IBM's AskHR mandate sank HR NPS to -35 — a user-led pivot lifted CSAT to +80s (interview)

In a July 2024 Fortune interview, IBM CHRO Nickle LaMoreaux recounts the rocky birth of AskHR, IBM's HR benefits-assistant chatbot, first introduced to staffers in 2017 as a purely technical change. "We started on it as a technical change: 'Here's this technical tool,'" she tells Fortune — "and what happened was nobody used it." When IBM realized employees would not take to it willingly, leaders brought down the hammer: in 2018 the company told more than 20,000 workers they were spending too much time on busywork and would have to use the chatbot instead of asking HR.

The mandate backfired. Before it, IBM's HR department had an employee CSAT score of +19 on a 0-100 scale; within the first year of mandatory use, HR's NPS plunged to -35 on a scale of -100 to +100. LaMoreaux calls it a wake-up call: "We were trying to force the behavior change, but we thought about it from the perspective of the HR team." Managers were asking why they should type into a chatbot at all — this was 2018. So IBM started asking managers directly how the chatbot could be better and what they liked about it.

What staffers valued, LaMoreaux says, was asking questions in real time and getting customized responses: AskHR could synthesize a worker's career data to point them to the policies most relevant to them, instead of forcing them through dense employee manuals. IBM reworked the bot to be more direct, instantaneous and personalized — more concise and digestible answers, faster help with job transfers, and role-play for nerve-racking conversations like performance evaluations. "By taking that feedback from them, being centered to who your end user is, that's what then got us out," she says.

HR CSAT rebounded to the +80s (self-reported in the interview); AskHR handles 94% of queries and ~10.1M interactions a year, with 765,000 tasks automated.

IBM launched AskHR in 2017 as a technology drop with no behavior-change plan, and employees simply didn't use it.

The 2018 fix — mandating use for 20,000+ workers to eliminate busywork — optimized for the HR team's goals and cratered employee NPS to -35.

Direct manager feedback about what users wanted (real-time, personalized, digestible answers) is what turned adoption around, per LaMoreaux.

Sensitive situations like low performance or misconduct stayed with human HR leaders, keeping the chatbot from overreaching.

A technology drop with no change plan fails; forced adoption optimizes for HR, not the user. Adoption follows real-time, personalized answers — gradual, user-centered automation.

The interview reports HR CSAT now consistently in the +80s, with AskHR handling 94% of employee queries and resolving around 10.1 million interactions a year, 765,000 tasks automated and the HR operating budget trimmed 40%. IBM has built 280 AI mechanisms in HR over the past decade and is moving AskHR to generative AI for better natural language. LaMoreaux warns that simply throwing chatbots or copilots at a workforce builds resentment against both the technology and its leaders.