The Human Factor in the Hybrid Contact Centre

Why contact centre leaders are rethinking “Automation First”

There is a particular kind of silence at the other end of the line when a chatbot replies, for the third time, “I didn’t quite understand. Could you try again?” It is not the silence of patience. It is the silence of a customer disengaging.

For contact centre managers, that silence is measurable. It appears later as abandonment rates, repeat contacts, negative sentiment and churn. What looks, on a dashboard, like “digital containment” can, in practice, be the quiet erosion of trust.

The drive toward automation in contact centres has been swift and, in many ways, unavoidable. Cost pressures are relentless. Volumes fluctuate. Clients demand faster response times and lower cost per contact. Research suggests businesses leveraging AI chatbots can save up to 30% in customer support costs while accelerating response times[i]. The global AI retail market is projected to grow from $9.4Bn to $85.1Bn by 2032, with a CAGR above 31%[ii].

For outsourced contact centre providers and in-house operations alike, the commercial attraction is obvious.

But there is another set of numbers that should give pause.

MindMetre’s nationally representative UK research of over 2,000 consumers reveals a striking dilemma for contact centre-dependent businesses. While 77% believe AI chatbots are increasingly being offered for device support, 75% say chatbots alone are not a satisfactory method of support and would weaken their loyalty to the brand[iii].

This is not resistance to technology as such. We’re all using it in our daily lives. But it is resistance to feeling trapped in a tech-only environment.

The Hidden Cost of “Bots as Gatekeepers”

Many service strategies have quietly repositioned automation from assistant to gatekeeper. The bot sits at the entrance, filtering, deflecting, containing. Only when the algorithm is satisfied does the human appear.

Yet according to Gartner, 64% of customers would prefer companies did not use AI for customer service, and 53% would consider switching if they discovered AI was being used[iv]. More telling still: only 14% of service issues are fully resolved through self-service[v].

For call centre leaders, that 14% figure is critical. It implies that the overwhelming majority of meaningful customer problems still require human intervention to reach satisfactory resolution. If the journey to that human is obstructive, repetitive or opaque, frustration accumulates before the agent ever says hello.

Qualtrics’ UK research reinforces the point: only 5% of consumers prefer automated assistants, while 58% worry they won’t be able to reach a human when AI is in play. Nearly half of bad experiences (44%) result in reduced spending[vi].

For contact centres operating under performance-based contracts or client SLAs tied to satisfaction and retention, these are not abstract concerns. They present huge risks – to margins, to client retention, to sustainable growth.

Why Complexity Changes the Equation

Let’s take the example of an industry that requires major customer support – consumer electronics. When something goes wrong, we all want to get onto the contact centre and get it resolved as fast as possible. We are so dependent on our devices.

Now, some industries can sustain higher levels of automation. Others cannot. Consumer technology support – and increasingly any environment shaped by connected products – sits firmly in the latter category.

Modern device ecosystems create what researchers describe as “double dependency”: work and life are both tethered to functioning technology. Research from Asurion Europe shows two-thirds of working adults consider effective tech support essential, and many are willing to pay for resolution within 24 – 48 hours[vii].

When something fails, it is rarely a single, neatly resolvable issue. Hardware interacts with operating systems, peripherals, cloud services, security layers and user accounts. Each household configuration is unique. Each user brings different levels of technical literacy and emotional tolerance.

Additional Asurion research indicates that 33 – 55% of consumers experience significant device breakdowns quarterly, and 66% want a unified support and insurance model to avoid navigating multiple hotlines[viii].

In such environments, the value lies not in speed of initial response, but in quality of diagnosis and confidence in recovery. Resolution, not deflection, is what customers remember.

When Automation Undermines Itself

There is also evidence that consumers are adapting defensively to automation-heavy environments.

A widely reported incident involving a well-known delivery company saw its AI chatbot swearing at customers and criticising the company, forcing a public rollback. The reputational risk was immediate and viral.

More quietly, ServiceNow’s 2024 Consumer Voice analysis found that 75% of UK consumers deliberately press no IVR buttons in order to force a human interaction, while 77% search online for ways to bypass chatbots[ix].

For contact centre managers, this behaviour has operational consequences. Forecasting becomes unreliable. Containment rates are artificially inflated. Agent queues spike unpredictably as customers “break through” en masse. What was designed as efficiency becomes instability.

Academic research in 2024 links chatbot-led service failure to frustration, aggression and reduced loyalty, while noting that easily accessible human touchpoints alongside bots mitigate that potential escalation[x].

Automation is great… until it becomes a barrier.

Where Intelligent Automation Belongs

None of this amounts to an argument not to leverage available technology.

Automation performs exceptionally well in structured domains: status tracking, password resets, appointment confirmations, triage, session summarisation and knowledge retrieval. In these contexts, it reduces friction and liberates agent capacity.

The strategic question for contact centre companies is therefore not “Should we automate?” but “Where does automation create value without eroding trust?”

Analysts at Forrester argue conversational AI works best when deployed as assistive rather than obstructive. In operational terms, that means:

Automation gathers context before escalation
Automation supports agents with real-time insight
Automation shortens diagnostic paths
Automation summarises and documents interactions.

But crucially, automation does not try to obscure the human option, nor treat escalation as failure.

Designing the Hybrid Contact Centre

The future contact centre is neither fully human nor fully automated. It is deliberately hybrid.

In such a model, escalation pathways are transparent and fast. Customers can see the human door and open it without friction. When they do, the agent receives the full conversational history and diagnostic data gathered by the bot. The transition feels continuous rather than adversarial.

Agents themselves are augmented by AI: knowledge retrieved at speed, suggested resolutions, case summaries generated in real time. Handling and resolution times improve not because humans are removed, but because humans are empowered.

KPIs also evolve. Average handling time and call-length management remain important, but they are intelligently combined with data on first-contact resolution, repeat contact rates, satisfaction post-escalation and – most critically – churn. Customers who abandon during automation rarely file complaints. They simply disappear.

A Strategic Choice for Contact Centre Leaders

The headline MindMetre finding – widespread chatbot exposure coupled with a majority who say bots-only weakens loyalty – should not be ignored.

In a world of growing technological dependency and rising consumer expectations, resolution is the true currency. Automation is indispensable in achieving efficiency and scale. But when deployed as a wall rather than a bridge, it quietly undermines the very loyalty it is meant to protect.

Business that depend on their contact centres have to be bold and smart when presenting the strategic value of call resolution and satisfaction rates to the financial markets. Short term cost reduction delivers far less value that long-term customer loyalty and upsell. Investors are clever people who understand this. They are actively seeking sustainable returns over time, rather than flash-in-the-pan metrics associated with the private equity world. So these metrics are not only for day-to-day management but are also for strategic corporate reporting.

For call centre managers and contact centre companies, the choice is not between humans and machines. It is between short-term cost optics and long-term customer value.

In a hybrid world, technology may open the conversation.

But it is still the human voice that most satisfactorily closes it.

Notes

https://www.invespcro.com/blog/chatbots-customer-service/

[ii] https://www.fortunebusinessinsights.com/artificial-intelligence-ai-in-retail-market-101968

[iii] www.mindmetreresearch.com

[iv] https://www.gartner.com/en/newsroom/press-releases/2024-07-09-gartner-survey-finds-64-percent-of-customers-would-prefer-that-companies-didnt-use-ai-for-customer-service

[v] https://www.gartner.com/en/newsroom/press-releases/2024-08-19-gartner-survey-finds-only-14-percent-of-customer-service-issues-are-fully-resolved-in-self-service

[vi] https://www.intelligentcxo.com/2025/11/06/ai-cynicism-threatens-loyalty-as-95-of-uk-consumers-would-rather-avoid-chatbots/

[vii] https://assets.ctfassets.net/16nm6vz43ids/6gJp3l0cFfVhUPjgPGSOup/a7733a61df1ed5b213747435b53fe367/Asurion-Home-Work-Report-eBook_2024-11-WEB.pdf

[viii] https://europe.asurion.com/europe/newsroom/

[ix] https://www.enterprisetimes.co.uk/2024/08/12/uk-consumers-are-avoiding-chatbots/