Earlier this month, we hosted a webinar with UPS’ CHRO, Darrell Ford and his team, to hear how the world’s largest package delivery company scaled AI across its HR processes and built a governance model to meet peak season demands over the holidays.
For member companies weighing how to move AI from pilot to enterprise, UPS offers a playbook on sequencing, vendor management, and building trust with skeptical operators.
By the numbers:
-
UPS hires roughly 200,000 to 300,000 seasonal workers between Thanksgiving and Christmas—a volume that pushed manual processes to their limits.
-
AI tools shortened offer turnaround for hourly seasonal roles from 48 hours to under 15 minutes, with background check and drug screen offers now under 7 minutes.
-
UPS projects $2.2 million in savings over five years, freeing up HR to focus on higher-value work like workforce planning and succession.
How UPS scaled AI:
-
Built the foundation first. Every vice president, member of the c-suite and the board went through digital fluency training before AI scaled, giving leadership a shared framework and language that UPS says improved adoption rates.
-
Let internal competition shape strategy. Of roughly 88 early AI projects, only 8 made the cut companywide, mostly revenue-facing ones. This project was outside of that focus—HR leaned on partners’ technology roadmaps instead of competing for internal capital.
-
Picked partners, not vendors. UPS had longstanding relationships with Accenture and Fountain and partnered with them on this initiative, aligning both around one shared business outcome rather than individual system priorities.
-
Governed it as a live operation. A cross-functional group met biweekly through peak season in a “hyper care” process focused on getting human escalation paths right before scaling further.
The trust challenge: Teams manually double-checked automated results even after the system worked. Out of roughly 15,000 hires processed, all but a handful cleared successfully. Allowing teams to verify at each step helped build confidence in the technology.
