Contact centres have spent a decade automating. Chatbots, advanced IVR, digital self-service: the goal was clear — reduce the volume of interactions handled by humans. The paradox is that advisers have never mattered more. According to Gartner, 85% of customer relations directors are currently broadening and increasing the complexity of their human teams’ responsibilities, and 75% are moving them into entirely new roles within their organisations. This is not a trend. It is a structural shift. Understanding what is happening in contact centres today means understanding something larger: how a profession reinvents itself when machines absorb the repetitive work, and what remains is the most difficult, most human, most decisive part.
Between 2015 and 2025, the obsession of customer relations leadership was "digital containment": channelling as many interactions as possible into automated journeys to reduce handling costs. Average Handling Time (AHT) was the dominant metric. A short call was a well-handled call. A fast adviser was a good adviser.
That model rested on an implicit assumption: most customer interactions are simple, repetitive, predictable. Which was true — until automation proved it by absorbing exactly those interactions.
What remains on the human floor is no longer simple at all. Complex situations, contractual disputes, customers in distress, cases that have already passed through two or three layers of self-service before reaching an adviser — often frustrated, sometimes angry. The AHT observed on these flows no longer runs between 2 and 5 minutes. It approaches 12 minutes on average in organisations that have genuinely begun their transformation. The industrial model, built on linear scripts, time-keeping, and procedural compliance, simply does not work for these interactions. It never worked particularly well. AI has just made the mismatch impossible to ignore.
Generative AI can today understand a complex request, synthesise a customer history, draft a coherent response, process a management action without error and without fatigue. On those tasks, it outperforms a junior adviser trained in two weeks. That is a fact, and denying it serves no one.
What it does not do is read the tension in the voice of a customer whose case has been dragging for three months. It does not perceive that an elderly person does not really understand what is being explained, even when they say yes. It does not sense that a customer on the verge of cancelling simply needs to be heard for two minutes before anything is proposed.
These capabilities have a name: emotional intelligence, situational adaptability, contextual judgement. They are precisely the competencies that the industrial model undervalued, because they cannot be measured in calls handled per hour. And they are today the only ones the machine does not replace.
81% of customers explicitly express the need for more personal and individualised contact. In a world saturated with synthetic interactions, authentic human conversation becomes a rare asset. What was once the norm becomes the differentiator.
There is a systemic phenomenon that few organisations measure correctly: Failure Demand. These are all the contacts arriving in a customer relations centre not because the customer has a new need, but because a previous interaction failed.
A customer calling because the chatbot did not answer their question. A customer calling back because the email response was generic and useless. A customer contacting the service because the online self-service crashed mid-process. These failure contacts are statistically longer, technically more demanding, and carry accumulated frustration that the adviser must absorb before they can even begin to help.
Several sectors illustrate this clearly. In energy, market liberalisation and price volatility have turned the electricity bill into a near-indecipherable document for the average consumer. Customer service has become a space of pure education, where advisers must explain complex pricing mechanisms to customers already stressed about their budgets. In automotive, the shift to electric has erased decades of purchasing reference points: incentive schemes, real-world range, charging infrastructure, fiscal options. The human becomes an indispensable guide through an entirely new landscape. In insurance, the multiplication of exclusion clauses pushes customers to refuse self-service on the moments that genuinely matter, and to demand a human voice capable of taking responsibility for an answer.
Each failure contact costs two to three times more than a standard contact. Reducing these flows does not come from more automation. It comes from resolving better on the first contact — which is precisely the work of the augmented human adviser.
The Gartner figures do not describe a hypothetical future. They document what is happening right now in organisations that have begun their transformation. 85% of customer relations directors are broadening their human teams' responsibilities. 75% are creating new roles. These decisions are not driven by theoretical anticipation — they respond to a concrete operational reality.
When AI absorbs simple calls, the residual flow changes in nature. Organisations that have understood this are not trying to maintain the same model with fewer people. They are reconfiguring what their advisers do and developing the competencies that correspond to these new flows.
This shift comes alongside another striking figure: 93% of customer relations professionals actively want AI to handle post-call tasks — summarisation, note-taking, file updates. This is not the resistance to AI one might expect. It is a demand to free up cognitive bandwidth for what genuinely has value.
Advisers working with a well-configured AI copilot are not doing less work. They are doing different work — denser, more complex, and often more fulfilling, because it requires being genuinely present to the person in front of them rather than following a decision tree.
The transformation does not simply mean "advisers become more versatile." It produces distinct career paths, with specific competencies and different value-added depending on the sector.
The complex situation expert handles cases that fit no standard process: disputes involving multiple parties, emergency situations, sensitive regulatory cases. Their expertise is not technical in the sector sense — it is relational and procedural. They know how to manage tension, rebuild a damaged relationship, and find a resolution within a constrained framework.
The pedagogical adviser is becoming essential in sectors undergoing major transformation: energy, electric mobility, financial services. Their role is to explain complex environments to customers who feel lost in the face of change. They do not recite a script. They assess their interlocutor's level of understanding in real time and adjust their register accordingly.
The long-term relationship manager is what some call "B2C clienteling," borrowed from luxury retail codes. Using customer data made available through CRM tools and AI copilots, they build relational continuity: they know what the customer has experienced, what they have asked for, what frustrated them, what satisfied them. Each interaction is no longer a ticket opened and closed, but a chapter in an ongoing relationship.
The omnichannel journey coordinator orchestrates interactions that have crossed multiple channels before landing with them: a customer who started on chat, sent an email, then calls because nothing worked. Their value lies in reconstructing the thread, sparing the customer from having to repeat themselves, and resolving holistically what became fragmented along the way.
The transformation of contact centre roles has direct implications across three organisational dimensions that leadership cannot address in isolation.
The first is recruitment. The profiles that were valued in the industrial model — fast script learning, resilience under call volume pressure, procedural compliance — are not the ones that succeed in the craftsman model. It is not that these profiles are poor; it is that they were evaluated on the wrong dimensions. What changes: the ability to manage ambiguity, navigate without a safety net in new situations, and adjust communication register in real time become the priority selection criteria.
The second is training. Training an adviser on a script takes two weeks. Developing situational pedagogy, aggression management, and emotional intelligence applied to customer relations is a continuous investment. The organisations that will successfully navigate this transformation are those moving from a single initial training model to one of ongoing professional development.
The third is performance measurement. As long as NPS and first contact resolution (FCR) remain secondary metrics compared to volume and cost indicators, the craftsman model is not truly valued. Aligning performance indicators with what is genuinely desired is a governance decision, not just an operational management one.
For companies considering outsourcing, these shifts also redefine what to expect from BPO partners. The "lowest cost per act" criterion has not disappeared, but it is no longer sufficient. What matters now is the outsourcer's capacity to train, supervise, and develop advisers on complex flows, to deploy effective AI copilots, and to share a performance vision centred on customer value rather than processed volume. Our guide on measuring outsourced contact centre performance through KPIs and SLAs details how to structure these expectations contractually.
The final implication of this transformation may be the most structurally significant for decision-makers. 85% of business leaders now consider customer relations a direct lever for revenue generation. That figure would have seemed exaggerated a decade ago, when customer service was accounted for purely as an overhead.
The logic is straightforward: a customer who experiences genuinely excellent customer service at a difficult moment is a customer who stays, who recommends, and who often buys more. Retention has a zero acquisition cost. In sectors where acquisition costs five to seven times more than retention, turning customer service into a retention lever changes the entire economic reading.
This is what performance data from contact centres in transformation is beginning to show: organisations that invest in upskilling their human teams and in the right AI tools see their NPS improve, their churn rate fall, and their long-term customer value indicators rise significantly. For a deeper view of how to connect these indicators to financial outcomes, our analysis of NPS, CSAT, and CES — which metrics to choose provides a complete framework.
The relational craftsman is not a romantic vision of the customer adviser. It is a rational response to a market where automation has made human expertise strategic.
Will AI eliminate jobs in contact centres?
AI is not eliminating contact centre jobs — it is reskilling them. Simple and repetitive interactions migrate to automated channels, freeing advisers for more complex flows. According to Gartner, 75% of customer relations leadership teams are creating new roles for their human agents. The transformation raises the skills bar; it does not produce net headcount reduction in organisations managing this transition actively.
What is Failure Demand in a contact centre?
Failure Demand refers to inbound contacts generated not by a new customer need, but by the failure of a previous interaction: a chatbot that did not respond, a self-service journey that crashed, a generic email that solved nothing. These failure contacts are longer, more tense, and more costly than standard contacts. Reducing them does not come from more automation, but from better first-contact resolution — which relies on the quality of human intervention.
How do you measure adviser performance in the new model?
In the craftsman model, the relevant indicators are NPS, first contact resolution (FCR), post-interaction retention rate, and real-time customer satisfaction (CSAT). Average handling time remains useful as a complexity indicator, but must no longer be a minimisation target. A long call that definitively resolves a complex problem has more value than a short one that generates a callback.
What should a company require from its BPO partner in 2026?
A relevant BPO partner in 2026 must demonstrate the capacity to train and develop advisers on complex interactions, to deploy operational AI copilots, and to manage performance on customer value metrics (NPS, FCR, retention) rather than volume alone. Contract models are evolving toward value-sharing mechanisms — success fees indexed to NPS, gain-sharing arrangements — that align provider interests with those of the client.
What is the difference between a contact centre and a call centre?
A call centre handles exclusively telephone interactions. A contact centre manages all channels: phone, email, chat, social media, messaging. The transformation toward the craftsman model operates in multichannel contact centres, where coordinating fragmented customer journeys across channels becomes a core adviser competency.
The transformation of contact centre jobs is not a technology question. It is a question of strategic choice. Organisations treating AI as a short-term cost-reduction tool will miss what matters: the opportunity to turn customer service into a durable competitive advantage.
Organisations that understand AI frees up human time for higher-value interactions — and invest accordingly in the competencies, tools, and performance indicators to match — are building something their competitors cannot easily copy: genuine relational capability, carried by teams trained to exercise it.
In a world saturated with synthetic interactions, that is the ultimate premium product.
→ Want to assess your organisation's maturity on this topic or understand how to structure upskilling for your customer relations teams? Speak to the Armatis team or explore our CX Horizon 2030 study at armatis.com/en/cx-horizon-2030.
Armatis is a European specialist in customer relations and business process outsourcing (BPO), operating across multiple continents with thousands of employees serving companies of all sizes and sectors. The company designs and manages end-to-end customer service operations: multichannel contact centres, complaints handling, technical support, back-office and digitised processes. Backed by integrated technology infrastructure and the ability to adapt to any sectoral and regulatory context, Armatis helps its clients combine operational performance, quality of experience and cost control, wherever they need it.
Contact our teams to discuss your challenges and find out how we can support you
Join the leaders who trust our multilingual and technological expertise.