Turning Analytics and AI Into Business Value

Hey everyone. We're back here at Inspire 2026, the big user conference for Alteryx. My next guest is Steven Birdsall.

Steven is the CRO at Alteryx, and first of all, Steven, welcome to Techshow TV. Thank you. Man, it's great having you on here.

It's great to be here. Nice to meet you, too. Nice to meet you as well.

" Well, that's one thing a chief revenue officer does. Right. But there's a lot of hats that go under that CRO title.

Steven, in order to successfully juggle those hats, though, you got to kind of know what you're doing. So I imagine you've had some experience, some journeys in your career. If you wouldn't mind, share with our audience a little bit of your story.

Sure. Happy to. I've been in software for about 30 years, so I'm dating myself a little bit.

I'm right there with you. Yeah. So about half my time I spent with SAP.

Okay. I was chief operating officer in North America, Japan, Asia Pacific, Latin America, and then globally. Mm.

And then I ran a bunch of businesses. This is the fourth time for me to be a chief revenue officer. So lots of experience steeped in private equity, taking a company public like Anaplan.

Thoma Bravo owned the last company I was with, Click. And I've been here for about 18 months, so yeah. Oh, really?

Okay. Super excited to be here. Sure.

Thoma Bravo. Run into a lot of companies that are- Oh, yeah ... Thoma Bravo companies.

We're with Insight and ClearLake now, Insight partners with ClearLake. Sure. I know Insight as well, too.

I have a lot of friends at Insight over the years. My best friend at Insight, I don't know if you know, he's actually from, well, he's US, but he lives in England now, is Emmett Keefe. Oh, sure.

Emmett, what a great guy Emmett is. I worked with Emmett when I was at SAP. Did you?

Yeah. Yeah, actually, yeah, he was there. So.

When Emmett was running the Insight Ignite- Uh-huh, sure, yeah ... events, I've been to a bunch of them. That's awesome.

Yeah. George Matthews over there from SRC. Yes.

Yeah. So, but yeah, those- Insight's a great- It's a small, kind of- It is ... close-knit community.

It's a small... Look, one thing I've learned in also 30 plus years in tech is, it's always a small group. It is.

There's six degrees of separation in tech, right? That's right. So you here about 18 months.

With all your experience, and great experience in this, what attracted you to Alteryx? Well, after I left my last company, I wasn't sure what I wanted to do. I was toying with the idea of maybe retiring, and so I went off and I took a couple of years.

" Yeah. " Uh-huh. " And Alteryx was in this unique position where we were acquired by Insight and ClearLake in January '24.

Took us private after being public for seven years. And I love Insight and ClearLake, so I know them as sponsors. My reputation, they were fantastic.

They had also done a great job at just managing the business and getting a lot of the expense out, and kind of setting up the new leadership team for success. And then that's what they did, and so they hired Andy at the end of '24. I came in about a month later, and then the entire leadership team came in.

And now here we are, 18 months later, the product is completely re-platformed. We've got an entirely new go-to-market. We changed the way we price to be more sensitive to our customers, make it more capacity-based, so you pay based on value creation.

All kinds of changes that we've made in the org. I didn't make major organizational changes, but I did change some of the structural things, with a focus on Alteryx One and our adoption of the platform. And then now this year, again, more and more of our customers have moved over into this capacity-based pricing model.

So I think those are important things for us now. Now it's a matter of just realizing the value of Alteryx One, and that's what our team is focused on. So yeah.

Absolutely. Exciting times. So I guess you're telling me you haven't been that busy.

Not that busy, yeah. Just by attitude, I love being busy. I am a self-proclaimed workaholic, and find a good life balance.

So, yeah, it's fun. I'm genetic. I'm in the same boat.

I don't know what I would do. I live here down in South Florida. There's enough retirees in South Florida.

Oh, yeah. I live near Boca Raton, and I just couldn't imagine myself... What, get up and play golf or pickleball or- Right ...

versus doing this? I'd rather do this. Yeah.

I do triathlons, so. Do you? Yeah.

Good for you. So that's my- Good for you ... fun side projects.

Well, you need it, right? That's right. We all need to be active physically, and especially when you're around a while.

You presented here- Yeah ... already today. A lot of people watching this didn't get a chance to see it, though you can, because I think a lot of the event is being done virtually, and so it'll be available.

But for those who haven't seen it, give people a sense of what you spoke about. Well, I think there's a lot of questions on: How is AI changing go-to-market, and what are we doing to show that we adopt it? So I spoke a little about, again, some of the organizational things that we've done.

But to me, I love the fact that everybody's using AI, they're experimenting with it, but what they're quickly finding is that while cloud is awesome, and OpenAI, and some of the other large language models, Groq, using Gemini, there's great things. But when you train it on a corpus of data that is just in its entirety in a cloud GitHub platform, and you're not really connecting all of your systems of record, like a Salesforce and everything else, it might sit in Snowflake or Databricks. How are you actually using that for the better good to run my business?

So if I want to run a calculation on commissions, for example, and I want to build an agent for commissions that I'm going to provide to my sellers, it has to knowHow I pay the reps, what their territories are, their compound, their on-target earnings. There's a lot of things that go into it. That corpus of data may or may not just be sitting in Salesforce.

It might actually be sitting in a Snowflake table somewhere. And so being able to connect the AI models with the business logic that sits inside Alteryx, where you're building all these workflows, and we have about 480 million workflow automations run in 2025, 774 trillion rows of data was run in Alteryx, and 258 petabytes of data kind of ran through our systems. That's a lot.

That's a lot of data. That is. So that business logic is tied in with thousands of customers, 700,000 active members of our community.

So people are using Alteryx, they love it, really high NPS scores. And then they have the cloud data platform that they can use where it stores all that data. So we have a product called LiveQuery, and I talked about that, that actually connects into the cloud data platform, and then as you build your workflows, as you enrich that data, you bring it from all these different sources, you bring in your large language models, and you train on it, and you prompt it.

Then all that enriched data then finds its way back into the cloud data platform. So we create this end-to-end unified model that allows everybody to connect data from any source, prepare and blend it, automate it, run a workflow around it, bring an agent into that model, into that canvas, and interrogate it through prompts, and then store it back down in the cloud data platform. Like that is odd.

Nobody does that. No. That's like a virtuous cycle.

That's right. Where you're taking the data from that data platform, analyzing it, working on it, the AI is using it and doing its magic with it, and then boiling down that work product back into the data to make the data better for the next iteration, for the next go around. It's a- Now people, what customers are telling me is they don't want a model that just is trained on this big corpus of data, and then they don't know what it's doing.

They can't just go, if you're a business analyst, you're not going to go review Python script. Mm-hmm. You're not technical that way.

You want to be able to build an agent, but you need to be able to know where it's pulling from. So if you need to make sure that GDPR has been applied, or personally identifiable information, PII, or if you want to know how the commission structure that it's been applied to that, or if you're looking at a certain customer segment, all those things need to be visual. So they want to have visualization.

They want to make sure that they understand the data. So how do they make sure that now that you visualized it, do they understand what it's doing and kind of that, again, managing on the corpus. And then they want to make sure that it's repeatable because now I'm going to build a system around it.

So now it's repeatable, and then it's auditable. I can always go back because a lot of times I don't want probabilistic data, I want deterministic data. Yes.

You're in finance, I can't say I have about $10,000 left in my account. I need to know precisely. So deterministic versus probabilistic is where things really come to light.

That business logic sitting inside Alteryx is where that all happens. And almost by default, AI is the reverse of that. That's right.

Exactly. It's not deterministic, it's probabilistic. When we look at this though, one of the biggest problems I hear, and I obviously speak to a lot of people doing what I do, is that these LLMs are great.

I'm not here to tell you they're not. And I don't care whether you're talking about Grok or Chat or Claude, obviously, or even the Meta models and stuff are great. But they're as good as the data they've been trained on.

" Right. The idea, and I think this is where the action's going to be for the next at least couple years, is training your model on your data because your data is the valuable data that is going to allow your AI to really provide value to you. It's not what it learned digesting every book ever written or something like that, that they didn't pay for or did pay for.

But rather the data that resides in your Snowflake, that resides in your data lakes. That's right. And it sounds to me like Alteryx is right in the middle of that.

Taking that data that resides in those data lakes and putting it into form factors and containers, if you will, that allow the AI to not only provide value today, but actually to train to be more efficient and more valuable going forward. That's right. Yeah, everybody that uses Alteryx are really business users.

These are analysts. Sure. They work in finance, they work in HR, they work in supply chain, they work in sales operations.

These are people that are managing the business day to day. And so they understand how the business runs. They need access to the information, so if I want to gain access to our data that's sitting in a cloud data platform, I have what I call a data concierge.

And I can actually request data. Maybe I want to look at my Salesforce data, I want to look at Marketo, I want to look at Optimizely, Six Sense, and I want to pull all this data together. I've got some information in Certinia.

I pull that in through a data concierge. It gives me access to it. So you have access controls, you have governance, you have auditability.

I can pull that into my canvas. And because it's really easy to use, it's all drag and drop. It's really easy for them, for an analyst to say, "I want to go through these steps.

I want to automate this process. " Again, if I'm in sales operations, there's certain things I want to do. If I'm in manufacturing or I'm in finance and manufacturing, how do I make sure I look at all of the different inputs into a manufacturing process?

I've got visibility into my supply chain. I have visibility into the products. If I change the nature of a price, and I want to look at elasticity, I've got all the demand planning that I've been doing.

And so you can't just take a large language model and go train on, again, this big set of data without having the business logic tied into it. Applications are where you store your system of record. You want to make sure that, again, that's auditable and you have that information, but the business logic is where all the definitions, the rules of the business are run, and that's where the analysts live.

That's where we have been for almost 30 years. That's what we built our product on. Sure.

You're always going to have the other technologies that are out there that can help you with governance, compliance, again, the auditability. You want to have the cloud data platform. So a lot of the IT organizations are setting up all of the governance aspect to that.

And then tying that between the systems of record and then where the business logic sits is where Alteryx is. I love it. We feel like we're in a sweet spot.

I don't disagree with that at all. I feel compelled to say, though, or to ask, a lot of people are not 100% sold on this AI issue. Right?

Can I trust-- Trust is- I bet ... the defining term. Can I trust the analysis?

Even if I have Alteryx and it's bringing up, it's surfacing this data for me, and now I bring my AI in to work on it, can I trust it to-- Is it giving me the right conclusions? Is it making the right decisions, or at least suggesting the right decisions choices? Because I may not autonomously allow it to do things anyway.

But can I trust it from a security point of view that I'm using one of these LLMs, and now, boom, everybody has access to my crown jewel data, right? Right. What do you say to people about that?

Trust is a word that we hear every single day in almost every single conversation. Trust, governance, compliance, auditability, access to the data. Not everybody has access to all the data, so you have to create access controls.

You need to make sure you have the right governance in place. So the IT organizations are making sure that a lot of those things are set up. But at the end, I need it to be visual.

If I can't see what's happening with my data, then I don't know, and I can't just trust it, it can't be just a black box. And so taking a large language model and running it on this corpus of data without knowing what's happening with my data, I can't trust it. So in order to have auditability, you need to be able to visualize it, and that's where our canvas really comes to life.

So most of the customers, the reason we have such high NPS scores with our customers is they love the fact that they can see it right in front of them, and it's easy to manipulate it. " IT can't own all of that. The business has to own that because the business changes every day.

And as tariffs come out and new tax laws and everything else, the business needs to be able to operate with speed, and they can't rely on having to rewrite Python script to do that. So there's got to be a happy medium, and that's where Alteryx fits in. Love it.

Of course, we're in this generative AI is so 2025. And now it's all about the agentic. Yes.

Right? And with agentic comes autonomy and more than just looking at things or analyzing things, it's actually doing things. Right.

How is that playing out with the Alteryx? What do you hear from customers? Because again, that trust is all wrapped up into that.

That's right. But it's that next step. How's that playing out with your customers?

There's a few things. So I'll actually answer it in two ways. One, internally, what am I doing?

Because I have about 1,000 people on my team, so I run sales, customer success, support, renewals, all of the go-to-market functions. So I'll talk about some of the things I'm doing internally, and then externally, what our customers are saying and how are we helping to support them. So first on the internal side, we've actually rolled out some capabilities.

We rolled out a product called OneMind that basically is an AI super agent or an AI agent. So Annie sits on our website. Yes.

So she's a person, if you've seen her. She's here. Well, Michelle spoke about her.

Oh, good. So Annie's here at the Inspire. She doesn't have red hair.

She's got brown hair. Okay. She's got brown hair.

Just checking. And she's awesome. So we trained her on all of the information on Alteryx, so she knows how to position Alteryx.

She can help you with a pilot, to go in and do a demo. She can help you with the partnerships we have with Google or Databricks or Snowflake. She can help you with how to set up a workflow.

Pretty incredible. So she's trained on everything. She's basically an inbound BDR.

Mm-hmm. She does pre and post form, but she sits there on the website and does that. Adam is our CSM super agent.

com. So if you're a customer, you log into My Alteryx, you get access right to Adam. So now instead of, I have about 100 customer success managers, they can't interact with 8,000 customers.

Sure. So how do they do that? Well, Adam is there to help support those customers, to answer quick questions and provide the support that they need.

And then we have Atlas, who is a solution engineer. He's a headless solution engineer. So it used to be if you're in inside sales and you wanted to do a demo on a call, a first call with a customer, it'd be hard-pressed because I don't have unlimited numbers of SEs.

So Atlas is now on a call with the ISR, and he can actually do product demonstrations right on the fly. So this agentic workforce that is there, for us, I'm not replacing any people. It's about how do I actually make people more productive?

How can we double our revenue without doubling our travel and expense that goes associated with it? So that's what we're focused on, is getting greater productivity out of the field by giving them agents to help them manage their day-to-day business. The whole idea of generative AI is, sure, people can just write easier emails.

They can be more productive in the way that they do. That's okay. That is 2025.

Mm-hmm. We did that. We thought it was cool, but that one feels like- That was last year Now what people are doing is, I've seen a lot of sales reps are creating these permagraphics of their customer.

They bring in telemetry data, they bring in pipeline data, they bring in all the interactions they've had, and they have a graphical illustration with all the value that was created, how many hours, how many use cases, how many workflows have been built. It's pretty amazing. It is.

And they built that with Claude. Yeah. So they bring it right in and they build that.

So I have this thing in Slack channel, it's really cool. In Slack channel, we have this Go-to-Market AI lab. That's the channel.

Okay. And I have people that every day are posting all these ideas. "Hey, look what I did.

" We're doing the same thing. It's so awesome. Yeah.

We love it. Yeah. Put in what you're doing so everyone else can share and learn from it.

Yeah. And it's like watching evolution. That's right.

Like God's creation. Just- Yeah. And I want that, and I want everybody to experiment.

But at the same time, I can't scale every single idea. Yeah. Some ideas are not really scalable.

So what I have is a forward-deployed engineer, an AI person on the team who has a small team of people, and what they do is they take all these great ideas, and then they figure out what we're going to have that should be repeatable, and then they build scale around it. They might need to build an agent, but that agent needs to be maintained. You can't just build stuff and then just throw it out there and have the rep be the one that's maintaining it.

You really need to be focused on building something that's scalable. So there's a team of people on the rev ops that can actually build this stuff for scale. Same thing Michelle would have on as our CMO, she would have that for an AI person for marketing.

Yeah. Someone in product would have that, someone in finance. But they all need to interoperate together functionally across all the different lines of business and make sure that we're building these agents at scale.

They're all doing that with the same kind of idea in mind. Now it's auditable. Now it has the compliance, it meets all the criteria.

So on one end, I've got a factory of ideas, this ideation that comes up. People love sharing all those ideas. The SEs get access to that, and they come up with new demos that they're going to do.

It's really, really cool. But then we figure out what we're going to go scale, and we take that into the market. Now we can take that with the data concierge getting access to all the data, being able to build all this stuff.

We actually have things we can show to our customers. So that's Alteryx on Alteryx, but how do we then show that with our customers? That's where we start talking to the IT organization.

Again, setting up the right governance, setting up the right data, giving them the ability to have access to all these different data sources. Not just build things in Python, but build it so that the business can actually own it and run it and operate it. So that's where these things start to converge together, and I love it.

It seems like every week something else is coming out, whether Anthropic is coming out with a new product or- Every- OpenAI. Yeah. It's incredible.

Well, they're all competing with it and- Yeah ... I think it's only going to increase it. It's so fun, though.

The market is so fun because it's moving so fast. I am thrilled by it. I know for people like us who've been around the block, this is fascinating.

Yeah. One day they'll do all of this in business schools, they'll study all these things. Right.

I got one last question for you. Okay. If you don't mind, Steven, I'd like you to address the audience.

Okay. What is the biggest misconception enterprise leaders are still having now about deploying AI successfully? I think there's an old saying that you can't let perfect be the enemy of good.

Yeah. And to me, I think experimenting, coming up with ideas, figuring out that, hey, your data's not all clean and it may not be everything that you need, but you need to have access to information. So how do you get access to the data?

How do you make sure that that becomes easy for your people that actually know they're running the business? How do you give them access to that? How do you let them experiment with it?

So I think the biggest misconception is I've got to get everything perfect before I allow the field to start engaging with it. That's a mistake. Let people go get access to it, find out, uncover where you have faults, where you have bad data, where you have issues, bring it up to the surface, address it, and then let people experiment further.

Now they're experimenting on good data, on good information, and then they can build and grow from there. But don't let the organization get so pinned down with having to get perfect data. It's okay.

Use a large language model to go solve the problem, and then test that with what you have in place today to make sure the answers are correct. Are they seeing things eye to eye? If not, then solve for that.

It's like any other math equation. You have a set of variables that are known and set that are unknown. You isolate the unknown and you solve.

Same thing with data. Identify where you've got bad data, where you have a need for cleaning up your data. Get your house in order on that side, and then you can start building from there.

But yeah, experiment. It's fun to experiment. Do it agree, right?

You can't let it- You can't just run a science experiment all the time. I get that. But you want people engaged with it, especially with what we do, and we're in the heart of AI.

That's what I love about Alteryx is we are that business logic layer that ties all the data together with the systems of record, and we allow people to actually leverage large language models right inside the platform. And because of that, I like the experimentation. I like our team to be engaged with that because it helps them understand the problems that our customers are facing and what they need to do about it.

So a lot of what I do is just knocking down roadblocks and making sure that they're enabled to be effective in the way that they're experimenting. I love it. Yeah, it's fun.

I like one last thing. I just want to reinforce this. Annie, Adam, Atlas.

Yes. Yeah. Those are the three inputs.

That's right. Whether you're an existing Alteryx customer or a new customer, go to the website, interact with them, and you're on your way. That's right.

com. com and you can talk to Annie, and she's awesome. Love it.

Steven, thank you so much. Yeah. Good to see you.

Steven Birdsall, CRO of Alteryx, here. We're at Inspire. We're going to have more coming at you in just a moment.

Stay tuned. Thank you.