The capacity question facing CX teams deploying AI is not whether automation works, but what organisations choose to do with the people it displaces. IKEA's chatbot implementation demonstrates this principle sharply: the company automated close to half its customer service queries, reducing wait times and saving EUR 13 million annually. Rather than treating the freed-up 8,500 agents as a cost to eliminate, IKEA's leadership identified a critical insight buried in the chatbot's failure data—customers were requesting interior design consultation services the company didn't yet offer at scale. This wasn't a technology gap to patch; it was a market signal that required human judgment to recognise and act upon. The company retrained its existing agents into a premium design consultancy that generated EUR 1.3 billion in revenue by 2025 and is projected to reach 10% of total revenue by 2028. The strategic choice here matters enormously for CX professionals evaluating their own AI roadmaps.
For support teams and administrators currently implementing Zendesk, Freshdesk, or similar platforms with AI capabilities, the IKEA case raises a fundamental question: are you treating agent capacity freed by automation as headcount to reduce, or as expertise to redeploy? Most organisations default to the former—viewing AI ROI purely through operational savings and efficiency gains. But the underlying principle that IKEA applied applies equally to customer experience functions: relationship depth, product knowledge, and the ability to identify unmet customer needs cannot be automated. When your chatbots and AI agents handle routine queries efficiently, the agents who previously managed those interactions possess accumulated knowledge of your customer base that becomes strategically valuable elsewhere. The real competitive differentiation in CX—whether in financial services, retail, or any sector—emerges from how intelligently leadership deploys that freed-up human capacity, not from the technology itself.
The implications for CX teams are practical and immediate. Rather than planning AI deployments around headcount reduction, teams should be asking what customer problems your agents currently encounter that your systems cannot yet solve, and whether those represent bugs to fix or signals of unmet demand worth pursuing. This requires a fundamentally different conversation between support leaders, product teams, and executive stakeholders before AI implementation begins. The question isn't whether your platform can automate more—it can. The question is whether your organisation has the strategic clarity to recognise what your freed-up people should do next, and the leadership conviction to invest in retraining them for higher-value work rather than simply offloading the cost.
When AI Frees Up Capacity, Where Should It Go? The Full FX