At Zoetis, widespread integration of artificial intelligence is firmly paired with a strict mandate to prove the tangible returns on heavy tech investments. The major animal health enterprise, which pulled in roughly $9.5 billion in revenue during 2025, currently deploys AI across research divisions, customer support channels, and daily staff workflows. Keith Sarbaugh, the firm's executive vice president alongside his roles as chief digital and technology officer, treats widespread platform adoption and financial oversight as deeply interdependent duties.

Having joined the organization as CIO in 2023, Keith quickly expanded his oversight to encompass technology, digital strategies, artificial intelligence, core data, cybersecurity, and global customer experience. He directly aligns capital allocation with the practical demands of veterinarians, livestock producers, and pet owners. Explaining his core focus, he noted that his position centers less on managing infrastructure and more on directing strategic technology and data investments to secure real commercial results.

Defining Value Before Investing

Zoetis implements a structured business value realization framework for major technological expenditures. Before any project kicks off, tech executives collaborate closely with corporate finance and business units to outline expected advantages, evaluate profit and loss impacts, and establish clear tracking metrics. Periodic evaluations during development give teams room to pivot, followed by thorough reviews of actual user engagement post-deployment.

Keith points out that this methodology stretches beyond simple metrics, helping optimize tools and continuously generate added value once systems go live. This framework also embraces strategic failures; certain underperforming projects have been pulled back, reflecting his preference for active deployment so teams can quickly assess, refine, or abandon initiatives.

Supply chain operations highlight these clear economic advantages. Zoetis utilizes smart agents as planning copilots to enhance demand prediction and inventory allocation. Keith anticipates these digital upgrades will cut millions in logistics expenses while accelerating product delivery for clients. In a separate technical domain, he noted a 30% to 35% boost in developer output driven by AI assistants, keeping these software engineering gains distinct from anticipated supply chain savings.

Making AI Accessible Across the Company

Initial enterprise rollouts of standard ChatGPT sparked internal debates regarding model adaptability and deployment economics, steering Zoetis toward a more versatile architecture. Their proprietary Zen AI system provides staff with access to roughly 10 or 11 distinct large language models through a unified dashboard. Keith reports that 95% of personnel actively utilize the workspace, with employees having constructed approximately 10,000 custom agents.

Maintaining multiple model choices proves crucial because underlying technologies evolve at varying speeds and offer specialized strengths. Offering diverse options and price points remains essential for extracting maximum business utility. Furthermore, the sheer volume of employee-built agents necessitates active oversight, prompting Zoetis to assign dedicated curators to highlight the most beneficial tools.

One standout application assists personnel in drafting personalized professional development plans through guided prompts. Keith cites this as a prime example of bottom-up innovation that formal corporate planning never would have surfaced. Broad engagement has been heavily reinforced by an internal AI Change Champion Network comprising roughly 350 global colleagues who offer on-the-ground peer guidance.

Connecting AI to Animal Care and Customer Experience

Artificial intelligence is equally revolutionizing how Zoetis discovers treatments and enhances animal care. Keith details successful implementations in target identification, computational drug design, and early failure detection. Selected workflows have slashed the duration required to draft study submissions in half, trimming months from regulatory timelines. Additionally, a clinical data management engine with native AI integration aims to elevate data accuracy while cutting down manual checks.

These enhancements directly translate to superior outcomes for animals and their caretakers by accelerating research schedules. Furthermore, AI tools embedded in veterinary clinics assist practitioners in interpreting diagnostic samples with greater speed and accuracy. Direct field visits to veterinary clinics and livestock farms ensure Keith understands how these digital systems perform in real-world environments.

This customer-centric philosophy informs a unified analytics platform that aggregates data from sales touchpoints, digital platforms, and support centers. Keith estimates this setup achieves roughly ten times the analytical coverage of legacy Net Promoter Score frameworks. Advanced AI parsing uncovered systemic bottlenecks previously hidden across siloed channels, supplying the customer experience group with an initial batch of 10 to 12 targeted optimization goals.

Building Adoption Through Early Involvement

Reflecting on past missteps, Keith concedes that Zoetis initially underestimated the complexities of change management. A compelling business case alone could never guarantee end-user adoption. He openly admits to previously overlooking the critical role change management plays in driving user acceptance, a lesson learned from early rollouts where stakeholder engagement lagged.

The enterprise now actively integrates key stakeholders into every phase, from defining core problems and conceptualizing solutions to guiding system design. This collaborative approach fosters genuine personal ownership, which naturally drives high adoption rates. Ultimately, accurately measuring financial value relies entirely on this active engagement: staff must help sculpt the initiative before technology can genuinely elevate performance.