Customer service in the AI revolution: Why human connection still wins

Two months ago, a scooter died on me mid-ride. Twenty minutes later, I was still hunting for a way to contact support. No chat button, no phone number, nothing. So, I gave up and paid for a ride that didn't work.

‍

Last month, my milk frother stopped working within warranty. Amazon told me to contact the manufacturer, the manufacturer's phone support told me to email, and their email team told me to call. I'm still in limbo weeks later.

‍

These experiences aren't unique; they're everywhere. And they've opened my eyes to something most tech companies miss: while everyone is rushing to automate customer service, the companies that maintain human connection are going to dominate their markets.

‍

Here's what we've learned about real support

‍

Most companies treat customer support as something to outsource or automate away. But when a company does things differently, like getting engineers to handle support directly, something powerful happens: they understand their users better, learn from real problems, and make smarter product decisions.

‍

I'm an Analytics Engineering Advocate at Lightdash. A few years ago I didn't see myself in this kind of role, but being involved in support has been surprisingly challenging. Being closer to customer problems has made me more product-focused and a better engineer. The thing is, you can't understand your customers from analytics alone. You need to experience the frustration when data won't sync, or talk to someone struggling with a slow query while trying to meet a tight deadline.

‍

At Lightdash, we don't have a dedicated support team. Instead, engineers rotate on support duty every week. Sometimes you might even get an answer from our CEO or CTO. This approach brings customers closer to someone who can actually fix the problem. When you tell us something isn't working as expected, the person helping you may just be the person who built the feature you're asking about. That's why we can often fix problems immediately.

‍

I mean, the results speak for themselves. In the last week:

• Our median response time was 5 minutes during business hours
• 90% of requests were answered within 66 minutes
• 93% of issues were solved within a week

‍

This rotation system means everyone gets the chance to understand the customer more. We've discovered that customer support is where you stay connected to what actually matters, making it one of our most valuable channels. You won't get automated responses when you reach out to us. You'll get someone who understands your specific setup and solves the real problem, not just the symptom.

‍

It's one of the reasons teams switch to Lightdash.

‍

But what about scaling?

‍

At this point you're probably thinking, "Ok, that works because you're a small company, but what about scaling?"

‍

Here's the thing: we're redefining how support works as we scale. The pressure to replace support teams with chatbots and LLMs becomes intense as companies scale. It's understandable because human support doesn't scale linearly, and automated responses seem like the obvious solution.

‍

At Lightdash, we embrace AI where it makes us better. We use AI-assisted development to ship features faster. Our AI agents help users get insights from their data in natural language. Our docs site also has an "Ask AI" button.

‍

But we draw the line at replacing human support with automation. While others are rushing to cut costs with AI agents, we're doubling down on human connection because we know it works. We use AI and tools to make ourselves more efficient, but support is fundamentally about listening, and we want to hear what our customers have to say because that feedback is gold.

‍

As we grow, we know that maintaining engineer-led support will get harder. But that's a problem we're determined to figure out because we can't lose what makes our support valuable.

‍

Why this gives us an edge

‍

We're competing with much larger BI companies that have been around for decades. We don't have everything they have, but we have something they've lost: the ability to move fast when customers need us to. We build what customers ask for, and we build it fast. When a feature request comes up repeatedly, we can do something about it in weeks, not in a quarter.

‍

This approach also delivers better business outcomes for our customers:

• Faster implementations. No waiting around for weeks to get help with your setup.
• Higher success rates. Customers get help from engineers who built the features.
• Product improvements. Feedback reached the people who can act on it fast.

‍

And we help everyone, not just paying customers. Lightdash is open source, so we help our community of free users too.

‍

The real opportunity in the AI era

‍

Think about the last time you had a great support experience with a company. You probably remember it because experiences like that are, unfortunately, rare.

‍

Here's what we've realized: the real win isn't replacing humans with AI. The real win is leveraging AI tools to improve processes while maintaining human connection. If we automate away all our customer support, we're missing out on vital insights for product opportunities.

‍

The companies that win are those that make customers feel heard and helped, not those with a comprehensive list of automated responses.  At Lightdash, we're proving this approach works every day, and I'm proud to be part of this winning team.

‍

Experience the difference yourself

‍

Ready to see what real support looks like? Try Lightdash and experience the difference human-centred support makes. Whether you're frustrated with unresponsive vendors, tired of chasing chatbots, or just want to work with a team that actually listens, we're here to help.

‍

Start your free trial and get your questions answered by an expert.

‍