The Quantanite POV

For decades, contact centres have been optimised around a familiar set of measures: average handle time, service levels, cost per seat, cost per transaction, and quality scores.

Those metrics still matter. They help leaders understand whether an operation is functioning efficiently. But they were never designed to answer the more important question: are customer operations actually creating value? 

Across this series, we have looked at that question from multiple perspectives.

• Atreya Chaganty challenged the traditional BPO proposition and the assumption that value should be tied primarily to human capacity
• David Brown argued that the executive conversation needs to move beyond operational KPIs toward customer and commercial outcomes
• Gavin Atkinson examined what happens to the operating model when technology begins to break the traditional relationship between volume, headcount and cost
• We then looked at how Quantanite is beginning to operationalise this thinking through QiQ
• At Kärcher UK, where those ideas are already producing measurable results

Taken together, they point toward a different model for customer operations.

Not a better version of the traditional contact centre, but a different one. A different one. 

The interaction can no longer be the unit of value

The traditional contact centre is organised around these interactions. A customer contacts the business. The interaction enters a queue. Someone, or increasingly, something handles it.

The operation measures how quickly that happened, how long it took and whether the interaction met established quality standards.

That model is useful when the primary objective is processing demand. But processing demand is not the same as solving the problem that created it. An interaction may be handled efficiently and still represent a failure somewhere else in the customer journey.

• An order update was never sent.
• A refund stalled.
• A warranty process created unnecessary effort.
• A customer was transferred because the issue reached the wrong team.
• The contact centre sees the resulting conversation.

But the conversation is often only a symptom. That means optimising the interaction can improve the operation while leaving the underlying problem untouched.

The unit of value must extend beyond the contact itself. The more useful question is, ‘What outcome did the interaction produce, and what does it tell us about what needs to change?’ 

Demand should be understood before it is absorbed

Most customer operations are designed around a simple assumption: demand will arrive, and the business needs enough capacity to handle it.

That creates an entire operating discipline around forecasting, staffing, scheduling and throughput. But not all demand is equally valuable.

Some interactions represent genuine customer needs. Others exist because the business created friction upstream. The distinction matters. If a customer contacts the business repeatedly because a repair process is unclear, the answer is not necessarily to become better at handling repair-status calls. It may be to change the repair process. If customers repeatedly ask for information the business already possesses, the opportunity may not be faster handling. It may be proactive communication.

This is why customer operations need to better distinguish between demand that must be served and demand that should not exist.

The most efficient contact is not always the one handled fastest. Occasionally it is the one the customer never needed to make.

Customer conversations are operational data

This changes the role of the contact centre. Customer conversations should not be treated solely as work to be completed. They are also evidence.

Every interaction contains information about where customers experience friction, where processes break down, what expectations are not being met and what is happening elsewhere in the business.

At scale, those conversations can expose patterns that individual interactions cannot. That is what makes the shift from traditional QA and analytics toward operational intelligence critical.

At Kärcher UK, for example, QiQ analysis showed that approximately 87% of unresolved interactions were driven by process dependencies rather than agent performance.

That single finding changes the improvement strategy. If the assumption is that unresolved contacts are primarily an agent-performance issue, the natural response is more coaching. If the evidence shows the root causes are documentation requirements, warranty validation, repair workflows and missing proactive updates, the intervention has to move beyond the agent. The problem is operational. So the response has to be operational too.

That is what it means to turn customer conversations into business intelligence.

Insight only matters when something changes

Customer operations do not suffer from a lack of dashboards. Most enterprises already have large quantities of information about customer interactions, operational performance and customer sentiment.

The harder problem is turning that information into action.

• Which issue matters most?
• What is causing it?
• Who needs to act?
• What should change?
• Did the intervention actually work?

This is where the traditional insight model often breaks down. Analytics identify a pattern. Someone interprets it. Someone else decides what to do. Another team executes the change. Eventually, someone tries to determine whether it had an effect. The value leaks through every handoff.

The next generation of customer operations needs to connect those stages much more closely.

Understand → Diagnose → Prioritise → Act → Learn. That is the operating loop. The objective is not simply better analytics. It is a system capable of translating customer signals into better decisions and, ultimately, better outcomes. 

Human + AI should be an orchestration model, not an automation target

The same principle applies to AI. Too much of the current discussion focuses on a binary question: What can AI replace?

That is the wrong operating question. The better question is, ‘What is the best way to deliver this outcome?’

Some work should be automated. Some should remain human. Some will require technology to analyse, recommend or prepare while a person makes the final decision. And those boundaries will continue to change as technology improves.

The objective therefore should not be maximum automation. It should be optimal orchestration. Use technology when it can produce the required outcome reliably and efficiently. Use people when judgement, empathy, complexity or ambiguity requires them. And design the system so each learns from the other.

That creates a much more flexible model than simply replacing individual tasks with AI.

‘Scale’ should mean more than adding capacity

This approach also changes the economics of customer operations. Historically, volume and headcount have been closely connected. More demand required more people. More people meant more cost. 

Technology creates the possibility of breaking that linear relationship. The question is no longer simply how quickly an operation can recruit additional capacity. It is how much additional demand the operation can absorb without requiring an equivalent increase in resources.

The Kärcher results provide an early example. During the comparable peak period, total customer interactions increased from 37,188 in 2025 to 44,616 in 2026, a 20% increase in workload.

At the same time:

• Voice transfer rates fell 33%
• Peak abandonment fell 52%
• Peak answer times improved significantly
• Email response times moved from as long as seven days to three to five days
• Voice QA increased to 97%
• Gaps to FCR and NPS targets narrowed materially.

More workload did not translate into worse performance. This is the kind of operating leverage that customer operations should increasingly be designed to create.

Better customer experience and lower cost should reinforce one another

There is a long-standing assumption that customer experience and efficiency sit on opposite sides of a trade-off. Improve the experience and costs rise. Reduce costs, and service deteriorates.

That does not have to be true. If the business identifies why repeat contacts occur and removes the cause, customer effort falls and workload falls with it. If interactions reach the right destination the first time, customers have a better experience, and the operation carries fewer unnecessary transfers. When proactive communication stops customers from having to chase information, it can lead to improved satisfaction and reduced contact volumes.

This approach is a more meaningful form of efficiency. Not simply doing the same work faster. Removing work that should not exist. Resolving the work that remains more effectively. And using human capacity where it creates the most value.

This changes what a BPO should be accountable for

Once customer operations move in this direction, the role of the BPO has to change too.

A traditional outsourcing model is primarily accountable for delivering the agreed capacity and meeting the agreed operational metrics. A next-generation model has to be accountable for something broader.

• Understanding why customers are contacting the business.
• Identifying systemic friction.
• Improving how work is distributed between humans and technology.
• Helping eliminate unnecessary demand.
• Creating operational leverage.
• Connecting customer operations to business outcomes

That is the distinction between providing resources and solving problems. It is also the distinction between being measured by activity and being measured by impact.

The next contact centre should be measured by what it changes

Operational metrics are not going away. Organisations will still need AHT, service levels, QA, workforce planning and capacity management. But those measures should remain what they are tools for running the operation.

They should not be mistaken for the value of the operation. The most important measures sit above them:

• Did we resolve the customer’s problem?
• Did we remove unnecessary effort?
• Did we prevent repeat demand?
• Did we improve customer satisfaction?
• Did we create more capacity without proportionally increasing cost?
• Did we identify something the wider business needed to fix?
• Did the operation get better because we learned from what happened?

That is the shift we believe is now underway. From interactions to outcomes. From processing demand to understanding and preventing it. From analytics to operational intelligence. From automation to orchestration. From scaling headcount to scaling capability.

And from a contact centre that reacts to customer problems to an operation that helps the business understand, solve and ultimately prevent them.

The contact centre has spent decades becoming more efficient at handling work. The next opportunity is bigger: change the work itself.