AI and the Future of Customer Success with Bill Tabbit-Humphrey

Hey everyone, we're back here in Orlando covering Inspire 2026. My next guest is Bill Tabbot Humphrey. I got to be honest with you, by the time you guys have watched this video, Bill and I have already spoken for about 45 minutes about everything and anything.

So now we're going to get around to doing our interview here, and I hope you enjoy it. Bill is the chief customer officer, and if you've been watching any of our coverage here, you saw we had the chief marketing officer, the chief revenue officer, the chief executive officer, but I always like to speak to a chief customer officer because they're the people who are really hearing it firsthand- Yep ... from customers.

No filter, real deal. Bill, welcome to Techstrong TV. It's great to have you on here.

Thank you, Alan, I really appreciate the time. So look, you've had a bit of a storied career. Who was it?

Branch Rickey said luck is 80% preparation and 20% inspiration. But you've been lucky enough to have had a string of companies that have had tremendous success. Yes, sir.

IPOs, exits, and so forth. A bit of a rabbit's tail, right? It has been.

Give people a sense of where you've been and what you've done. Well, and I think I mentioned to you too, I grew up in a really small town in Pennsylvania called Sweet Valley, but I started my career as a COBOL developer. Really?

So I wrote COBOL back in the early '90s- Great ... for EDS, started doing the developer route. I thought that was the way to go, and a VP of engineering I had once said I was too personable and I belong in sales.

But I'm like, "You're kidding, right? " I'm like- Oh my God, they're banishing me ... " Yeah, and he said, no.

He's like, "You bring a technical depth to a customer conversation in a relatable fashion that- And that is invaluable. Yeah. I will tell you, as someone who's also been in tech a long time.

It's like normally AEs and SEs are just really technical focused or really just close the deal. Mm. So I ended up doing that, and then I spent seven years, eight years with Tableau.

Helped them build up from one-person services to $100 million in services a year. Wow. Built out the customer experience part with them, working on education services and customer experience.

Then I moved over to Informatica, so big culture shift. Yep. From very business user-focused over to IT-focused- Mm-hmm ...

and now you're looking at big enterprise platforms and multiple facets on that. Helped shift that organization from an on-premise to a cloud-based, and then also move from a one-time revenue model to recurring revenue. Revenue.

And then at 18 months ago, Alteryx came, and we had a great conversation with the board here, and lo and behold, I made the move over here, and now I run the whole customer experience organization here, which is everything from when the ink signs to customer realizing value and making sure that they get the most and the best ROI out of the platform as quickly as possible. And it's interesting how you described what customer experience is, because I've spoken to some execs who are chief customer officers and so forth, and for them it's more about help desk and tech support and that kind of thing. But I think that misses it.

That's like the complaint department is upstairs. Yeah. You know what I mean?

Remember you used to see those signs? It was a different finger they would use, but- ... the complaint department is upstairs.

It should be about customer success, not about fielding customer complaints. If you're doing your job right, you get less complaints and more partnership, more input. It's more proactive.

And I'll tell you, coming to Alteryx, a lot of good intended people through the changes. Mm-hmm. But it was a little bit like a Whac-A-Mole.

Yeah. And I had to shift a little bit of the culture. Everybody sells.

Every time you talk to somebody, you're trying to help them realize the value. ABC. You don't need to sell them software.

Mm. But I do need to sell them on other use cases that they have. And so, it's a lot of an evolution that I've been working on with the team, but support, they're a lead generator.

They're proud of that fact and closing the leads that come out of that. Our services folks engage with retention for our customers. It really is about that proactive outreach.

Mm-hmm. In a world where, I'm sure you've heard it before, the intelligent enterprise- Sure ... things are changing so fast, and if our only responsibility is to whack the nail on the head, we're not doing our customers any service.

They need a trusted advisor on what are you seeing in the market, what are our competitors doing, what are our peers doing? Well, that's what they're looking to you for. Yes.

It's funny, I spent most of my career, Bill, on the vendor side, right? I co-founded a bunch of venture-backed companies. And then, 12, 13 years ago now, I moved to this side of the camera, and I got a chance to talk to a lot of people.

And the misconception over how to deal with customers, right? Because a lot of people say, "Ah, he's a vendor. " Well, he's the vendor who's talking to a thousand people like you, so he doesn't just hear your use case.

He's hearing a thousand- A thousand, yeah ... use cases and picking out the trends and what works and what doesn't work. Who better to know that?

It's interesting you say that. I finally qualified for the coveted ConciergeKey with American Airlines this year. Really?

Good for you. I've been on the road every week this year. We're now just the end of May.

Visiting somewhere between three and 12 customers in a week. And what I found along that journey is thatCustomers don't want you to just go, "Yes, Alan. Yes, Alan.

" Yeah. No, they don't want that. "Alan, how you're thinking about that probably isn't the best way.

It is a way, but have you thought about this? Have you thought about that? " I'm sorry I touched the mic, sorry.

I do it all the time. Don't worry. How am I using AI?

How are my competitors using AI? How is data playing into that? Who do I empower?

Where do I set up enablement organizations? What's the COE? Why is it important?

I do. Those are the things that customers, they hear enough from marketing and enough from sales about what they think they want to hear. They really want to know the beat on the street.

Yeah. No, they want the real deal. Yeah.

It's the real deal. I want to pivot a little bit. Okay.

We've been talking a lot today, as I mentioned, with a lot of the exec team here, a lot of some customers. AI is top of mind. Yeah.

It's game-changing. I have a small company here at Techstrong, but even for our little group, it's game-changing. Mm-hmm.

You look at a company like Alteryx and what you guys do, it truly turns things on its head. 18 months ago, you started changing the culture. How has AI and agentic AI caused you to maybe rethink what turning on its head means here?

What are you doing around that? So let me give you some perspective, at least from inside of my four walls, because customers love to hear what I'm doing directly with the platform, with the product in hand. 18 months ago, we had a small sliver of AI, and we were predominantly focused inside of our technical support organization.

Great. It was predominantly around case deflection, helping customers- Yeah, chatbot stuff. Yeah.

And we were doing probably 7, 8% case deflection. Good, not amazing. But as all the boards now are saying, "We want you to be more efficient," many companies are trying to figure out how do you grow 30% but only grow 10% more cost?

Or how do you grow 20% with no cost in that process? So we took a hard look inside of my organization, and we've gone from two or three to 36 different use cases. We went from that 7, 8% deflection to almost 40% deflection.

Really? Great. That's great.

And repurposing those support folks into helping customers realize value faster with the product. This is not about shrinking teams. No.

This is making them more efficient and getting them more in front of our customers. Our CSMs used to take weeks to build out QBR decks. They pull the data out of the system, spit it into a PowerPoint creator.

Right. And it spits it out, and then they spend 30 minutes just reviewing it and putting the customer logo on it. That was a day, two days, three days worth of work.

Two, three weeks. I do the same thing. And so those are the kind of things, and one of the things that was powerful with Alteryx is the platform is so business-focused, any user can do it.

It's a renewals person, it's a success, it's a consultant, it's a support person. Every month we do an ant hackathon, for lack of a better word, and we have them figure out what makes them more productive and start making those automated. So we took the QBRs, for example, and rolled that into our portal for all of our customers to get a quarterly update on their tickets and their usage and their volumes.

That wasn't even happening before. And so all of that was built with Alteryx- Right ... on top of our data system, and it's powerful.

It's drinking your own champagne, eating your own dog food, whatever you want to call it. Which I think any good marketer out there would love to admit they're doing, yeah. But it makes all the sense in the world when you think about what that means.

And like you said, it doesn't necessarily mean a reduction in headcount. Yep. It's about being 4X, 5X, 10X more involved in your customer's experience, making it better, hopefully.

We're growing this year, and I didn't ask for a single headcount. Really? Not one.

That's- So we're growing double digits. What a great story. And I'm like: I don't need them.

" No. I'm making my CSMs more productive. I'm helping my support folks be more productive, getting the mundane out of the business.

That's the promise. The employees are happier because they love customers, and so you're giving them more of that and less... Well, and they're less complaint takers and more, as you mentioned, proactive helpers.

Yep. We've heard from a few folks that I interviewed today about Annie and Adam and Atlas. Yep.

And they all start with A. I don't know what the- I think it's an Alteryx thing. It's an Alteryx A thing?

I think so, yeah. Okay. I got to imagine that that's a tremendous help when you talk about deflecting cases and everything else.

These have really got to-- Do they live in your organization primarily, or? So Adam does. Okay.

Annie lives with sales and marketing. Mm-hmm. Because it's really more of a lead generation- Sure ...

kind of get somebody involved. The next thing we're working on with those two is how can you talk to Annie, and Annie will pass you right over to Adam. Without having to rehash everything, your name again, everything.

Adam can get involved and talk about the use cases and talk about how you implement and how you onboard them. So it's kind of like taking Annie and Adam on the road as a duo, huh? Yes.

Yeah. So no, however, I will say one of the things with AI, it sort of shifts the paradigm a little bit. We talk a little about developers, like when we spend 70 to 90% of your time building and 10% testing.

Now you're spending 20 to 30% building and 70 to 80% testing and iterating. A lot of people start their project, and it sort of falls off because there's a data element, there's a data management element. Mm-hmm.

And I think that's the piece that my team is running for all the different agents that we're running in now is making sure that the data is relevant, the data is populated. We keep updating it with every new release in that system. We have a use case repository that literally has tens of thousands of use cases in it that we feed into Adam.

" And it spits out five different use cases. Well, this isThis addresses the next thing I wanted to bring up with you, and it's around this issue of trust for AI. Yep.

Right? AI's great with coming up with ideas. Mm-hmm.

And, when you tell it, "Oh, this isn't right," it's very good at saying, "Oh, of course it's not right. My bad. I'm sorry.

" But the fact is, actually, I was talking about it with Andy right before you came in. We seem to hold AI to a higher standard than we hold even humans. Yep.

Right? Humans make mistakes. We're human.

Yep. AI makes a mistake, I'll never trust it again, right? That's the kind of place where we're at right now, it seems.

And when I say AI makes a mistake, it said something wrong or whatever, right? Knowing that that's the reality we're dealing with, how does that factor into how you're using AI in your customer solution? That's a great question.

And you think about the evolution of those bots, those agents- Mm-hmm ... those super humans, for lack of a better word. They start off with only providing information that you feed it.

Mm-hmm. So, Adam, for example, only knows what we told it. We told it how to do upgrades.

We told it how to onboard with Alteryx when we told it the use cases. We also gave it permission to say, "I'm not sure of that answer. " That's one thing that a lot of those agents and bots don't put in there, is that if it's not in the system, the default is to make something up, if you look at ChatGPT.

You need to give it an off-ramp. If you can't find the data, say you don't know. And I think that helps make it more human.

I think the other piece of this, and this is where we're starting to evolve in Alteryx, is now, because of data security and GDPR and whose data and where is it protected, getting expressed consent. So if I'm Adam and you're calling in and I can say, "Alan, I'd like to look up your account. " You hit a little button.

Then it can go back to the CRM to figure out what do you have installed, how many users you have. We don't put that in there initially because we need to make sure that we get the consent. So we're putting the guardrails on it- Yeah ...

to make sure that- No, I think that's the way we have to-- Look, I believe at some point, it's going to be more in the EU- Yep ... than here. But that kind of stuff is going to be mandatory.

The AI just can't decide to jump into your stuff. 100%. It's funny, I've been experimenting with an AI EA.

Uh-huh. And I'm going through this right now. I call her she because I named her Crystal.

But she asks my permission every time I'm asking her to do something new, or I have integrated with Slack and a bunch, Google and stuff, HubSpot. And I notice it's done, and it's probably because it is a European-based company that's doing it. And you're right, it gave me a sense of, you know what, they're respecting my boundaries.

Your boundaries. Yes. And I like it a lot.

So you're 100% correct there. And I think as customers are doing the same thing, they're looking at, from a trend perspective, how do I put my toe in the water? Mm-hmm.

Looking at efficiencies with ChatGPT and making an employee more efficient. Then they look at the headless superhuman- Mm-hmm ... that you're feeding it information.

And then the third evolution is now looking at agents that sort of connect together. And you start looking at a sales agent and a marketing agent, or a success agent, and now how do they talk to each other? Each other.

I didn't have to build Annie. My team didn't build Annie. We built Adam.

But we're now figuring out how do we both have agents out there- How does Adam and Annie- ... that talk to each other ... talk to each other?

Yeah. Yes. Just like you and I would work together.

" And pass information. Yeah. Yeah.

It's fantastic. I love the whole idea of it. Let me ask you a harder question.

Okay. And this has always been a question in customer success. What's success look like?

What is the measurement, especially now in this AI world- Yep ... you gave me one metric. You're deflecting more with 40% of your cases.

Yeah. That's a good one. Give me some other kind of marks.

Let me get some clarity. Good. When you say how do I measure success, are you measuring my organization internally, or how do I measure it with my engagement with customers?

Because they're different perspectives. How do you measure, let's call it customer sat. Okay.

You could do MPC and all that sort of- I don't like customer sat. Okay. Honestly, I think it's- Tell me what you think is the right- I think it's tactical.

It's a point in time where you do a survey or something and you sort of get it. Yeah. I look for that referral.

" That's a more sticky measure to me. I look at it from a success perspective internally. Do we have the right leaders at our accounts that are engaging in the QBRs?

Is it just a director? Is it a VP? Do we have an SVP?

Do we have a C-suite? Where is the level of investment? Do we have documented use cases in where the customer can go talk to their board about demonstrative value they're getting out of this solution?

I think those are really big. I think the next one is going to be, like I said, around referral business. Are they open to having a conversation, whether public or private?

" I think that's a measure that's really powerful for us. I do think that one of the marks is how often are they using support, and how often are they using education? If they're not calling a support or they're not using education, that's a risk flag for me.

That's a red flag. Yeah. Yeah.

But ultimately, all of these things lead into the NRR and the GRR. Are they renewing? Are they finding value?

Are we able to articulate that value, and are we expanding that with the customer to go forward? And again, it's not about selling them more, but it's are you capitalizing getting the maximum out of what you purchased with Alteryx? But those are the things I look for.

Love it. Yeah, no, I think that's a great way of looking at it. One of the other things that we're seeing, we talked about trust in AI, and that's one aspect.

The other aspect is this whole when you're moving at AI speed and AI scale, it's very hard for the humans to keep up, right? There's this term human in the loop. Other people say human at the helm.

But how do you in a customer success organization keep that human in the loop? We're all running as fast as we can. AI is exponentially deeper, wider, faster.

Yeah. How do we keep the human in there? I don't think that the human can ever come out of it.

And it's interesting because I may not always be what everyone wants to hear, but people buy from people. Yes. There'll always be a need for a customer success manager.

They'll always have to have some kind of interaction with that customer. How that CSM is set up to be successful, to help make that customer more productive with more use cases, more value on that platform, more usage across that platform, that I think is more of that measured piece of that for us to go forward. Got it.

Hey, we're almost down to the bottom here. Let's take a look forward. Okay.

Make believe we're here in 2027. Okay. And the way things are moving, it's almost mind-boggling to even think, but what do you think the big stories are there for 2027?

I think you're going to hear many more. So you're going to hear about the customer awards on Thursday. Mm-hmm.

And I have a nice black brocade jacket I'm going to wear on stage. Not quite a black tie, but pretty- Close enough Yeah. I think you're going to see, we had a phenomenal amount of submissions, more than we've ever had.

I think we're going to start seeing more and more of our customers start embracing AI in real valuable use cases. Say people are putting their foot in, they're trying small little things. But I do think at the speed at which AI, at the speed at which the intelligent enterprise is moving, it's almost like every month is a year, and you're going super, super fast.

We had a decade to get through. Everybody's wanting it and we're all trying to get it, figure out where we're at. Yeah.

No one's solved it yet. No. But there'll be some companies- There are companies that are running- Yeah ...

so I use this term living below the AI poverty line. There are some companies that are embracing and running so fast, and they're going to be fat. Yep.

And then there are some companies, your traditional sort of cautious, I want to see what everyone else is doing first. I want to see some success before I jump in. And they're going to live below the AI poverty line.

Yep. Right? Because they didn't invest.

They didn't move fast enough. They didn't trust it. They didn't embrace it.

They pulled back. And for a lot of companies, that's going to be a life or death kind of- Yeah ... decision.

I will say, as you look at the space, to your point, the more traditional companies are going to be more of a traditional waterfall project. Right. IT's going to manage it.

It's going to put up this big wall. Alteryx actually has a secret weapon for everybody who has it. They're already connecting to data today.

Right. We have access to help them connect to LLMs, either their own internal LLMs or public LLMs. Right.

Frontier, yeah. Yeah. You can start building those proof of concepts out.

I was at the Gartner conference last week in London. Mm-hmm. And one of our customers, a large financial institution over there, was talking about how IT told him it'd be nine months to get something done.

He went over the weekend. Friday night, he was told nine months. Monday morning- He also- ...

he put the concept up there, and they're like- Who did this? Yes. With AI pulling it through.

And this is not about removing IT from the picture. Not at all. We still need the governance.

We still need the right data sets. We still need to make sure it's managed. And you'll start hearing more of that come out with Ben's keynotes this week.

But his point was all the data that I already have access to, I now automated it and I have an agent making it work. " Alteryx- That's how game changing it is. Correct.

Yeah. I don't know whether there's another way of saying it. There isn't.

Yeah. It's a game changer. Yeah.

Bill, thank you so much. Thank you so much, Alan. It's a pleasure.

I enjoyed the conversation. All right. Bill Tabbert Humphrey, Chief Customer Officer here on Alteryx.

We've got a few more interviews today, so stay tuned for that. You're watching Techstrong TV.