BlogMiikka Rahikainen 23.03.2026

Building brand experience in the age of AI: Swap AI tinkering for process blasting

Brand development Marketing development Marketing technologies New technologies

If your company´s AI strategy is for humans to chat with AI, you don’t have an AI strategy yet. Real change only comes about when processes and customer journeys are completely rethought.

Agentic AI changes the rules of the customer experience

Most organizations are still at a stage where artificial intelligence acts as a single tool: the chatbot answers questions, the recommendation engine suggests products, and the content creation tool generates text. Each tool works separately and does not share information with each other. The customer experience remains the sum of separate points.

The next step is agentic AI, i.e. artificial intelligence, which does not wait for a question, but observes, plans and acts independently. It manages entire workflows: it tracks data, makes decisions, takes action, and learns from the results. Humans set goals and boundary conditions, while AI takes care of the implementation.

Change can be perceived through three levels.

• At the first level, AI is a tool. The process remains the same, AI just speeds up parts of it.
• On the second level, AI acts as an assistant. AI personalizes messages, optimizes broadcast times, and segments audiences automatically. The workflow is the same, but the execution is smarter.
• At the third level, AI is an agent. AI autonomously controls the customer journey, detects the signal, anticipates the need, selects the channel and formulates the message.

Simply adding technology is not enough

According to BCG‘s Nordic AI Inflection Point study (2026), Nordic companies spend a disproportionately large share of their AI budgets on ready-made tools such as Microsoft Copilot, ChatGPT, and other general-purpose assistants. These are aimed at gradual productivity gains: an individual employee writes an email faster or receives meeting notes automatically, but the workflow remains the same.

At the same time, global AI pioneers are directing their investments towards renewing entire processes and building new revenue streams.
There is an instructive example of this. Klarna implemented the OpenAI chatbot, which replaced 700 agents and handled 2.3 million conversations in the first month. The resolution time dropped by more than 80 percent, but customer satisfaction dropped sharply. The brand experience suffered because the handling of complex issues deteriorated. The old process remained, and AI only replaced humans without renewing the workflow. Eventually, Klarna hired people back and switched to a hybrid model.

Air India, on the other hand, built its AI solution around customer service from the very beginning. The process was designed on top of AI, and escalation to humans is a seamless part of the whole. The result is a 97 percent success rate in customer surveys and only three percent of conversations are transferred to a human agent. Both efficiency and customer experience improved at the same time, because AI did not replace the old one, but was a completely new process.

Actionable data is the foundation of everything

Agentic AI doesn’t work without high-quality data. Without unified customer data, the agent does not personalize but automates guesses.

Effective customer understanding is created from three layers:

1. Data collection: First-party and zero-party data form the foundation. They arise directly from the customer relationship and genuinely tell about the customer’s actions and choices.
2. Identity aggregation: Signals generated from different channels and devices are combined into a single profile so that personalization doesn’t remain channel-specific.
3. Profile enrichment: The profile is continuously enriched with attributes related to behavior, value, and purchase probability. Only then can marketing anticipate and schedule actions more accurately.

A prerequisite for all of this is effective consent management. Without a valid consent template, data cannot be collected, combined or utilised. Consent management is not just a legal requirement, but the foundation of the entire data strategy: if the customer has not given their consent, the agent AI has nothing to learn from.

Four guidelines for the AI age

1. AI change is a change in processes. It is not enough to just implement tools, but customer journeys and workflows must be rethought.
2. Actionable data is a prerequisite for agentic AI. Without high-quality, connected customer data, AI cannot personalize meaningfully.
3. Continuous work supports AI and automation. Agentic AI doesn’t benefit from individual campaigns that start and end. It needs a constant flow: constant content, constant data and continuous optimization. When marketing is continuous, automations can learn from every encounter and improve results over time. Instead of individual campaign spurts, we need systematic development work in which we learn and unlearn quickly.
4. AI requires rules of the game and new indicators. As AI guides entire customer journeys, measurement must also change. Channel-specific conversions no longer tell the truth, because a single conversion can be the result of multiple channels and touchpoints. The focus of the metrics shifts towards the customer’s lifecycle value (CLV), customer retention and the profitability of the entire customer relationship. Only with these indicators can we see whether automation genuinely produces better business or just more measures.

The brand experience is not created by individual tools, but by holistic thinking that combines technology, data and human understanding.

Do you want to build a customer experience where AI and data work together? Whether it’s renewing customer journeys, building a data strategy, implementing AI automations, or developing metrics, don’t hesitate to contact us!

Author

Miikka Rahikainen

Martech Director

miikka.rahikainen@dagmar.fi

+358415062689