amplify them.

Organizations may find themselves balancing competing priorities, such as designing for disaster recovery versus optimizing for day-to-day operational performance.

What improves resilience during an outage may introduce latency, complexity or cost during normal operations.

These tradeoffs matter because performance directly affects productivity.

Small delays, multiplied across dozens or hundreds of employees, translate into lost time, frustrated staff and reduced output.

In industries such as manufacturing, health care, logistics and professional services, performance directly impacts operations, productivity and service delivery.

As a result, performance decisions are increasingly viewed not as technical preferences but as business risk and operational design choices that require executive input.

Lift-and-shift didn’t deliver as promised

Many early cloud migrations followed a “lift-and-shift” model: moving existing servers to the cloud with minimal changes.

This approach was fast and relatively low-risk, but it rarely delivered transformational results.

Legacy systems moved without redesign often consume more resources than necessary, cost more to operate than expected and provide little improvement in performance or reliability.

A key reason for this mismatch is that legacy systems typically operate with an “always-on” mentality.

They were designed to run continuously, regardless of demand.

Cloud platforms, by contrast, are economically optimized for on-demand usage and pre-reserved capacity, rewarding workloads that can scale down, pause or dynamically adjust.

When always-on systems are placed into environments that price and architect for elasticity, the economic advantages of the cloud largely disappear.

Instead of benefiting from flexibility, businesses end up paying a premium to keep systems running exactly as they always have, just in a different location.

In many cases, organizations moved existing problems into a more expensive environment without gaining cloud-native benefits.

When simplification became more complex

Cloud adoption was frequently positioned as a way to simplify IT – and in some respects, it does.

Physical servers, power and cooling are no longer concerns.

However, many SMBs discovered that cloud environments introduce new forms of complexity, including:

• Cost monitoring and optimization
• Identity and access management
• Shared security responsibilities
• Backup and disaster recovery planning
• Managing multiple vendors and subscriptions

Though providers manage the underlying infrastructure, organizations remain responsible for configuration, access control, data protection, backups, compliance and cost management.

For SMBs without deep IT resources, this shift in responsibility can be significant.

IT vs. operational technology: A critical distinction

As infrastructure decisions mature, another factor is gaining attention: the difference between traditional IT systems and operational technology, or OT.

IT systems include email, collaboration tools, business applications and data platforms.

OT systems include manufacturing equipment, medical devices, building controls, sensors and systems that interact directly with the physical world.

OT systems often have unique characteristics such as low latency requirements, continuous uptime and safety/regulatory impact.

Moving OT-related workloads to the cloud without careful consideration can introduce operational risk.

Even brief delays or connectivity issues can disrupt production lines, patient care or facility operations.

However, the cloud can play a valuable role when used appropriately.

Many organizations keep real-time control systems local while sending operational data to the cloud to support digital twins, simulation, predictive maintenance and long-term performance analysis without impacting day-to-day operations.

As a result, many organizations are choosing a blended approach by keeping latency-sensitive OT systems close to the operation while using the cloud for analytics, reporting and long-term insights.

The rise of hybrid infrastructure – by design

Rather than choosing between cloud or on-premises infrastructure, more SMBs are intentionally combining both.

In a well-designed hybrid environment:

• Core predictable systems stay private/on-prem
• Cloud supports DR and external workloads
• Placement is reviewed periodically

For many organizations, infrastructure placement is no longer a one-time decision but an ongoing financial and operational discipline.

Disciplined business owners’ approach to technology

Business leaders who get this right approach technology with operational realism, not ideology.

They start by asking consequence-driven questions tied to their industry:

In manufacturing

• What stops production if latency spikes or the internet drops?

In health care

• Which systems affect patient flow, diagnostics or compliance?

In professional services

• How much billable time is lost when systems slow down?

In distribution

• What does system lag mean for picking, shipping and inventory accuracy?

They focus on real economics, not projections:

• What does the environment cost at steady state?
• How does spending change during testing, audits or peak demand?
• Can finance clearly explain the bill to non-technical leaders?

Most importantly, they place workloads intentionally.

Always-on, latency-sensitive systems stay close to operations, while variable or external-facing workloads move to the cloud.

“Cloud-first” is replaced with a fit-for-purpose approach.

Reassessment is a sign of maturity

Revisiting cloud decisions is not an admission of failure – it is a sign that organizations now have enough experience and data to refine their approach.

This maturity is not limited to companies that moved to the cloud early.

Many SMBs that have not yet made the switch are also taking a more deliberate approach, learning from others and asking more informed questions before committing.

Many organizations are now repositioning select workloads back on premises to improve performance and cost control.

This selective repositioning, often called cloud repatriation, is becoming more common and more accepted.

Just as often, the opposite shift happens at natural lifecycle moments.

Server replacements, storage expansions or major upgrades create inflection points where the economics change.

Services like file sharing, document management and workflow automation can be expensive to maintain on premises but are often bundled and simplified in cloud platforms.

Cloud savings rarely come from moving everything.

They come when a refresh forces the math to change.

Whether reassessing past decisions, acting at lifecycle milestones or delaying change altogether, disciplined SMBs are applying the same principle: technology strategy should follow business reality, not industry momentum.

The bottom line for SMB owners

The cloud is not going away.

Neither is on-premises infrastructure.

The future for small- and mid-sized businesses is not about choosing sides – it’s about making informed, intentional decisions.

Most SMBs don’t have the time or internal expertise to evaluate these tradeoffs across finance, operations and risk.

Most SMBs don’t have dedicated leadership focused on technology strategy.

Independent, business-focused guidance, often fractional, helps owners evaluate tradeoffs across cost, risk and operations and turn technology from a reactive expense into a deliberate business asset.