An AI assistant has been launched, but only 10% of employees are using it. How would you investigate the issue, what metrics would you analyze, what stakeholder interviews would you conduct, and what recommendations would you provide?
I would treat low adoption as a business problem requiring both quantitative and qualitative analysis. I would examine usage data, identify barriers to adoption through stakeholder interviews, and determine whether the root causes relate to awareness, usability, trust, training, process fit, or perceived value before recommending targeted improvements.
Full Answer
When adoption remains low after an AI assistant launch, the first step is to understand whether the problem is awareness, accessibility, capability, or value perception. A Business Analyst should avoid assuming the technology is the issue and instead investigate how employees discover, access, and use the tool within their daily workflows.
I would start by analyzing adoption metrics such as total eligible users, active users, frequency of use, repeat usage, session duration, feature utilization, task completion rates, and adoption trends across departments. Comparing adoption rates by team, role, location, and manager can reveal whether the issue is organization-wide or concentrated within specific groups.
Next, I would conduct stakeholder interviews and gather qualitative feedback. Employees who actively use the tool can explain what value they receive, while non-users can identify obstacles such as lack of awareness, insufficient training, poor user experience, concerns about AI accuracy, security concerns, or workflows that do not align with the assistant's capabilities.
Based on the findings, I would recommend targeted actions. These might include improving onboarding and training, integrating the assistant into existing business processes, addressing trust and governance concerns, refining features that users find difficult to use, and establishing measurable adoption goals with ongoing monitoring to evaluate improvements over time.
Sample Answer
I would approach this as both a data analysis and stakeholder engagement exercise. First, I would review adoption metrics, including eligible users versus active users, frequency of usage, repeat usage, departmental adoption rates, and the most commonly used features. I would look for patterns that indicate where adoption is succeeding or failing. Next, I would interview key stakeholder groups. I would speak with sponsors to understand expected outcomes, managers to understand team-level adoption challenges, active users to identify benefits, and non-users to uncover barriers such as training gaps, usability issues, trust concerns, or lack of perceived value. After identifying the root causes, I would prioritize recommendations based on business impact. These could include targeted training, process integration, communication campaigns, feature enhancements, governance improvements, or revised success metrics. I would also define KPIs and monitor adoption trends to measure the effectiveness of the changes.
How This Applies by Industry
In a large enterprise rollout, adoption may vary significantly between departments. Analysis may reveal that teams with strong management sponsorship adopt the AI assistant more frequently than teams with limited communication and training.
In a SaaS company, employees may avoid the AI assistant if it is not integrated into tools such as CRM, ticketing, or product management platforms that they use daily.
Healthcare staff may hesitate to use an AI assistant due to compliance, privacy, or accuracy concerns, requiring additional governance controls and education.
Frequently Asked Questions
Want personalized interview coaching?
Work with a CBAP® certified consultant
Vikrant Chauhan has reviewed and coached candidates across 30+ real BA/PM/PO hiring processes in healthcare, SaaS, and fintech.