“We Are the Communications Infrastructure of the AI Era” — How Twilio Is Shaping the Future of Customer Communication

Angelique Szameitat: Chris, what is Twilio in the age of AI?

Chris Koehler: We believe that we are the communications infrastructure of the AI era. Today, companies no longer only have to communicate with customers, but increasingly also with AI agents. Our focus is on connecting this communication across all channels and enriching it with context, persistence, and intelligence.

Angelique Szameitat: What exactly does that mean?

Chris Koehler: The central problem is that conversations between brands and consumers are still very fragmented today. Marketing, service, and sales each conduct their own dialogues with the customer, but these conversations are often not connected. That creates a broken customer experience. We want to ensure that companies understand a customer’s entire communication history and can use it intelligently in the next interaction.

Customer Memory Instead of Channel Silos

Angelique Szameitat: You speak of “Customer Memory.” What is behind that?

Chris Koehler: Customer Memory means that all of a customer’s conversations with a brand are brought together across channels. That way, the next conversation can build on the previous one. The customer does not have to start from scratch every time, and companies can communicate far more consistently and relevantly.

Angelique Szameitat: What role does orchestration play in this?

Chris Koehler: A very large one. For a long time, the principle in marketing was: the right message at the right time in the right channel. But the “right channel” is not necessarily the one the company prefers — it is the one the customer prefers. One person wants to communicate via WhatsApp, another via email, another prefers voice. Companies will need to orchestrate these preferences intelligently in the future.

AI as a Productivity Engine — Internally and Externally

Angelique Szameitat: What does AI mean internally for Twilio?

Chris Koehler: We see AI as an accelerator for everything we do. First, it was about personal productivity: employees become faster, more efficient, and more capable. Then comes team productivity: better collaboration, more efficiency, less friction. And the third phase is the real transformation: we are completely rethinking our customer workflows.

Angelique Szameitat: Do you have a concrete example?

Chris Koehler: Yes. Internally, we have an agent named “Issa.” This agent now handles more than 90 percent of all incoming inquiries. In the past, there were capacity limits because people could not respond to unlimited inquiries at the same time. Today, the agent can communicate continuously with prospects and customers. If a human needs to take over, the entire conversation to date is already documented. That significantly improves the customer experience.

Angelique Szameitat: How large is the time gain from such solutions?

Chris Koehler: Enormous. In the past, availability, time zones, and language barriers were limiting factors. Today, inquiries can be handled immediately. That saves thousands of hours of productive time and, above all, significantly reduces the time to the first relevant response.

Content Creation: No Longer Starting from Scratch

Angelique Szameitat: Which marketing tasks should no longer be done completely manually this year?

Chris Koehler: Above all, content creation. Whether text or visuals, no one should still be starting from a blank screen today. AI should provide the first draft. Humans remain important, however, to ensure tone, context, brand identity, and quality.

Angelique Szameitat: So humans remain in the loop?

Chris Koehler: Absolutely. For the foreseeable future, we need people to ask: Does this sound right? Does it fit the brand? Is the message nuanced enough? AI can prepare a great deal, but the fine control remains human.

Angelique Szameitat: Where do you currently see the greatest productivity gains in marketing?

Chris Koehler: Clearly in content creation. At the same time, it is also a risk area. If content becomes too easy to produce, there is a danger that everything will start to sound the same. That is why we are thinking very intensively about how to preserve our voice, our identity, and our brand character.

Angelique Szameitat: So is there a risk of a kind of “AI sameness”?

Chris Koehler: Exactly — that is my concern. Just because we can produce more content does not automatically mean that this is good for marketing. If everything becomes interchangeable, that ultimately harms the discipline.

Trust Becomes a Strategic Currency

Angelique Szameitat: How does AI change consumer behavior from your perspective?

Chris Koehler: Very significantly. We see this especially in search. In the past, users worked with keywords; today, interaction is becoming increasingly conversational. People simply say what they need and expect a system to provide the right answer. That changes behavior radically — and marketing must adapt.

Angelique Szameitat: Personalization has been considered the supreme discipline in marketing for years. Why is it still so difficult?

Chris Koehler: Because several things have to come together: clean data, the right consents, regulatory certainty, and the ability to derive truly relevant content from that data. In practice, that is very difficult. And complexity is increasing because there are ever more rules and requirements internationally.

Angelique Szameitat: How well are companies managing this balancing act today?

Chris Koehler: Some are already doing it well, but overall the industry is still not where it should be. Many companies have data, but do not use it consistently because the foundation is not clean enough. As an industry, we are still not truly good at personalization.

Angelique Szameitat: Are consumers still willing to share their data?

Chris Koehler: Yes — if they clearly see the added value. The problem is that customers often do not feel their data is being used meaningfully. That creates frustration. If companies use data, they should use it to create real relevance and real value.

Angelique Szameitat: Another issue is trust. AI also makes fraud and scams more believable. How big is that challenge?

Chris Koehler: Very big. AI will make it easier for fraudsters to create credible deceptions. If consumers eventually no longer know what is real and what is not, we lose trust on a broad scale. That is why, from our perspective, compliance, security, and responsible platforms are absolutely essential.

Angelique Szameitat: What exactly is Twilio presenting at the SIGNAL Conference?

Chris Koehler: We are talking about our new Conversations platform and, above all, about three things: Customer Memory, Orchestration, and Intelligence. In other words: How do we create context across all conversations? How do we deliver communication in the customer’s preferred channel? And how do we use conversation data intelligently to improve the next interaction? We are also talking about messaging and EU data residency, as well as concrete customer examples.

Angelique Szameitat: Are there use cases that are especially relevant for retail and brands?

Chris Koehler: Yes, absolutely. One exciting area is voice in customer service. For example, we are talking with Otto about how AI can support incoming calls, capture information, and route customers to the right place. Equally interesting is the question of how outbound AI agents can support sales before being seamlessly handed over to a human. In retail overall, we see many use cases around personalized, cross-channel communication.

Retail: The Biggest Mistake Is Silo Thinking

Angelique Szameitat: Retailers have been talking about omnichannel for years. What are they still getting wrong from your point of view?

Chris Koehler: The biggest mistake is silos. Companies often believe they are omnichannel because marketing is active on multiple channels. But from the customer’s point of view, omnichannel means every interaction with a brand — including service, support, and sales. As long as those areas are not connected, the experience remains incomplete.

Angelique Szameitat: If we look toward 2028: which marketing activities will largely be automated by then?

Chris Koehler: I believe many content and production workflows will be much more heavily automated. At the same time, we will still need humans — not necessarily for every single step, but for strategy, quality control, brand management, and defining the right guardrails.

Angelique Szameitat: So will the marketer of the future be more of an orchestrator than a specialist?

Chris Koehler: I think we will need both. Today, many marketing organizations are heavily shaped by specialists. In the future, there will likely be more people who orchestrate end-to-end workflows and use AI in a targeted way. Specialists will remain important, but their role may shift more toward oversight, expertise, and quality.

The Most Important Advice for Marketing Leaders

Angelique Szameitat: What one piece of advice would you give marketing leaders for the next twelve months?

Chris Koehler: The biggest opportunity lies in not merely optimizing what already exists. My advice would be: don’t just make small improvements to the old way of working — throw it overboard and truly think from a blank page. Ask yourself: if I were setting up this task today with AI and modern technology, how would I design it? That is the only way real change happens, rather than merely incremental change.

Angelique Szameitat: So not just improve — but rethink?

Chris Koehler: Exactly. If you only optimize the existing process, the result will also only improve step by step. The bigger lever comes from fundamentally redesigning processes. The best approach is to start small, for example with a campaign workflow, learn from it, and then transfer that learning to other areas.

Angelique Szameitat: Do you see differences between Europe, Asia, and the USA in how AI is being approached?

Chris Koehler: To be honest, I see a great deal of curiosity and optimism worldwide. The differences are less between countries than between companies and industries. In more heavily regulated sectors, companies are naturally more cautious. In retail or in less regulated environments, testing and implementation often move faster.