Part 2 of Tenet’s Enterprise Operating Models with AI series
There's a fork in the road that every organisation hits once AI starts delivering results.
One path treats AI as a labour replacement engine: automate tasks, reduce headcount, and take the cost savings where you can.
The other treats AI as an amplifier of human capability: free people from low-value work and redirect them into work where human judgment, creativity, and relationships generate value that AI cannot.
The human amplification model in practice
In 2021, IKEA's largest franchisee introduced an AI chatbot to handle routine customer service queries. Within two years the bot was resolving close to half of all inbound contact centre enquiries.
IKEA could have taken the cost savings and reduced headcount. Instead, they retrained over 8,500 contact centre workers as remote interior design consultants. These were people who already knew the product catalogue and were skilled at navigating customer conversations. IKEA built on that foundation, upskilled the workforce into a paid advisory service, and saw significant growth in their interior design advisory revenue in the years following. The same workforce shifted from processing inbound queries to nurturing and converting sales leads.
This is the outcome of a deliberate operating model design decision. The question facing most organisations is whether they plan for this or default to cost reduction because it is the easier outcome to explain.
Start with business capability, not technology
The organisations getting this right are starting their redesign from what the work needs to be, not from what AI can automate. That sounds obvious, but the default approach still seems to be technology-first: here is the AI capability, here are the tasks it can handle, here is the headcount implication. That sequence consistently produces narrow outcomes.
A business capability approach starts from a different place. What outcomes does a function need to produce? Which parts of that work requires human judgment, creativity, empathy, or accountability? What parts are repetitive, rules-based, and high-volume? Design the AI layer around the execution work and design the human roles around the work that humans do best. Then make sure that the two are appropriately connected, resourced, and governed. The difference in outcomes is significant.
Consider the HR function. A technology first approach looks at the current team and asks what can be automated. CV screening, interview scheduling, onboarding paperwork, leave management, standard policy queries. Automate those, reduce the administrative headcount, and reduce the footprint of the function. The end result is an internal HR efficiency story.
A business capability approach starts from the purpose of the function: attract, retain, and develop the right talent to build an organisation that performs.
With AI handling the repetitive transactional layers continuously, HR will have the capacity and data to operate differently. The function can analyse internal capability against what the strategy requires, identify critical skill gaps before they become execution failures, and surface retention risk months before a resignation. The business case for keeping a high performer typically gets built after they have already decided to leave. A senior commercial or technical hire carries a replacement cost of one to two times annual salary before accounting for the revenue impact of the transition period. Getting ahead of that risk even a fraction of the time pays for the capability investment.
The HR function is not only more efficient, but it becomes commercially relevant in a way it rarely has been, contributing directly to the organisation's ability to execute strategy and retain the people it cannot afford to lose.
When you design with technology first, the tendency is to eliminate roles and, in doing so, lose institutional knowledge, employee trust, and the capacity to handle complexity. When you design with business capability first, every function becomes a potential source of competitive advantage rather than merely a cost to be optimised.
How are you designing your future workforce?
Every organisation deploying AI is making a choice about what kind of workforce it is building, whether it has articulated that choice or not.
The path most organisations default to produces a narrower, cheaper operation. The path that requires more deliberate design produces a more capable one, where AI handles the execution layer and the human workforce is concentrated in the work that creates the most value.
Getting to the second requires starting from business capability, designing the operating model around where humans create the most value, and being willing to invest in the workforce transition that a genuine amplification model requires rather than taking the efficiency savings early and calling it done.
In Part 3, we look at the practical design decisions that determine whether AI-augmented roles actually hold up under real world conditions.