MarTech360 Interview with YouMon Tsang, Founder and CEO at ChurnZero

"AI should ultimately make your customer team more human. It should take away the work that machines are increasingly good at, so your people have more room for judgment, relationships and the situations where getting it wrong really matters."

YouMon Tsang, can you tell us about your professional background and your current role at ChurnZero?

I’m a repeat entrepreneur, with ChurnZero being my fourth company. For more than three decades, I’ve been building and operating software businesses across product, marketing, and general management. A big part of that career was in marketing technology. Before ChurnZero, I was a CMO and later a business unit leader at Vocus, where I got a firsthand look at how much technology companies had built around acquiring customers compared with how little they had built around keeping and growing them. My time in that role was what inspired me to build ChurnZero.

I’m now the co-founder and CEO of ChurnZero, the platform and partner for customer growth. The teams we work with and support own the customer after the sale: onboarding, adoption, renewal, expansion, advocacy. What we do is connect all of that. Your customer data, your systems, the way your team already works and AI built specifically for customer success sitting on top of it.

You’ve spent more than three decades moving between product leadership, entrepreneurship, marketing technology, and building software companies. When you look across that journey today, is there a particular problem or pattern you realize you’ve been trying to solve in different forms all along?

There’s definitely been a theme to my career: how to help people make good use of information.

Early in my career, my focus was all about data. The first problem was how to access locked-up data. Then it evolved to, okay, having access to data is interesting, but you still have to manipulate it and turn it into analytics, summaries and things you can actually make business decisions on. Then the problem became: how are you actually going to act on the data? With ChurnZero, my focus moved from accessing the data to analytics, to taking action on it and creating the workflows to make that happen.

With AI, the next step is possible: to just do all of that automatically. By putting the structure in place, AI and automation can take that next step on your behalf, and you can feel confident about it.

Also Read: MarTech360 Interview with Cobus Greyling, Chief Evangelist at Kore.ai

Your experience at Vocus has been particularly important to the creation of ChurnZero. You had access to increasingly sophisticated intelligence about prospects before they became customers, yet the post-sale side had far less visibility into customer behavior and health. What did that contrast teach you about where software companies were fundamentally underinvesting?

The contrast was stark. When I was CMO at Vocus, I had access to unbelievable technology. Sales had great technology, marketing had great technology, finance had great technology. But there were places where the tech really wasn’t there yet, and customer success was certainly one of them.

In retrospect, it’s easy to understand how companies got there. When companies start, we all worship at the altar of new sales. You have to get new sales. You have to prove somebody wants to buy your product. That goal becomes so ingrained that companies can lose perspective as they mature. At some point, you have to shift the investment because you’ve created an asset that’s even more important: your customer base. So for me, it wasn’t a hard step to say that we have to bring the same level of sophistication to the people taking care of customers.

The economics of SaaS have only made that question more important. ChurnZero’s research found that 74% of leaders say most of their company’s revenue now comes from existing customers. From your perspective, when does customer success stop being a post-sale function and become part of the company’s core growth strategy?

For most companies, customer success becomes a growth function much sooner than they realize. In the early days of starting a business, almost all of your energy goes toward proving there’s product-market fit and winning new customers. But if you carry that mindset too far, you can keep operating as if growth only happens before the contract is signed, even when a meaningful share of your future revenue is already sitting in the customer base.

Investors also care deeply about retention now because growth and retention are inseparable in a subscription business. If you’re constantly replacing revenue that walked out the door, it’s very hard to build durable growth on top of it.

I wouldn’t tie the shift from post-sale to core growth strategy to a certain company size or stage. It starts with the first customers. Are they adopting? Are they getting value? Are they renewing? Is there room to grow the relationship? Once you start looking at Customer Success through those questions, it stops being the team that manages what happens after the sale and becomes part of how the company grows.

Customer success is being given more commercial responsibility. Your 2026 research found that 83% of CSMs now have some revenue responsibility, while only 60% feel prepared to own it. What do you think companies are getting wrong when they ask Customer Success to become more revenue-oriented, and what would a healthier version of that transition look like?

First, I’d just say that it’s really good to see the percentage of teams owning revenue getting higher. An empowered CS team is instinctively more capable, more resilient and more valued within the organization. Historically, owning a share of that revenue has been one of the differences between having power and not having it within a company.

But the shift is tough. Many customer success teams started without owning a number, and now companies are asking the function to own revenue. The obvious mistake, and we all make it, is assuming that all the same people can simply do that new job.

There are people who are great at being customer-facing. They want to make the customer happy. That’s what they’re really good at. The commercial side can be very challenging for some of them. So when you make that transition, you have to ask: are you training your people to do the new job, or are you changing the mix of talent in the organization to put the team in a position to succeed? If people don’t feel prepared to take on that ownership, you have a training and talent-mix problem. The good news is that both are fixable.

You’ve spent years watching software evolve from tools that helped people manage information to systems that can increasingly act on it. With ChurnZero now moving toward AI agents that can detect signals, use business context, and execute workflows, where do you think the boundary should sit between what an AI agent can decide for itself and what should always remain a human judgment?

I think that boundary is going to keep moving. There’s already a class of work we’re going to stop worrying about, like transcribing and summarizing meetings, gathering information, enriching account data, finding signals. We’re going to increasingly accept that AI can do those things extremely well.

But as the models get better, we’re also going to ask them to do harder and more important things. So I actually don’t think human involvement necessarily goes down. The role of the human just moves up the stack to oversee more complex work.

AI can build the picture, surface risk and opportunity, draft the work and automate workflows. But for many customers, you still need a person to make the call, be the strategic driver and advise. The CSM role is too relationship-dependent to simply hand over, and AI should ultimately make your customer team more human. It should take away the work that machines are increasingly good at, so your people have more room for judgment, relationships and the situations where getting it wrong really matters.

There is an interesting shift happening here. If AI can eventually handle much of the research, monitoring, summarization, follow-through, and routine intervention that consumes a CSM’s time, the question may no longer be how to make CSMs more productive, but what the CSM role should become. What do you think the best Customer Success professionals will be uniquely responsible for five years from now?

In the end, I think CSMs finally get to do what we’ve always hired them to do. From the beginning, companies have wanted customer success to be a partner to the customer. A subject matter expert, a consultative partner, a strategic advisor. Those are the words we’ve always used to describe great CSMs.

But CSMs rarely have the opportunity to spend most of their time in those roles. Instead, they’re chasing down issues, coordinating and summarizing meetings, gathering information, and herding cats.

AI promises to solve many of those tasks. So if you want to know what the best CSMs will be responsible for five years from now, just pull the same job description you wrote 10 years ago, five years ago or even last week. That’s the job they’ll be doing because they finally have the capacity to do it.

One of the strongest ideas behind ChurnZero’s current AI strategy is that intelligence without context is not enough. An agent needs to understand the customer, the business, the company’s knowledge, and the desired outcome before it can act responsibly. As foundation models become increasingly accessible to everyone, do you see proprietary context and customer intelligence becoming the real competitive moat in enterprise AI?

Absolutely. Every vendor now has access to roughly the same models, so the model can’t really be the differentiator anymore. Context will be what makes your product competitive. Generic AI knows a tremendous amount about the world, but it doesn’t know your customers. It doesn’t know your product, how your company operates, what this particular customer is trying to accomplish, what they said on the last call, what their usage looks like or how you define risk and opportunity.

ChurnZero provides AI with a system of context. You need the account data, product usage, emails, call transcripts, surveys, company-level insights and your own knowledge brought together. Without all those details, you get generic next-best actions that sound reasonable but don’t really help anybody.

So when everyone has access to the same underlying intelligence, the advantage becomes the context you can put around it and your ability to make that context actionable. The model is important, obviously, but the model by itself doesn’t know your business.

You’ve now built multiple companies across very different technology cycles, from the early internet era through SaaS and into the agentic AI era. What feels genuinely different about this AI moment compared with the technology shifts you experienced earlier in your career? What part of the current excitement do you think the market is getting wrong?

When the AI transformation started in 2022, I felt like I had a handle on it. I had survived the dot-com boom, the cloud transformation, and the mobile transformation, and I felt those experiences had prepared me to lead confidently through another major technology shift.

What has felt different over the last year is that suddenly things feel unknowable. With most transformations, you can imagine drawing a line on a chart, probably up and to the right, and you can see the milestones along that line. You may fail because you don’t execute well enough, but you basically understand where the line is going. With AI, I’m no longer sure where that line is moving or even what direction it’s heading. The uncertainty changes how you lead. You have to be more agile, have really good focus, expect change and get comfortable dealing with the unknowable.

I think we’re getting things wrong by overpoliticizing AI. Being suspicious of it or uncomfortable with it is a normal, credible reaction. But I’d rather see the debate center on technology and safety than turn AI itself into a political belief.

If you look beyond ChurnZero itself and imagine the Customer Success industry ten years from now, what is the assumption about customers, software, or human relationships that you think we are most likely to get wrong today?

With the caveat that everything is unknowable, I think we’ll be surprised in both directions. We’ll be surprised by where the technology can take us. AI will probably do even more than we expect, and things that feel like they require a human today may become perfectly normal for technology to handle. And I think we’ll also be surprised by how important the human remains in a business relationship. The amount of human involvement may not disappear nearly as much as people assume. It may just move to higher-impact moments and harder problems.

Thanks YouMon!