AI & Customer Service

By: Carlos Francisco Restrepo ~ Manager To Maximize Computer Consultant 

An affirmation to start: the central purpose of The service to the client is to give an effective response to the needs and expectations of the customer, creating a satisfying experience that will contribute to strengthen the image, reputation and sustainability of the organization.

From this definition it follows that the service experience of the customer is not the fruit of chance, but something intentional, should be designed taking into account the promises that the organization does, and putting the customer at the centre of every decision.

In line with the above, and given that the AI has erupted with force and came to stay, a question arises: how can the AI be an effective solution for strengthening the experience of service in an organization with a business model B2C[i]?

Like any process, the Customer Service is likely to be automated and, when done in the proper way, it can generate benefits for both the organization and its customers. Given the power and versatility of the IA, it is logical to consider it as a key tool to strengthen the customer experience. The central question, however, is where to focus efforts and what to prioritise when you incorporate them.

To do this it is useful to distinguish two fundamental components of Customer Service: Attention and Performance:

Care refers to the treatment and the contact between the organization and its clients: accessibility, connection, contact channels (face-to-face, by phone, or digital), waiting times. It requires skills such as communication and empathy, warmth and assertiveness, and is based on the recognition[ii];

The provision, in contrast, relates to the operation of the service promise. That is to say, the set of actions that satisfy a need, through which the delivery of what the customer pays. Relies on processes and requires effectiveness, ease of use, quality, knowledge, skill and technical ability.

Though distinct, attention and performance are interdependent components of the service experience and must be considered together in the design. Technology can enhance both: At the provision, it facilitates the gathering and processing of large volumes of information, streamlines workflows, and allows you to automate operations, reducing errors in the performance of the service; In care, enabling customers to self-manage their needs, at the time and channel that is most convenient for them.

However, if I peguntaran, I would suggest to start by the provision. For one simple reason: less failures of service you have, the less attention will be required; and in an ideal world, a perfect not generate claims, or complaints, requests only.

But I want to focus on the care:

Given the fierce current that the AI has been generated, many companies are implementándola actively in their customer service channels. In this regard, there are statistics[iii] we show that the adoption of chatbots in business B2C is reaching significant levels, marking a structural change in how organizations manage the requests, complaints and claims of its customers, and the trend shows a greater dynamism for the next few years.

Such behavior may be due to several reasons, within these (just speculating a little bit), seeks to reduce operating costs.

While it is legitimate for an organization B2C seek to reduce their costs, automating customer care using solutions AI, it is worth remembering that the purpose of the customer service is not to reduce costs but, rather, to generate a great experience. The key question is to what extent these investments improve really the service from the customer's perspective.

Responding appropriately means recognizing that behind every client is a person, with emotions, expectations, and needs. Therefore, the incorporation of AI in the customer service channels should be from that perspective. But the evidence suggests that is not always the case.

According To Botpress[iv] (2025), 62 % of consumers agree to use chatbots only if you reduce the waiting time; that is to say, they value the immediacy in front of the delay for access to a human agent. However, their acceptance is limited to simple queries, and the satisfaction drops when you do not understand the context, generating discomfort in about 38 % of the cases. Consistently, KatanaMRP (2025) [v] indicates that 49% of consumers prefer human attention, compared to 12 % who opt for chatbots, while 25% said that the preference depends on the complexity of the case: the greater the complexity, the more important of the human factor.

This trend is reinforced by other studies: II Observatory of Trends in Customer Experience Foundever [vi] (2025) shows that about 60 % of the respondents prefer to be served by a person, especially after negative experiences with chatbots. In the same line, a survey reported by Kinsta[vii] (2025), reveals a clear preference for the human attention: 78.3 % believe that humans solve problems faster and 88.8 % believe it should always be the option to interact with a person.

Although there is no specific data on satisfaction in the use of AI, voice (IVR advanced) in contact centers, it is reasonable to infer that the conclusions are similar. What is clear is their rapid adoption: a report of Resonate AI[viii] he points out that approximately 79% of contact centers in the united States are already using technologies of AI in interactions spoken. And it is worth noting that the phone still the middle of preferred contact: according to Foundever, in 2025, what you choose to 64.2% of the customers, followed by WhatsApp and email.

As can be seen, the statistics give us an interesting picture, and sure we can draw some conclusions of value, not necessarily this means that an understanding of the problem or identify clearly the areas of opportunity. So, in order to achieve a better understanding, I invite you to review a real-life case:

In Colombia Claro, Tigo, ETB and other companies offer solutions for communication, both mobile and home, and even packages that integrate both. All of them have invested significant resources in the automation of your customer service channels, and some have already built solutions with AI.

My mother-in-law, who is over 80 years old, has a package of one of these companies with mobile telephony, fixed telephony, tv and internet. We recently commented on some problems with their services, which I summarize in 3 points: (i) The internet signal is not stable; (ii) Not all cable tv channels that you prefer to work well; and (iii) The remote control of the decoder has several keys damaged.

Apparently the only two services that functioned as the promise of service from the company who provided were the signal of cellular and fixed-line telephony.

Keeping in mind that required a technical visit, we proceeded to call the service line. Suffice it to say that for more than 2 hours to rehearse all the existing options in the IVR, but we never got to talk to a person. “The machine”, each time you mark, assumed that if we called it was because at the time of the call, we didn't have service, so I was limited as to say that, after a diagnosis had not found anything out of the ordinary, also claimed that it had carried out actions to restore the service; immediately after, hung.

Incorrect assumption, because the intermittency in the service of the internet, as well as the low image quality of the TV, it does not imply that at the time of the call did not service, but no option in the menu made it possible to make that exception.

It simply was not possible to get what we needed.

On the basis of this experience, what would it be valid to say that the problem was in the technology and the solution IA implemented yet do not have the capabilities required to ensure adequate care?

May be, at least in part, but, in my opinion, it is more evident that the process of care, this company does not have a proper design. I would say that, at the time of defining the cycle call, never consider the experience of the client, or contemplated the possibility that there might be features or exceptions in their needs, which is not the fault of the technology, but from the one who designed the process.

Although the incorporation of AI in the customer service may seem relatively simple (given the exponential growth of solutions like chatbots and agents of voice), do so with alacrity involves significant risks. Any failure or mistake, it will be perceived directly by the client, affecting not only your satisfaction but also the image of the organization. In addition, it is likely that the company is not aware of it, because there will be no one who can listen to what the customer is experiencing and communicating it internally.

On the basis of this example, and the statistics presented above, we could say, to make a good deployment of AI in the service process, and to capture the expected benefits, it is advisable to take into account some key recommendations.

First, the AI itself can function as a first line of care, but should be guaranteed, always a transition simple and accessible to a human being. Because, although the technology arouses interest, people seek to be heard and understood. This is especially relevant in organizations B2C with clients of different age ranges, where older adults often require more support human. But, beyond the age, when the customer perceives that the AI doesn't understand your situation or does not offer a solution in line with their expectations, the need for human contact is intensifying, and if that option does not exist or is difficult, you can generate frustration and dissatisfaction.

Second, the design of the service experience should not be delegated. The challenge is not the technology, but in how it is integrated within the customer experience. This requires strategic vision, business understanding and context, and understanding of customer needs, which are diverse and changing. A poor design can lead to the IA does not capture the nuances of each case, assume, wrongly, that has solved the problem or ignore certain requests require the intervention of other internal processes that, if not functioning properly, undertake the solution.

In synthesis, incorporating AI in the customer care is not an operative decision, but a strategic; it does not belongs to the area of technology, but the high address; and it is not an exercise in simplistic, but a process that requires rigorous planning and conscious, involving those responsible for the service.

To close, I offer some reflections.

Are you a client of a company B2C using IVR or automated AI of voice in customer service?

In case you have had to make any application or submit a claim for non-standard, could have easy access to a human being? And not to be so, how did you feel about that experience?

Now look around your own organization. Do you have invested in chatbots or AI voice to automate the attention? And, if it has, how they are doing something really different to what most of the companies?

If the answer is no, hardly should expect different results in terms of satisfaction. In the end, why do your customers have to feel better than you as a customer of another company, when they interact with a machine and don't have access to a person?

As well said by Albert Einstein: “Insanity is doing the same thing and expect different results”.