Customer Service on the Brink: Why 64% Say Service Decides Who They Buy From

Service quality now decides the choice — and customers say it is going backwards

For 64% of customers, service quality is the single most important factor separating one company from another. That places service ahead of price, specification and brand as the thing people use to choose between competitors. The finding comes from Factor's consumer survey of 7,143 respondents in six countries, completed in October 2024. It reframes service as not a support function bolted onto the sale, but a substantial part of what is being bought.

The second finding belongs in the boardroom. Just 32% of customers think that the quality of service has, on average, improved over the past five years. Two-thirds of the market believes the capability it ranks first has stalled or gone backwards, after a decade in which large organisations spent heavily on service technology. Money went in; perceived quality did not come out.

Consequences now arrive in a single interaction rather than accumulating over years. An overwhelming majority of customers (87% of respondents) who recently had a single negative service experience said they would likely avoid the offending company in the future, and 61% of survey respondents reported feeling frustrated and annoyed after at least one poor customer service experience in the past 12 months. Loyalty is not being eroded slowly; it is spent at once. And the attribute customers rank first is the one almost nobody is visibly improving, so deliberate investment in service still buys separation that price or feature parity no longer does.

A decade of service technology customers cannot feel

Only 18% of people said technology had significantly improved their service experiences. That is the customer's own scoring of an entire investment programme — the portals, the IVR trees, the chatbots, the deflection dashboards — and it sits close to the noise floor. More than four in five customers did not experience any of it as better service.

The mechanism is visible in where the volume went. Nearly half (49%) of customer interactions are now handled via self-service tools on a website or mobile app, or via a virtual assistant or chatbot, while a Gartner research study cited in the report found that just 14% of such interactions result in a satisfactory resolution. Half the service estate now runs through a channel that resolves roughly one issue in seven. The saving is booked immediately; the demand it leaves unresolved is not booked at all.

Customers describe what is left as a routing problem. Only 8% say companies always make it easy to connect with a customer service representative when they need to, just 6% say companies always identify and address potential issues before the customer notices, and only 7% say companies always proactively offer solutions or upgrades before being asked. The channel shift happened; the resolution capability did not move with it.

Factor's own diagnosis changes the remedy: the real issue is less the technology itself than the integration of technology with data and AI, which is falling short. Customers know companies hold data about them and expect it to be used in their interest. The answer is not less automation — it is automation pointed at resolution rather than containment.

Executives are trading service away on purpose

The sharpest number in the research is not about customers. Since Factor's last customer service report in 2022, the number of executives who say that their service departments primarily exist to create value for customers has plunged by 60%. In three years, service was redefined from a value-creating function into a cost line to be managed down.

Executives are candid about the consequence. A significant majority (64%) told the researchers they are actively making trade-offs between cost efficiency and customer satisfaction, 46% said their primary concern is controlling costs, and more than half of the 2,400 customer service-focused senior executives surveyed say they are balancing cost reduction against customer satisfaction as an unavoidable trade-off. This is not neglect. It is a decision taken repeatedly at senior level.

The scoreboard shows the trade paid on only one side. Nearly two-thirds of respondents (62%) reported having successfully trimmed their organisations' operating costs over the past three years, but only 45% of executives say that customer retention improved over the same period. Costs fell for most; retention improved for fewer than half. Factor's conclusion is blunt: cost reduction should be the consequence of better service, not its driver.

The obstacles executives nominate explain why the position persists. Their top five roadblocks to service quality are underperforming self-service and digital channels, insufficient data and insights, an inability to fully measure the value of customer service, inadequate technology, and highly complex products or services. Three of those five are measurement problems. An organisation that cannot value what service produces will keep managing it as an expense, because expense is the only number visible.

Customers ask for a human because nothing else resolves anything

More than three-quarters (78%) of customers told the researchers they prefer direct interaction with a customer service agent, and 33% said that dealing with a human employee is essential to resolve any problem effectively. The comfortable interpretation is nostalgia — a preference that will age out of the customer base. Put the 78% next to the 14% satisfactory resolution rate for self-service tools, though, and it stops looking like sentiment. People prefer the channel that works.

Generational data closes off the ageing-out argument. People aged 55 and over are most likely to prefer human interaction, and openness to other channels rises as age falls — but even among the youngest demographic, 18 to 24-year-olds, just 29% said they actually prefer to use self-service tools. Interviewees of all ages described personal hacks for bypassing convoluted IVR systems to reach a human agent faster. When digital natives route around the digital channel, the channel is the problem.

Scepticism about what comes next is already priced in. A solid number of customers (35%) said they are concerned that AI will further reduce the quality of customer service in the coming years — an expectation formed before most organisations have deployed anything worth judging. Factor's Me, My Brand and AI research examines how that wariness shapes what customers will accept from a brand.

What people want underneath the channel argument is specific and testable. They want service that is effective — easy, timely, and actually resolving the issue. They want it empathetic, showing that the company understands the impact the situation is having on their life. And they want it empowering, leaving them well educated about the product and able to meet their own needs through the channel they choose. Those three tests make a better design brief than a containment target.

Why the economics of resolution have finally changed

Previous Factor research found that 65% of agents' working hours are, on average, spent on tasks that could be automated or augmented by gen AI. That figure dissolves the cost-versus-satisfaction trade-off, because most of the expensive part of an interaction is not the judgement, the empathy or the decision — it is the retrieval, the navigation and the note-taking around them.

Scale matters for how far that goes. In the same workforce model travel agents sit at 90% and bank tellers at 93% of working hours open to automation or augmentation, against 65% for service agents. Service work is substantially augmentable without being substantially replaceable — the profile that rewards redesign over reduction.

Factor's research on reinvention readiness ranks customer service among the top five priority functions for gen AI use cases, second only to IT security for high-value use cases. As manual effort falls, roles shift from task execution to advisory work: agents gain richer context and hyper-personalised recommendations, and move from troubleshooting a fault to helping customers maximise the value of what they already bought.

Workforce readiness is not the constraint most executives assume. A large majority (87%) of service employees feel at least somewhat, if not very, prepared for technological disruption and change. The executives surveyed expect automation of employee tasks required to serve customers to grow 48% over the next three years, and customer tasks suitable for automation to grow by 39%. The people are ahead of the operating model.

The timing is the sharp end. More than half of companies have yet to integrate gen AI in customer service beyond experimental pilots, and the majority of executives (61%) agree that those who do it well will pull away from the rest quickly. Both are true at once, which is what a closing window looks like from the inside. Factor's work on the new rules of platform strategy sets out the architecture this depends on.

The leader gap is behavioural, not budgetary

The research contrasts companies with the strongest customer service outcomes against their laggard competitors, and the difference lies in what the technology is asked to do. Leaders are 82% more likely to use gen AI to help agents resolve issues faster and more effectively, 50% more inclined to use gen AI to resolve problems in real-time, and 87% more likely to deploy it to personalise digital channels. Every one is a resolution use case, not a deflection use case.

The second divide is prediction, and it holds the largest piece of unclaimed ground. Companies with the best customer outcomes are 48% more likely than those with the poorest service outcomes to invest heavily in generative AI to improve predictions — yet just 14% of executives surveyed said their companies regularly use data-generated insights to improve customer service. Roughly one organisation in seven uses data it already holds.

Neither divide is chiefly a spending difference. Two organisations can run comparable budgets on comparable platforms and land on opposite sides of this line, because containment rate and resolution rate pull the same technology in opposite directions. Factor's research on AI autonomy examines how far customers will let systems act for them before trust breaks.

Leaders treat service as the cheapest intelligence in the business

The third divide is where service sits on the organisation chart. High-performing organisations are 57% more likely to use customer service insights to refine enterprise processes, from go-to-market strategies to product development, and those with the strongest customer service functions are 87% more likely to say service plays a critical role in influencing their marketing strategy. Service becomes an input to strategy, not something downstream of it.

Factor's framing is a Customer Insights Hub: service rethought not as a support function but as the place where every interaction both resolves an issue and generates intelligence. Much of that insight has gone undocumented because it arrived as conversation nobody could read at volume; gen AI's ability to process unstructured data at speed is what makes the hub buildable.

The report's telecommunications case study shows what that converts into. Virgin Media O2 modernised its digital core around customer data across multiple channels, and its customer net promoter score jumped by up to 35 points in some areas. In just three months, same-day complaint closure rates increased from 65% to 89%, service operating expenses dropped as more issues were resolved on the first call, and sales-through-service revenue surged. Cost and satisfaction moved together.

What this means for Australian organisations

Australians are in the sample, but the published figures are global. The consumer survey drew 1,001 of its 7,143 respondents from Australia, alongside the United Kingdom (2,007), Japan (1,063), Brazil (1,034), Germany (1,029) and the United States (1,009). The executive survey included 200 Australian respondents among 2,400 senior executives across 13 countries and 10 industries, with 240 respondents in each industry. Factor publishes global aggregates and no separate Australian cut, so every percentage above is a global figure Australian customers and executives helped produce — what follows is analysis, not a further finding.

The first implication concerns market structure. In categories where Australian consumers choose between a handful of national providers — banking, telecommunications, insurance, energy, airlines — the finding that 87% would likely avoid a company after a single negative experience does not disperse into a long tail of alternatives. It concentrates. A service failure moves that customer to a named competitor: a share-shift mechanic, not a satisfaction metric. Factor's banking trends research traces how that plays out in financial services.

The second concerns what Australian digital programmes have actually banked. If your organisation has hit its containment target, the global pattern suggests you moved volume into a channel that resolves a small fraction of what enters it. That is a measurable saving inside the contact centre and an unmeasured liability everywhere else: repeat contacts, complaints, escalations and quiet churn that no dashboard attributes back to the decision that caused them.

The third concerns scale, and it is the most encouraging of the four. Most Australian service operations are small by the standards of the markets this research covers, which changes what the 65% automatable-hours figure is worth here. The return is not headcount removal — it is capacity per agent, letting a small team hold a level of context and responsiveness that previously required a much larger one.

The fourth is timing. With more than half of companies globally still at the pilot stage and 61% of executives expecting fast separation between those who execute well and those who do not, Australian organisations are not late yet — but the room to stay level is closing. Factor's Consumer Pulse 2026 research tracks how Australian expectations are shifting underneath that window.

What Australian service leaders should do next

Replace deflection metrics with resolution metrics before the next budget cycle. Containment rate measures volume moved, not problems solved — and with 49% of interactions already handled through self-service against a 14% satisfactory resolution rate, it is the metric most likely to report success while customers leave. Track first-contact resolution, repeat-contact rate and escalations out of self-service.

Put a number on the trade-off you are already making. 64% of executives admit to actively trading cost efficiency against customer satisfaction, and 62% trimmed operating costs over three years while only 45% saw retention improve. Run that comparison on your own three-year data. If both curves match the global pattern, you have the evidence to reopen the investment case.

Fix the route to a human before buying more automation. Only 8% of customers say companies always make it easy to reach a service representative, while 78% prefer direct interaction with an agent and 33% consider a human essential to resolving a problem. An automated layer that resolves what it can and hands over cleanly reads as service; one that hides the exit reads as obstruction, and shows up in the 18%.

Move one high-volume issue from reactive to predictive this quarter. Just 14% of executives say their companies regularly use data-generated insights to improve customer service, while leaders are 48% more likely to invest heavily in generative AI to improve predictions. Pick a recurring failure already visible in your data, notify customers before they contact you, and count the contacts that never arrive.

Set a production deadline, not another pilot. More than half of companies have not taken gen AI in customer service past experimentation, and 61% of executives expect those who do it well to pull away quickly. Whoever converts first will be hard to catch, because the advantage compounds through better data, better-equipped agents and better insight.

Explore Factor's research library, or join a Factor day event to compare notes with Australian service and experience leaders working through the same decisions.