A management team usually expresses the ambition in one sentence: working more customer-focused. Responding faster, fewer handoffs, customers who don't have to repeat their story three times. Everyone at the table nods, and everyone means something different. The sales director thinks of response time. The operations director thinks of fewer escalations. The CFO thinks of lower cost per customer contact. Those are three ambitions, not one. As long as the ambition remains at the level of vision, that is not a problem — only once someone has to determine what will actually change in the work itself does the difference of opinion become visible.
To make the ambition measurable, it must first be broken down into layers: the vision (working more customer-focused), the vision state (what that means concretely for the customer and the organization), the target (a number of steps achievable this year) and the target state (the situation that exists after those steps). Only at the level of the target does it become clear which work must change, and therefore which capacity is needed for it.
In practice, working more customer-focused mainly affects the work around contact: intake, routing, status updates, summarizing a file before an employee calls the customer, looking up previous interactions. That is work that is partly already being taken over by AI, partly can only happen with human oversight, and partly remains human work.
A summary of a customer file, a first draft response to a frequently asked question, recognizing a pattern in complaints — a system can pick that up. Assessing whether that draft response fits the relationship with this customer, or escalating a sensitive situation, requires oversight: someone who approves or rejects, with a reason. And the conversation in which a customer is angry about something that does not fit a template remains human work, not because that is sentimental but because there are not enough comparable cases to base a pattern on.
The difference between companies that already organize this way and companies that do not yet do so rarely lies in the technology. It lies in whether it has been established who decides on that second category: who approves the draft response, based on what information, and what happens if that person is not available. Without that decision, AI work remains alongside the existing process, instead of within it.
The ambition test places eight readiness dimensions against the target state: do you have the data to recognize a customer across channels, is there a system that can produce a summary, is there a role that does the approval, is there an escalation path if something goes wrong, is there a way to measure afterward whether the customer actually experienced the contact as better. Five confidence gates test whether you know that or assume it — an assumption about data quality that has never been tested is not readiness.
From this follows a capability translation: not "AI in customer contact" as an abstract idea, but a role that owns the quality of generated responses, a decision right over when a response may go out without oversight, and a capacity shift — hours freed up from looking up and summarizing, which can be redeployed to the conversations that remain human work. This affects how roles are structured, not how many people are needed; what an employer does with that falls under its own legal requirements and is not part of this test.
This question — what must be organized internally before work shifts — runs through almost every ambition. Anyone working on the ambition to increase quality at the same time will recognize the same eight dimensions, and anyone who wants to combine working more customer-focused with growing without hiring extra people will see that the hours freed up from customer contact are precisely the budget with which that growth is absorbed elsewhere.
After a year, the target state is not an atmosphere but a set of observable things: the number of customer contacts resolved in one go, the time between question and first meaningful response, the share of cases in which an employee approves an AI draft response versus rewriting it themselves, and the hours this has freed up for the contact that needs attention. Without those figures, "more customer-focused" remains a word everyone nods contentedly about, without anyone being able to say whether it has actually succeeded.
Which part of customer contact in your organization truly qualifies for AI takeover differs per task and is made concrete with the work scan from FTE TO AI. That also requires an answer to who will decide on a response drafted by a system, and we explain that separately at what you establish about decision rights during a change — as well as the trade-off between growing faster and the margin that remains intact, addressed at how you weigh growth against margin in a strategy.
The full ambition test, with the eight dimensions and the capability translation into roles and decision rights, is under construction. Ahead of that, there is the free readiness check: eight short questions, one per dimension, which show within a few minutes where your organization is furthest along and where the greatest distance lies to the customer focus you have in mind.
Vertel wat u wilt bereiken, dan kijken we samen wat daarvoor moet staan.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.