How Public Sector Procurement Should Think about AI
Technological innovation is no longer confined to specific industries or market niches. And public administrations can no longer be mere spectators.
Our research shows that scaling generative AI from pilots to full-scale production remains fraught for enterprises – and now agentic AI must be adopted too.
Access in-depth AI adoption and market data, including what enterprises are trying, where they are getting value and what our findings suggest about the best practices needed to increase adoption, scale and return on investment.
Download the ReportAgentic AI often stalls at the pilot stage. We share a framework to align your agentic AI investments to strategic intent, measure performance of agents in business terms and enable continuous improvement.
Matt Aslett
Director of Research Analytics and Data
ISG served as a strategic advisor, fostering a collaborative bridge between the business and its technology data team, and steering a project aimed at unlocking significant value from data across key business domains.
Read More
Is your firm struggling to get value out of your predictive, agentic and generative AI investments?
Get to the next level with ISG Explore more insightsAI is everywhere, and spending is expected to continue growing. But many organizations are failing to translate investments into tangible, repeatable business value. Pilots stall. Hype clouds judgment. Value remains elusive. It's not about infinite trial-and-error PoCs - it's about making the right moves from the start.
Whether it's navigating the AI ecosystem, revamping your existing contracts for efficiency gains or addressing your data foundations and organizational readiness, ISG is your independent advisor to deliver on the promise of applying AI at scale.
Our research shows providers are experiencing 30% to 58% cost savings as a result of AI adoption. Are you seeing resulting price savings? Let's talk AI cost optimization.
AI is enabling providers to deliver solutions faster, better and cheaper. This is impacting almost all traditional IT managed services contracts. Are your providers passing along cost savings as price savings to you?
43% of contact centers have already adopted AI technologies, leading to a 30% reduction in operational costs. But despite these efficiencies, research consistently shows that 75% of consumers still prefer to engage with human agents when dealing with complex issues. So what is AI’s role in the contact center?
Read MoreGenAI in software development leads to substantial productivity gains, estimated at 30-42%. By leveraging AI, machine learning and GenAI, you can enhance and automate IT operations (AIOps). Combining these two ideas is revolutionizing software development.
Read MoreJeff Orr
Director of Research, Technology Research
Traditional AI typically handles things like organizing information, predicting what will happen next and making processes more efficient. Generative AI can produce original content like text, images and even computer code. Agentic AI isn’t simply a tool; it’s more like a digital colleague that can independently solve problems, take action and continuously improve its performance.
AI empowers businesses to operate more efficiently, reduce costs, and improve customer experiences. Generative AI can create content, but agentic AI goes beyond that – it proactively works to achieve goals independently. Enterprises are seeing the positive impact of AI in use cases that have to do with providing excellent customer service, simplifying access to information, streamlining content creation and automating everyday tasks.
Make sure you have a well-thought-out plan, reliable information, a designated owner for the data and basic safety precautions ready to go before you begin. Assess where you are now – looking at your data, technology, team's skills and willingness to adapt – so you can identify where value is most attainable and create a straightforward plan to achieve them.
Start with the outcome you want and your data and work backwards. When choosing a solution, consider its security and compliance, how easily it integrates with what you already use, its total cost, how accurate its results are and the vendor’s reliability for the future. Try it out with a small group initially, and roll it out more widely if it goes well.
Make sure everyone understands the system’s purpose, how it will be run and that people – and not the system – are ultimately in charge. Ensure fairness, document decisions, manage access and monitor performance. Before letting AI act independently in your environment, thoroughly establish safety protocols and review procedures, then consistently monitor them to ensure continued effectiveness.