top of page
Search

AI as the Customer: Designing for Our Most Curious New User

  • sheharav
  • Apr 10
  • 3 min read

We’re used to thinking of AI as a tool —there to help us, automate for us, or co-create with us. But something interesting is happening: AI is stepping into the role of the customer.


From booking services and purchasing products to requesting information and generating creative briefs, AI is beginning to act on behalf of others—as a consumer in its own right.


It’s a quiet shift, but one that could radically change how we design customer experiences, build products, and even define value.

 

🤖 What Does It Mean for AI to Be a Customer?


Today, autonomous AI agents can:

  • Schedule meetings via email

  • Shop for products online

  • Select software or tools for deployment

  • Inquire with customer support bots

  • Review CVs and advise on hiring

  • Summarize content and recommend actions


And this is just the beginning.


As we evolve toward agentic AI, we’ll see systems that evaluate options, make purchasing decisions, and initiate service requests with minimal human intervention.

Which means: your next customer might not be human—but it will still expect a high-quality experience.

 

🧠 The Shift in Design: From Empathy for People to Empathy for Agents?


Traditionally, product and service design has centered around the human experience.


But what happens when AI is navigating your website, your chatbot, or your sales funnel?


We’re entering an era where we need to consider machine empathy too—not because AI has feelings, but because:

  • AI agents need clean, structured information

  • They rely on logic flows, metadata, and promptable interfaces

  • Their “experience” affects the outcomes they create for humans


Designing for AI as the customer means:

  • Clear APIs and documentation

  • Conversational interfaces that are machine-readable

  • Removing ambiguity in UX and communication

  • Ensuring your content is AI-optimizable


In short: digital front doors need to be agent-friendly.


💡 Why It Matters


1. AI agents are intermediaries.

Just like a human assistant might book a hotel for an executive, an AI agent could choose your service over a competitor based on clarity, quality, and integration.


2. B2B and B2C are evolving.

We’re shifting toward A2H—Agents to Humans. Your platform may be judged and selected by AI systems on behalf of a client or end user.


3. Bias and influence are baked in

If your systems aren’t easily “read” by AI, they might be skipped. Worse—biased AI might misinterpret what you offer.We need to think critically about how we present ourselves to AI, not just to people.


🌱 A New Frontier: Designing With AI and For AI


The idea that AI might be your next customer is more than a trend—it’s a call to reimagine the relationship between creation and consumption.


And it builds on something I explored in my last post: The way we interact with AI shapes how AI interacts with us.


If we’re thoughtful, structured, and intentional—we’re not just training better AI models. We’re designing better futures for both human and machine collaborators.

This also highlights the importance of a more diverse range of people interacting with, building, and using AI—so we can better train the AI tools (and customers) of tomorrow.

 

✨ Final Thought


Your next “customer” might not say “please” or “thank you”—But it’s still deciding whether to engage with you.

Design accordingly.

 

🔗 Tools and Links

🧰 Agentic AI & AI-as-Customer Tools


 

📊 AI Optimization & Readiness


 

🔍 Now For Some Practical Suggestions - Where To Start


AI agents aren’t just helping people make decisions—they’re making them.

That means when you're putting in a job application, submitting a proposal, or requesting access to a platform, the first reviewer might not be human. It might be an AI agent acting as the “customer,” assessing fit, filtering content, or deciding what gets passed through.


So where to begin?

  • Audit your digital content: Is it structured, accessible, and clear for both human and machine readers?

  • Review your application materials: Are your CVs, cover letters, and proposals AI-parseable? (Think metadata, clarity, keywords)

  • Evaluate your tone and clarity: Would your site or service make sense to an AI scanning for decision triggers?

  • Explore how AI agents engage: Try tools like ChatGPT, Claude, or AgentGPT to simulate how AI might navigate your content or offerings.


You may not know when AI is visiting.

But it’s best to be ready when it does.

Commentaires

bottom of page