We sat down with Thumbtack’s VP of Commercial Operations, Lin Zhao, to pick his brain about how he vets enterprise AI solutions and the pivotal factors that led him to Maven.
You spent six months evaluating the space that Maven is in. What was happening in the broader AI landscape that made you take the time to go deeper?
There’s this belief that emerging generative AI LLM technology will tremendously benefit the customer experience. Part of that belief is that it will provide accessible, effective support for all customer questions, 24/7. At Thumbtack, we share that vision. We spent quite a bit of time intensely studying the AI space to make sure we picked a solution that could live up to that.
You clearly saw an opportunity in the space, but was there also a problem or pain point you were solving for?
It came down to wanting to debunk negative associations with chatbots. When people hear “chatbot,” they often have a negative reaction because, in the past, bots haven’t been very smart or helpful. Talking to one can feel like a waste of time. You just want to speak to a real person.
But with advances in large language models and conversational AI, things have improved. Bots can now provide more natural, human-like responses and genuinely helpful answers. Plus, they can support customers at any time, no matter how late it is or how long the queue is.
At the end of the day, AI helps us deliver a smoother, more satisfying experience for the homeowners and the small businesses we serve.
How did you decide what really mattered in a sea of AI hype and fast-moving promises? How did you structure your evaluations to stay grounded in your team’s needs?
We started with a large list of over 50 key players in the space and worked closely with our technology team to evaluate them. We had really clear rubrics for both tech integrations and customer experience possibilities.
When we whittled down the lists, we ran multiple demos and had robust conversations to land on the finalists. Then quickly, we conducted a proof of concept using real customer data to determine which solution fit our needs.
What were the hard questions you asked every vendor, and how did those questions evolve as you learned more?
We focused on three main areas. First, we looked at the tech integration capabilities and whether the technology was mature enough for our complex systems.
Second, we assessed how vendors approached the customer experience. Were they just trying to get people off the contact lines as quickly as possible, or did they share our focus on truly helping customers?
Finally, we considered whether their solutions could grow with us and extend beyond support to create new touchpoints across the customer journey.
Being able to deliver a full-funnel experience seems crucial here. Can you describe how you think about the customer journey and which touchpoints you plan to prioritize at Thumbtack?
Many support leaders see the customer journey as starting when someone reaches out for help. At Thumbtack, we see it a little differently. The journey actually begins the moment a customer interacts with our app or platform.
Every engagement, whether it’s a positive experience or a confusing one, is part of that journey. By the time a customer contacts support, they’re already well into this process.
That’s why we focus on supporting their needs from the very first touchpoint, offering guidance, education, and helpful tips throughout their experience with our product. And, of course, if an issue arises, we’re still there to offer quick, effective help right when it’s needed most.
After you completed the evaluation process, how did you decide that Maven was the right partner for Thumbtack?
A few things. First and foremost, Maven shared our vision of using AI agents not just for support, but also for customer success and sales enablement. We knew they would help us enhance the entire customer journey and show up for our customers at every touchpoint.
Second, we love their team. They’re so passionate about the space and so responsive to our questions. They also execute at lightning speed.
Finally, the proof was in the pudding. When we put Maven’s AI agents head-to-head with the top players in the space, the tool consistently delivered the most human-like and effective customer experiences while resolving issues faster. Our internal CSAT assessment showed them to be 15%+ better than everyone else. Knowing they would help us deliver the best outcome for our customers gave us the confidence to move forward.
What advice would you give to other operators who are just starting to evaluate AI solutions?
Don’t just trust the demo. Look under the hood at the vendor’s tech maturity and integration capabilities. If possible, run true comparisons to see how their product performs against others. Several vendors might appear similar in demos, but proof of concept helps you spot clear differences in resolution rates and customer satisfaction.
Also, don’t get distracted by flashiness. It’s tempting to flock to whoever raises the most money or spends the most on marketing. But it’s more important to ensure the vendor’s vision and growth stage align with yours. This is the true indicator of their willingness to form a partnership and adapt their product to meet your evolving needs.
That’s great advice. Looking ahead, what are you most excited about now that all this is in motion?
I'm excited to be able to give our customers the answers they're looking for exactly when they need them, at scale. I also know Maven’s been a great partner in looking at how we may unlock a lot of net new use cases across the funnel, and I can’t wait to see how those play out for us.