AI implementation? Take small steps for big returns

Thinking how you can implement AI successfully?

Then there may be good reasons why you should start small, think big, and follow the evidence.

That’s one of the conclusions you can draw from SoftwareOne’s latest research, showing that 84% of leading mid-market companies master AI internally before expanding to customer-facing applications.

These high performers—we call them Optimised Innovators—create a practical pattern anyone can follow: begin with focused internal projects, build expertise through hands-on experience, and then scale to more complex and ambitious customer-facing applications.

The results these trailblazers achieve speak for themselves: twice the ROI of their peers, and a clear path to sustainable growth.

Let's look at how these companies make this approach work in practice.

Findings like this confirm some of the lessons we have learned through our own direct client engagements.

As an example, SoftwareOne customer AAMI demonstrated how this AI progression can work in practice.  This leading semiconductor manufacturer, is transforming its workplace with the help of SoftwareOne and Microsoft 365 Copilot.  Building on its established Microsoft 365 environment, AAMI partnered with SoftwareOne to adopt Workplace AI technology, aiming to modernise operations and foster a culture of innovation.

Through SoftwareOne’s Copilot Advisory Service, the client implemented Microsoft 365 Copilot for a controlled group of early adopters during its first week of general availability. This strategic rollout emphasised breaking away from outdated practices (like saving files on local drives) and shifting toward OneDrive for improved security and collaboration. By aligning staff workflows with cloud-based tools, AAMI maximised the value of Microsoft 365 and ensured Copilot could deliver its full potential. The company now has a proven platform to revolutionise how its teams work, moving beyond automation to fundamentally reshape their approach to productivity and collaboration.

Starting with what you know—your internal processes and data— in this way creates a strong basis for more ambitious AI projects. But how do you build on these early successes? That's where systematic skill building comes in.

Build your skills

Redington Group's experience shows how this systematic skill building can work in practice.

Harsh Ramling, their Vice President of Infrastructure, Cybersecurity and Digital Practices, led the company through four strategic stages:

• PRODUCTIVITY: this involved harnessing AI to evaluate code, summarise documents and create effective content.
• PROFICIENCY: this focused on developing an AI assistant to help staff quickly find and use internal documentation.
• PREDICTION: this centered on building data management capabilities and predictive analytics systems.
• PROTECTION: this applied their growing AI expertise to strengthen cybersecurity through automation.

Supporting the Redington example, SoftwareOne has seen the same positive progression with our own clients.

Luxembourg law firm Arendt, for example, exemplifies how this systematic approach works with Microsoft 365 Copilot. Starting with document creation efficiency, they built expertise through small-scale proofs of concept comprising Copilot's integration with familiar Microsoft business applications such as Word, Excel, PowerPoint, Outlook, and Teams – highlighting its potential to streamline content creation, data analysis, document translation, and more. Each success created the confidence to tackle more complex applications, comprising a succession of strategic steps accelerating the firm's path to innovation.

This step-by-step approach to building AI skills delivers a double benefit: immediate productivity gains plus growing capability for more ambitious projects.

Significant value—the “big returns” I mentioned in my title—is delivered when these enhanced capabilities translate into measurable business gains,  fuelling what we at SoftwareOne call the flywheel of cost-optimised innovation.

Consolidate your gains

The commercial impact of this flywheel effect is clearly reflected in our research. In fact, the experience of our “Optimised Innovators” shows that an incremental approach can quickly deliver measurable gains for internal AND external tasks.

For internal processes, 41% generate measurable value versus 28% of others. For service operations, 53% create value through AI enablement, compared to 39% of their peers.

This creates a powerful cycle. Initial efficiency gains free up resources for new AI projects. These projects generate their own returns, which then fund more advanced implementations—the flywheel effect in action.

Our research indicates that within two years, 95% of companies following this progressive approach will advance to sophisticated AI implementations, up from 71% today. The pattern is consistent: start small, measure carefully, reinvest wisely.  It fits in with a more general pattern identified in the study: as middle-market companies become Optimised Innovators, they are twice as likely as others to see improved ROI because of their use of modernised IT platforms.

SoftwareOne client QNET has shown  how well this works in practice. Partnering with SoftwareOne, QNET explored several use cases for Microsoft 365 Copilot, focusing on overcoming writer's block, improving document formatting, streamlining data analysis, and enhancing compliance documentation. To start, 300 early adopters participated in workshops to build familiarity and confidence with the tool. This phased approach allowed QNET to identify practical applications of AI in their daily operations. Employees reported increased enthusiasm for AI-driven productivity, and the company is optimistic about scaling these efforts further in future.

With internal gains consolidated, putting AI expertise to work on customer-facing innovations often offers the next step forward.

The outcomes of following this path are compelling.

Optimised Innovators are 53% more likely than their peers to create value through AI-enabled service operations. They're also significantly more likely to see faster time to market, improved time to value, and greater business agility. And because these improvements build on proven internal capabilities, they tend to be more sustainable and scalable too.