Health systems don’t need more data. They need faster access to the insights that drive better decisions. Despite years of technology investment, critical information about spending, supplier performance, and savings opportunities often remains difficult to access and act upon. The data is typically there, but it’s scattered across systems, departments, and reports. As a result, leaders can spend more time searching for answers than evaluating opportunities and taking action.
Conversational AI offers a different way to approach that by allowing users to interact with existing spend data using the kinds of questions they would naturally ask a colleague. Rather than requiring someone to know which report to pull or where specific information is, users can begin with a business question and quickly access the information needed to answer it. The benefit is not simply convenience. It is the ability to move from a question to an informed decision faster.
Health systems are already investing in AI, but many are still working on how to translate that investment into measurable results and scale it effectively. A 2026 report from UPMC Enterprises found that more than 90% of surveyed health systems had deployed third-party AI solutions, but only 44% had a dedicated platform to test them. Reflecting that gap, 63% described their AI strategy as “developing or ad hoc rather than fully established or advanced.” Focused applications, such as purchased services and indirect spend, can offer a practical path forward by applying AI to a clearly defined business problem.
The First Answer Usually Leads to Another Question
Spend analysis rarely stops with the first answer. A team may start by looking at spending in one category, then want to understand which vendors are driving it, whether the pattern varies across the organization, or what is behind a change they are seeing.
Those follow-up questions are important because the first piece of information often provides only part of the picture. A higher-than-expected total spend, for example, may appear to be an immediate savings opportunity until the team understands which facilities are involved, what the existing contracts cover, and whether there are operational reasons for the difference.
Conversational AI fits that kind of back-and-forth because it can carry context from one question into the next. Teams can stay with the same issue and keep narrowing their focus as they learn more, rather than repeatedly pulling new reports or rebuilding an analysis from the beginning.
This approach also expands access to information across the organization. Leaders who understand the business problem no longer need deep knowledge of reporting systems to begin exploring an issue. They can investigate initial questions independently while still relying on subject matter experts when findings require validation or action.
Context Matters as Much as Speed
Getting to an answer faster is useful, but speed alone does not make the answer meaningful. Purchased-services decisions depend on understanding how spending fits within a larger operational and financial picture.
That becomes especially important when information comes from different parts of a health system. The same vendor may work across several facilities or departments, while contract terms and purchasing responsibilities may vary across the organization. A number viewed on its own can mean something very different once those details are considered.
Conversational AI has more value when it helps users stay within that context as they explore the data. Being able to ask a follow-up question based on an earlier response can make it easier to move from a broad finding to a more specific understanding of what is driving it.
The underlying data still has to support that process. A 2026 Guidehouse-HIMSS analysis found that data quality and governance remain barriers to AI implementation for health systems. Easier access only helps when teams can trust the information behind the answer.
Make the Data Useful to More People
Indirect spend knowledge can also become concentrated among a relatively small group of employees. These are often the people who know where certain information lives, which reports are dependable, and how the data needs to be interpreted before anyone can use it. That expertise is extremely valuable, but when every question depends on the same group of individuals, analysis becomes difficult to scale.
Conversational tools can give people a way to investigate those initial questions on their own. But the real transformation isn’t just automation. It’s accessibility. AI opens enterprise data for direct inquiry by non-technical leaders. Executives are no longer limited to pre-built dashboards or fixed SQL queries. They can ask natural-language questions and explore insights in real time. That fundamentally changes decision velocity.
Someone who understands the operational issue can begin exploring the data directly, while employees with deeper purchased-services expertise can become involved when the findings need to be validated, or a decision is ready to move forward.
For experienced team members, that can mean spending less time responding to routine information requests and more time evaluating opportunities where their judgment matters. It also gives other stakeholders a better understanding of the information before those conversations begin.
None of that removes the need for human review. Purchased-services decisions can affect vendor relationships, contracts, and operations across a health system, so the people responsible for those areas still need to decide what the data means and what should happen next.
Follow the Opportunity Beyond the Initial Finding
One of the harder questions surrounding AI investment is what organizations should measure. Adoption metrics can indicate whether employees are using a tool, but they reveal very little about whether the investment is generating meaningful business value.
For indirect spend, there is an opportunity to follow the impact further. A health system can assess whether teams are receiving useful information sooner, whether potential savings opportunities are identified earlier, and whether those opportunities ultimately move forward.
The distinction matters because an insight on its own does not reduce spending. Someone still has to determine whether the finding is meaningful, understand what is driving it, and decide whether there is a realistic action the organization can take. That gives health systems a more useful way to think about return on AI investment.
The value can be traced through the work, from the initial question to the decision that follows, and eventually to the financial result. Insight alone does not reduce costs. Organizations realize value when an identified opportunity moves through evaluation, decision-making, implementation, and ultimately produces measurable savings.
Define Success Before the Rollout
If health systems want to know whether an AI investment is working, they need to decide what success looks like before they start measuring it. Qventus found that 62% of surveyed leaders viewed revenue generation as a measure of AI success, while 59% prioritized hard-dollar cost savings. In purchased services, realized savings may be the ultimate goal, but they are not always immediate. An opportunity may still need to undergo analysis, stakeholder review, sourcing, or contract negotiations before any financial benefit is realized.
That makes it helpful to establish a baseline for the work surrounding those decisions. A health system can compare how long it takes to investigate an opportunity, how many identified opportunities move forward, and, over time, how much of that work results in savings.
Those measures give leaders a better way to assess whether AI is contributing to the process, rather than relying on adoption numbers alone. They also make it easier to distinguish a tool people are simply using from one that helps the organization accomplish something.
Keep the Use Case Grounded in the Outcome
The strongest case for conversational AI in purchased services is not how quickly it answers questions. It is whether those answers help organizations make better decisions and achieve measurable results.
Health systems do not need additional AI pilots that generate interesting insights without clear business outcomes. They need solutions that help turn information into action. Purchased services and indirect spend provide an ideal environment because organizations can follow the impact from the original question through opportunity identification, decision-making, implementation, and ultimately realized savings.
Because when that foundation exists, AI doesn’t just improve processes. It changes how leaders ask questions — and how fast they can act on the answers. And that’s where real competitive advantage begins.
When health systems can trace that complete path from insight to financial outcome, AI becomes more than an experiment. It becomes a business asset.
