Successful AI Transformation Starts with People

Rusty, you work in marketing technology, but your role covers a much broader range of perspectives. What does your work involve?

Rusty Warner: Marketing technology is a fairly general description. I spend time with marketers, helping them understand their strategy and identify the right technology for their needs. I also work with IT teams, enterprise architects, and data scientists on the wider ecosystem in which marketing operates.

In addition, I speak extensively with technology vendors. My role is therefore about bringing these different perspectives together.

When we last spoke, you argued that technology alone cannot create genuine customer centricity. What has AI changed or made more urgent since then?

What we said a year ago is still true, but it has become even more apparent. Without the human element, AI can accelerate and scale bad practices that technology already enables.

Companies are beginning to realise that they cannot simply switch AI on. Nor can they automate processes and operations that have never been properly documented or understood. They therefore need to spend more time with employees. How do employees talk to customers? How do they encourage customers to share their needs? How do they connect those needs with the company’s products and services to create a good outcome?

Companies also need to spend more time with customers and analyse customer feedback more effectively. There are no shortcuts. You have to engage with people, capture the right data, turn it into insight, and only then determine how it can be programmed or supported by AI.

Many vendors promise to help companies perform existing tasks faster and more cheaply. Is that not enough?

One of my frustrations with MarTech vendors is that they often use new capabilities to do the same old things better, faster, and cheaper. They build an audience, create a campaign, automate it, and measure conversions. I would have hoped that we could use this intelligence, together with employee and customer insights, to evolve marketing. Marketing should not simply send people more offers. It should engage with them based on their actual needs.

If the best we can do with AI is repeat what marketing has always done, only faster, we are simply replicating and scaling poor marketing and poor engagement.

You refer to the «human foundation» of an AI-powered enterprise. What does that mean in practice?

The difficult part is convincing people that we do not need to start from scratch. The foundation already exists: people, process, and technology. The problem is that companies have often abused that model.

Too frequently, they have skipped over people and processes, bought new technology, and then adapted their processes to fit the system. Employees are left with broken processes and technology they are forced to use, rather than being actively involved in defining the processes and technological requirements.

The human foundation uses the same three pillars, but people must become the most important one. And people no longer means employees alone. It must also include customers, prospects, and partners.

Processes then need to be designed with those people and agentic AI capabilities in mind. Technology investments should directly support the people and the processes, rather than purchasing attractive new AI tools and only afterwards deciding how to use them.

Why is this broader understanding of people so important?

Even companies that describe themselves as customer-focused often define customers in terms of the products and services they want to sell. They do not necessarily think about the person’s real needs or desired outcome. Nobody particularly enjoys shopping for a mortgage. The desired outcome is to live in a dream home, perhaps with enough space for a growing family. The mortgage is a necessary part of achieving that outcome, but it is not the outcome itself.

Thinking from the customer’s perspective may show that a bank needs to train advisers differently, work more closely with estate agents or public authorities, and determine which parts of the process can be offered through self-service. It also needs to understand which customers are financially confident enough to proceed independently and which require more support. That involves many more people than simply having employees who know how to sell a mortgage.

The company is therefore only one part of the customer’s wider value network.

Every customer is the star of their own movie. A company is usually only a small part of that customer’s value network. Consumers work out what meets their needs, what gets in their way, and how they want to realise value. A company’s products and services may contribute to that value, but the company is rarely the centre of the story.

You reframe people, process, and technology as human-led strategy, human-focused operations, and human-first transformation. What does this help leaders see?

The traditional model remains valid, but it needs to be interpreted differently. A human-led strategy considers employees, customers, prospects, and partners as people. Their needs should guide everything from business goals to investment decisions.

Human-focused operations adapt processes to the way people work, engage, buy, and interact. People should not have to adapt themselves to poorly designed processes.

AI also gives us a major opportunity to reinvent. Companies cannot discard everything they already have, because that would not be cost-effective. But instead of simply adding AI, they should ask how they would operate differently with these new capabilities. Transformation should not be data-first, AI-first, or mobile-first. It should be human-first.

What do companies need to understand about customers, employees, and their brand promise before investing in AI?

The easy answer is that it comes down to data that can produce meaningful insight. But that is not easy to achieve. You need to bring the right people together to understand what really matters.

In the automotive sector, demographic data or the fact that someone visited a dealership can help make marketing more specific. But companies would take a very different approach if they understood why the person wanted to buy a car. Is safety important? Do they need space for a family? Are they planning holidays, or do they mainly need the vehicle to commute? Too often, companies simply add another database parameter so that they can place someone into a segment. That is not the same as genuinely understanding the customer.

The same applies in health and beauty. Rather than focusing on age, gender, or ethnicity, companies should understand how someone wants to wear their hair, which products complement their skin tone, or what would make them feel attractive and confident. The challenge is to convince marketers to capture what is truly relevant to the person’s desired outcome, rather than continuing with marketing as usual.

In many companies, marketing, service, and IT still pursue their own AI initiatives. What happens when AI serves departmental priorities rather than customer needs?

That is precisely why people, process, and technology need to be reconsidered. Service had its own people, processes, and technology. Marketing had its own people, processes, and technology. IT believed it owned the technology and often paid too little attention to the people or processes. This created silos within what was intended to be an integrated infrastructure.

If all three dimensions become people-driven, and if people includes employees, customers, prospects, and partners, companies can use this moment of AI transformation to rethink the foundation holistically rather than by department. It involves a great deal of work. There will be mistakes, and many organisations will resist it. But without this approach, I do not believe there is a path to sustainable success.

How far have companies progressed on this journey?

We are not seeing this change at scale. In fact, we are not even seeing AI itself deployed at scale in production. Most AI applications are still in experimentation or pilot mode. Where they have been deployed, their scope is generally limited. They may support an employee assistant or a customer-facing assistant, but they usually remain confined to individual silos.

Are there companies that are beginning to connect the dots?

Club Med, for example, started in Brazil because it recognised that most customers there communicate through WhatsApp rather than visiting the website or app. The company analysed customer feedback and found that customers mainly wanted information about products and prices. It made that information easily accessible through WhatsApp and then provided the same capabilities to employees. When the interaction moves from self-service to an employee conversation, the customer experiences continuity.

Home Depot is another example. It provides kiosks in stores to help customers find products, while employees use the same application on their phones to assist customers in person. The company has also recognised that more people are managing their own renovation projects or working as independent plumbers, builders, or carpenters. Those customers may need project-management support, bulk purchasing, deliveries to a building site, and advice about the tools required for a particular job. Home Depot is therefore connecting customer support with its stores, supply chain, inventory, delivery, and logistics. It is not perfect yet, but I like the way the company is thinking about the challenge.

How does this relate to connecting customer experience, brand experience, and employee experience?

Those three perspectives form the total experience. We now understand the critical drivers of customer experience, brand experience, and employee experience. A company might begin with a specific customer-experience driver, such as ensuring that its website or app is useful, provides the necessary information, and enables self-service. That driver will create technology requirements. But it will also create process dependencies, because once a customer makes a decision, the organisation needs to act on it. Depending on the situation, employees may also need to follow up. Starting with an experience driver allows companies to connect the required investments in people, processes, and technology and translate them into a roadmap.

Before AI is scaled across the enterprise, what evidence should leaders demand?

The first question I would ask is how many parts of the organisation were involved. If marketing presents the proposal, I would want to know whether it has spoken to customer experience, service, sales, and IT. If that collaboration has not happened before the proposal reaches the executive level, the proposal is not yet ready.

The second question concerns measurement. I would assess success through three dimensions: Does it matter to customers? Does it matter to the brand? And does it matter to employees?

From a financial perspective, there are also two components. Companies need to understand the short-term objective and when they can expect a return. But the more important question is what long-term, sustainable benefit the investment will create for the business.