The San Antonio Spurs have spent decades building an organization known for finding an advantage early and developing it patiently. RC Buford has been one of its central architects.
Buford joined the Spurs as an assistant coach in 1988 and returned as head scout in 1994. He later became general manager, served in that role during four NBA championship seasons, and was named NBA Executive of the Year twice. In 2019, he became CEO of Spurs Sports & Entertainment, giving him responsibility for the larger organization behind the franchise. His career has spanned coaching, scouting, basketball operations, and enterprise leadership.
That history gives important context to the way Buford describes the Spurs’ approach to AI. He has spent more than three decades helping the organization identify talent, build systems, and sustain an advantage in a league where successful ideas spread quickly.
"We were looking to create real competitive advantages," Buford said when discussing the organization’s adoption of ChatGPT Enterprise.
The rollout focused on the work surrounding the team. Employees across business intelligence, fan engagement, retail, leadership development, partnerships, and internal operations began building GPTs around problems they already owned.
Within months, OpenAI reports, employee AI fluency rose from 14% to more than 85%. Staff saved over 1,800 hours each month, and 94% of users reported greater confidence with large language models. OpenAI published the results in its case study of the Spurs’ rollout.
Those figures describe an organization learning how to use AI and creating meaningful capacity in the process. They also create a new management question for Buford and his leadership team: where should 1,800 recovered hours go each month?
21,600 hours creates an allocation decision
Annualized, 1,800 hours becomes 21,600. Dividing that by a conventional 2,080-hour work year produces a little more than ten full-time work years of capacity. That calculation sizes the opportunity. The underlying figure comes from OpenAI.
The case study documents the workflows that became faster and the scale of the capacity created. It doesn’t trace those hours into a financial return or identify where every employee redirected the time.
For an executive team, the hours become a resource to allocate. Leadership can direct that capacity toward customer problems, new sources of revenue, work that has remained under-resourced, or greater depth in an area where the organization already has an advantage.
An employee may use the time to complete existing responsibilities with less friction. Elsewhere, a team could increase its output or pursue work that previously sat outside its available capacity. Each outcome carries a different kind of value and needs a different measure.
My read is that this is where the Spurs story becomes more useful than a typical productivity case study. The 1,800 hours establishes the size of the resource. Leadership still has to decide which problems deserve it.
The pilot made usefulness visible
Charlie Kurian, the Spurs’ Director of Business Strategy and Innovation, introduced ChatGPT Enterprise in phases, beginning with 150 pilot users. The rollout included in-person training, custom onboarding guides, peer learning, and an internal hackathon.
The Spurs’ own AI playbook says employees were encouraged to experiment without usage quotas or rigid use-case requirements. OpenAI also reports that people began emailing Kurian’s team to ask for access. Interest spread as employees saw the tools operating inside work they recognized. The Spurs describe the approach as a phased model built around curiosity, practical use, and continuous learning.
A pilot of that size can produce a wide range of internal evidence. Employees see which applications survive contact with daily work. The people leading the rollout learn where users need support and which problems deserve further investment.
The hackathon moved employees from learning about a tool to testing it against their own responsibilities. It also gave the Spurs a growing base of people capable of identifying applications without waiting for a central team to define every opportunity.
OpenAI says more than 150 staff members later participated in an AI Challenge Day and prototyped their own GPTs. By then, the organization had moved beyond a small group learning how to prompt. It had created a broader mechanism for finding work worth changing.
The strongest applications have clear owners
The GPTs described in the case study are tightly connected to specific functions.
Business intelligence analyst Daniel Arzouman built Fan Voice GPT to process thousands of post-game comments and produce sentiment summaries in minutes. Faster analysis gives the organization a chance to see changes in fan reaction while the experience remains current.
Learning and development manager Sarah Cardenas created Transformative Leadership GPT to adapt workshop content for different cultures. OpenAI reports that work previously requiring weeks can now be completed in hours.
Retail strategist Andre Moyce uses ChatGPT to flag suspicious resale listings by comparing their images and details with official Spurs inventory data. The organization has also developed GPTs for sponsorship insights, internal culture, and employee feedback. OpenAI documents the applications and their owners in the customer case study.
The published examples share a useful design characteristic. Each application belongs to someone who understands the underlying work. The owner can recognize a credible result, identify exceptions that need judgment, and fit the output into the way the function operates.
This also makes evaluation more concrete. Fan Voice GPT can be measured through analysis time and the decisions informed by its output. The leadership tool can be assessed through preparation time and the reach of localized programming. Counterfeit detection can be evaluated through the volume and accuracy of listings reviewed.
AI fluency expands the number of employees capable of finding opportunities like these. The business case becomes clearer when each opportunity has a defined owner and an observable result.
Adoption is one scorecard
Moving AI fluency from 14% to more than 85% within months is a significant organizational change. The confidence figure suggests that employees became more comfortable applying the tools to their work.
Those measures establish whether people are using the capability. Financial and customer outcomes require another scorecard.
A faster sentiment summary gains commercial value when it changes how the organization responds to fans. Quicker localization can extend leadership programming into more markets or reduce the time required to prepare it.
Following those outcomes requires more than a company-wide estimate of hours saved. Each application needs an intended result that can be measured over time. Leaders can then see where further investment makes sense, where an application needs to change, and which experiments have become part of the organization’s normal operations.
The 1,800-hour figure can eventually become a portfolio of outcomes connected to specific workflows. That would give the Spurs a clearer view of the return created by the capacity it recovered.
Where the capacity can go next
Buford has said he hopes the efficiencies created by AI will give the people around him more room to solve bigger problems. The Spurs are already identifying some of those problems.
The organization’s upcoming Spurs Performance Institute will include a research track on AI and coaching. The Spurs plan to work with UTSA and the Air Force to study decision-making and cognitive performance. This work opens an area of inquiry that could influence how the organization develops coaches, supports players, and makes decisions.
The institute offers a concrete place to examine how new capability supports bigger questions. The reported time savings and the institute’s research plans are separate developments; the case does not establish that one funded or caused the other.
Sources: OpenAI’s San Antonio Spurs case study and the San Antonio Spurs AI Playbook.
