Hajra Naz
Having led a business for nearly 20 years, I’ve seen more product launches than I can count. And yet, every single one teaches me something new. But our most recent launch was unlike anything we’d done before. We stepped into uncharted territory with the release of our first AI agents.
This wasn’t a snap decision. Our journey into agentic AI began back in 2023—long before “AI agent” became a buzzword. Back then, we were brainstorming how these autonomous tools could simplify life for our customers. Two years later, AI agents aren’t just a concept—they’re a force transforming how businesses operate.
The hype has been deafening. From bold claims about AI replacing entire industries to warnings of a dystopian takeover, the AI agent debate has been intense. No matter your view, one thing is clear. AI agents are changing the game. More leaders want to integrate them.
Read More: Agentic AI vs. AI Agents: The Truth You Must Know
Here are the biggest lessons we learned.
1. Listen to Data, Not Hype
Every great product evolves between idea and launch. Our first step into AI came during one of our regular hack weeks, when our team pitched an idea for AI-powered forms that users could create and customize simply by chatting with the form itself.
We quickly rolled out a beta version and put it in users’ hands. The adoption rate was promising—engagement was high, and retention looked solid.
Then the data surprised us. Ninety percent of users weren’t using the feature for forms at all. Instead, they were using it for customer service inquiries.
It wasn’t what we’d planned, but it was what the market wanted. So, we pivoted. We optimized the agent for customer service and integrated it into our support ecosystem.
Stat to note: A 2024 McKinsey survey shared a clear insight. Companies that adapt products early see better results. Using user data quickly boosts long-term success. They are 60% more likely to keep adoption rates high after launch.
The takeaway: Be willing to throw your roadmap out the window if the numbers tell a different story. Hype may get attention, but data will keep your product alive.
2. Prioritize Efficiency Over Friendliness
The rise of AI companions has made headlines, sparking debates over replacing human relationships with machines. That might tempt businesses to believe their AI agents need charming personalities to succeed.
We found the opposite. Users don’t want their agents to be their “friend.” They want them to be fast, accurate, and available.
In our testing, customers preferred agents to human reps in certain situations—not just because they were available 24/7, but because they could ask questions without feeling judged. They didn’t hesitate to repeat themselves or ask “basic” questions.
And here’s an important nuance: too much human-like personality can backfire. Research from Salesforce found that when agents are overly human in tone, customers expect them to perform like humans. If they fail, trust in the brand drops.
As Yvonne Gando, Senior Director of UX at Salesforce, puts it:
“It’s more effective to keep AI interactions helpful and approachable without creating a false sense of human connection.”
The takeaway: Keep your AI agents helpful, clear, and approachable—but don’t pretend they’re human. Customers value efficiency over charm.
3. Meet Customers Where They Are
One of our biggest wins came from making our AI agents available everywhere our customers already were—on our website, through WhatsApp, over phone lines, and in multiple languages.
The multi-channel approach removed friction. Customers could reach us anytime, from anywhere, without needing to learn a new platform. And because AI agents never sleep, our global users got the same experience whether they contacted us at 2 p.m. or 2 a.m.
Stat to note: Zendesk’s 2024 CX Trends Report revealed a shift. Seventy percent of customers now expect real-time support. They want it across multiple channels. Sixty-three percent say they’re more likely to return to brands that offer it.
The takeaway: Accessibility is just as important as capability. An AI agent that’s hard to reach isn’t much use.
Read More: Are AI Agents the Future of Customer Service? 7 Facts You Must Know
Final Thoughts
AI agents are not just another tech fad. They’re transforming how companies serve customers, solve problems, and scale operations. Our journey taught us that success depends on:
-
Listening to real-world data, not just industry hype
-
Designing for speed and utility, not artificial charm
-
Integrating across channels so customers can connect effortlessly
And perhaps the biggest lesson? Stay adaptable. The AI landscape is evolving daily, and so must the products built for it. As with any innovation, flexibility isn’t just a bonus—it’s a survival skill.
