Afraid of AI? Your Team Isn’t: Lessons From Early HubSpot to the Agentic Era

Thanks for that introduction, and thank you all for coming to my session today. I'm Frank Gozier. I'm the COO of Onramp, and I'm here to talk to you about some of the lessons I've learned over the years, going back to my early days at HubSpot to the present time in the age Antiguera. Hi, Mike.

And a little bit of background on myself. I have had the privilege of building and leading CS teams for almost two decades now. Big companies, small companies, some successful, some unfortunately not. Those were not my fault.

And most notably, I spent 10 years on the early executive team at HubSpot. I built with my team all of HubSpot's initial post-sales processes, and then we went public, and then I became the CIO. And then the company was 3,500 people, and it wasn't a startup anymore. And so I left, and I bounced around to a few startups, and then I found Onramp, a company that is purpose-built for onboarding, which is something I'm very passionate about.

So I joined Onramp to build their CS team, and then of course, I bet lots of you in here have been promoted after doing good work, and you know how that works. The reward for good work is you get a harder problem. And so now I find myself as the COO at Onramp. And one of the things that I've observed over the past couple decades is CS has evolved a lot during that time.

I like to think of it in several eras of CS, but at the end of the day, I find I'm always solving for the same thing, which is how do I keep my customers, my team, and my CFO all happy at the same time. That's been the challenge back in 2010, and it's the challenge now. So our agenda for today, in the next 45 minutes, I'm going to start with what I'm calling the CS Aries Tour. Now for any of you Swifties in here, just be prepared.

This will not have the same level of pageantry and sparkle as Taylor's Aries Tour, and I'm definitely not going to sing. But a side note about me, my wife and I are competitive ballroom dancers, and if she were here, she would be bringing the pageantry and the sparkle. But she's at the spa, and so you're going to have to just listen to me for the day instead. So I'm going to talk about the impossible trade-off that we all deal with between how do we have CS that is effective on one side of the trade-off and on the other side of the trade-off, how do we have CS that scales.

We've been grappling as a function on trying to find the optimum way to do this forever, and so we're going to talk about that a little bit. Then I'm going to talk about how a genteck AI maybe changes the game a little bit. Maybe the impossible trade-off is no longer as impossible as we think it is. Talk about that, and then we'll talk a little bit about what this means for each of us and our teams.

I'm going to try to give you something concrete and actionable that you can take back with you and use with your teams and your businesses right away. But before we get to that, let's go to our first poll. So if you can get on Slido, our first poll question is, right now, what's your biggest onboarding and engagement challenge? Is it that it's too expensive and time-consuming, or that your customers are disengaging?

Is it slow time to value, or is it too many customers to keep up with, and things are falling through the cracks? Now, I hear all of these from our customers and prospects and from my colleagues all the time. I'm sure many of you hear these same things or have felt these same things. And if you look closely at these, and as you're trying to choose which one to pick, you notice that they're all sort of related.

If we start at the bottom, if you've got too many customers and things are falling through the cracks, then it's not unreasonable that that would be a cause of slow time to value. And if you're slow of providing time to value to your customers, then they might disengage. And solving any of those things is expensive. So these are all sort of closely related to each other.

And I see that customers disengaging and too many customers to manage are leading the poll right now. But the point of this poll is that no matter which lens you look through when you think of onboarding and engagement challenges, whether you look through the lens of cost, or engagement, or time to value, or volume management, fundamentally, we're trying to solve the same problem. And that is, how do we provide service to our customers so they can get value from our products in a timely way without breaking the bank? Sure, each of our businesses is a little different, and we have nuances, right?

But at the end of the day, if we're in CS, and most of us are, probably all of us are here, then we're trying to figure out how to do more with less. So I want to talk about that problem of how we handle doing more with less, or how we handled scaling through a few different funny characterizations of the three eras of CS. So the first one, which I call the open bar era, this is 2015-ish in prior. Now in the open bar era, everybody drinks, and somebody else pays.

In this era, capital was free and abundant, and if you had a problem, you solved it by spending. Just throw money at the problem. Go big or go home is what Brian Halligan used to say to me all the time. And that era was a lot of fun while it lasted.

But eventually we figured out that we had to pay some attention to margins, that they did matter. And so we, as an industry, we embraced segmentation, and this leads to what I call the velvet rope era. In the velvet rope era, the party was still raging, but you had to be on the list to get in. If you were in the wrong customer segment, then you had to accept a lesser customer experience, and maybe certainly a lesser service level.

Unless you wanted to buy your way in, then you could cut the line. And that brings us to around the time of 2020, the COVID era, or as I'm calling it on this slide, the virtual party era. At this point, the party was shut down, cost control was paramount, the club was closed, but you could stay home and sit on your phone and interact with technology and tell yourself that it was just as good as interacting with human beings. But maybe it wasn't.

So let's take a more serious look into each of these eras one at a time. For the first era, I want to tell you a real life story of the early days of managing churn at HubSpot. And there's a couple people who worked with me back then to keep me honest, and I'm sure this PTSD moment will be ingrained in their memories. So in the early days of HubSpot, around 2010, 2011, our churn rate was between 5% and 7% per month.

Now, if you do a little math on that, that means we were turning over almost half or more of our customer base every year. Not a healthy place to be. But back in 2010, there was no, there were no platforms like Gainsight, there were no tools for CS, there weren't even best practices that were well established. We were all sort of making it up as we went along.

And so in those days at HubSpot, we didn't even have a CSM team. And we tried several things to get the churn down from the stratosphere into a manageable place. And those things didn't work. And then one day, we decided, well, let's try this CSM thing.

And so we spun up our first CSM team. I got budget after a hallway conversation, I got budget to hire a team of nine people. That's the first part, looking back on this story that I think is amusing, that I could have a hallway conversation and get budget for nine people. How hard is it for us to get nine people on planned headcount budget today?

Things have changed a lot. In any case, we had three or four thousand customers at the time. I don't remember the exact number. It was too many to assign out to nine CSMs.

So I divided the customer base into two identical cohorts, 50-50, not grouped by segment, but identical cohorts. And I used one as a control group. And they didn't get assigned to CSM. And I assigned the other group to the CSMs, as well as the hundred or so customers that we were signing up each month.

And we watched. And the good news was our customers at that period of time, they would go off the rails really fast. And so it didn't take long for us to validate that the CSM model was working. So that was great.

It was working. So I started to hire one or two CSMs per quarter to keep up with customer-based growth. And I thought, okay, well, I don't know what the right number of customers per CSM should be. There weren't established best practices like there are today.

So I thought, well, I'll pay attention to the metrics. I'll watch the churn numbers. I'll pay attention to the morale of the team. And the numbers will tell me when I hit the ceiling.

And so every month for two years, we inched the workload of the CS team up, up, and up. And we also grew the team. And every month for two years throughout all of 2011 and 2012, the churn number came down every month. By the end of that, I was feeling pretty good about myself.

The company was feeling pretty good about the numbers. And we were in the early stages of talking to bankers about IPO because we had finally solved our churn problem. So after two years, 24 consecutive months of progress, January 2013 came and the churn number went through the roof. And I don't mean a little bit of a spike.

It really spiked. And we thought, this must be an anomaly. After two years of consecutive progress, this must be a blip. Well, then the blip came back in February and I thought, I'm going to get fired because of this blip.

And I did not. And we dug into the numbers. And what we realized was that the metric I was looking at, our monthly churn, was in fact a lagging metric. Now, you're probably sitting there today going, no kidding, Frank.

We all know that churn is a lagging indicator. But in 2013, it was a bit of a new revelation to us. And so it was worse than we thought. We had been like six or eight months past the point where the team was reasonably staffed.

So now here's the funny part. My proposal to solve the problem, I went back to the leadership team and I said, we need to hire 30 CSMs in 30 days. And instead of being laughed out of the room, I was applauded for the boldness of the suggestion. The multimillion dollar unbudgeted expense was approved.

And me and Julie Hogan, who some of you might know, a good friend of mine, she was my right-hand person at the time in our recruiter day, Fernandez. We spent all day every day for a month. We hired 30 people. We spent millions of dollars.

And a couple months later, the churn number had stabilized and it came down. Point is, in the open bar era, when you had a problem, the solution was get a fresh credit card, walk up to the bar, slap it down, extend the tab. That's the way we did things back then. That was scale through headcount.

Can't do that anymore. One of the lessons I learned from that experience that I've never forgotten is to pay attention to the difference between lagging indicators, which we have lots of in CS, and they're all important, but know the difference between your lagging indicators and your early indicators. We all need to have things that we watch, metrics that we watch with respect to our customers from the very earliest moments. How long does their handoff take?

How well do they get through onboarding? When they complete onboarding, are they actually at value or not? There are early indicators that are critical, not just the bottom line lagging indicators that we're all used to watching. Okay.

Well, after this era, then the industry started to evolve a little bit, and we started to all pay attention and focus on self-service. If you've ever had to pay for an open bar, which I had to at my daughter's wedding, then you know that the way to control the cost is to limit the guest list. We used segmentation, simple concept really, to say, "Okay, we have high value customers over here, and we have low value customers over here." Regardless of how you define value or how many segments you chose, the theory is you provide the same high touch, high value, context driven, responsive, empathetic service to your high touch customers, and you funnel your lower value customers to some kind of a self-service model. Back then, we leaned into a lot of things like knowledge base articles and video libraries, group webinars and trainings.

The phrase "one to many" started to get a lot of attention, and it was all about saving money through self-service. Now, you had to backstop these self-service resources with a couple of CSMs, operated at much higher customer counts, and we sort of knew that those customers in the lower value segment were going to churn at a slightly higher rate, but if you worked your model properly, the money that you saved from reduced headcount could be greater than the dollars you lost from the increased churn. And so, net-net, this was a win for your margins. Now, it was a bit of a loss for some of those customers, but hey, business is business, and we had to do what we had to do, so we just accepted that reality.

Well, the problem with this model is it does provide you some scalability when you roll it out, but what about in the next year? In the next year? In the next year? Where you have to provide increasing amounts of contribution to that margin gain?

What choices do you have? Well, one choice is you can continue to expand the number of customers who get a reduced service level in the low value model, or you can make the service level you provide to everyone in that model even less than it was. And if you follow this pattern for long enough, self-service isn't really self-service anymore. It starts to approach being no service, right?

It gets to the point where we're not really servicing these customers so much as we are managing the economics of not servicing them. So then we get to the third era, which I like to call the rise of the bots, the chat bots. They came drifting in. And in this era, one of the things that we realized based on our experience from the self-service era was that even the longest, far out on the long tail customer, sometimes needs some interaction.

It's not enough to just give them access to a bunch of self-service resources. They need help finding the right resource. They need help following up. There's times when they need to interact.

And so bots promised to solve this in a scalable way. If you throw a chat bot in between your customer and your ever-growing set of self-service resources, then there's a scalable way for them to interact and get what they need. Well, it's a very good theory. But the challenge with the first generation and the early generations of chat bots is that they have limited ability to really understand and incorporate context into the response to the customer.

In fact, they're oftentimes built on some form of a decision tree and you have to anticipate what your customer is going to ask in order for the bot to be maximally effective. And that's hard to do. And I like to think of these bots as certainly an improvement over what came before them. But in some ways, they were more like an advanced phone-based IVR system.

Now, who of us hasn't been in a loop with a phone-based IVR system where you're just pounding zero on your phone and screaming to talk to a human being? We've all been there. Now, maybe none of us, maybe some of us, have yelled and screamed at a chat bot. Maybe some of us have thrown a keyboard.

Not me. But the point is that in certain edge cases and in certain situations, the first generation of bots, sometimes it provided as much or more frustration to some customers in some situations as it did benefit. And the point here is through each of these eras, the technology got better, but we never really fully solved the problem. There's always been an impossible choice, an impossible trade-off that we've had to accept.

On the one hand, we know that what really works is if you can just understand your customer, build trust with them, be a partner, the thing every CSM person wants to do. If we could do that with every customer, we know it works. The problem is it's very expensive. So that forces us to explore different forms of tech touch, and they just have limited effectiveness.

So we spend a lot of time trying to figure out the right balance of these things, and it's sort of like pick one, lose either way. It's a real impossible choice, and we've been stuck with this choice for so long that sometimes we start to use it as an excuse not to innovate in our process as much as maybe we could, because we know that no matter what new thing we adopt, we're still going to be stuck with the same all impossible choice. Well, I think that that world is now ready to change. Agenic AI really does change the game.

It is now possible for that trade-off to not have to be impossible. We can take context-aware agenic technology and deploy it side by side with our teams and with our customers, and we can have all the benefits of context-aware interaction and all the scalability of technology. The good news is the trade-off is not impossible anymore, but the imperative for all of us is that means we can no longer use the difficulty of that trade-off, the impossibility of that choice. We can no longer use that as an excuse not to change.

Every one of us has to have a strategy to embrace agenic AI in our businesses and with our customers and our customer-facing teams, where once it was aspirational to say that we wanted to provide the absolute best service at a very scalable cost. That's not aspirational anymore. That's the new operational imperative for our teams. So let's bring this back to our eras tour here.

We started on the far left with opulence and we made our way to era three in scarcity, and now we're in the agentic era, and I like to call this the "everyone's a VIP" era. We no longer have to use segmentation to create haves and have-nots in our customer base. Yes, we'll still have differences in the way we treat a million-dollar customer versus a 25-dollar customer. Segmentation is here to stay, for sure, but the harshness of the choices that we've had to make and the wide disparity in the engagement model, that can come together a little bit.

And so that is exactly what we are doing with Onramp Arrow. So I'm proud to say today that Onramp is launching our suite of agentic AI for customer engagement at scale. We're calling it Onramp Arrow. This is not a product pitch session, so if you want to learn all about it, please go to our booth and talk to our folks.

They'll tell you all about it, but briefly, we provide agentic capabilities for all the important constituencies in customer onboarding engagement. It starts with agents for your CSOps people and your IT people. Often overlooked is the cost initially for implementing a new platform and then an ongoing basis for maintaining it and keeping your playbooks up-to-date and all of that. We have an agentic technology for those folks.

For your customer-facing teams, we have agents that help you engage with your top-tier clients and the furthest out on your long tail and everywhere in between. And then most importantly, we have agents that at your discretion and under your control will interact when and how you want them to directly to your customers. So again, feel free to stop by the booth and talk about that. Now, anybody can get up on a stage and talk to a slide and make claims and value propositions about all the great things that AI is going to do.

And we're all in CS, and so by definition, we are risk managers where we're used to managing risk and we know that one source of risk is buying into hype. So I'm not going to ask you to buy into any hype. I'm going to hold myself accountable to proof points. And the proof point I want to show you is one of our customers push operations.

They are a payroll and HR solution provider. They actually have a lot of businesses. I don't really understand the mix of their businesses, but those are the main ones. And they're an early customer of ours and an early adopter of our AI technology.

And they have cut their onboarding time by 61%. They've increased the accuracy of their MRR forecasting. They work on a monthly basis. And they've improved their customer adoption rates.

Now, Kelsey is the manager of professional services. She manages all the implementations there. And she says that her team is going to spend half as much time figuring out what's happening across their book in twice as much time on real customer relationships. Now, what I'd like each of you to do is do a little mental calculation.

What is the hard dollar value of half of your team's time? Now, imagine you had discretion over what you could do with that amount of money. What would you do with it? Would you reinvest it into your CSMs to have them go even deeper with their partnerships and their relationship with your customers to maybe accelerate your expansion cycle and bring the revenue in quicker?

Or would you have them spend time going deeper on the adoption of your product and maybe get increased retention and renewal rates from that? Or would you take some of that money off the table in the form of cost savings? Or maybe a little bit of each, right? The point is half your team's time is not a soft number.

That's real dollars. And you can do a lot with it. So this is one real customer with real business really using our tools to get real ROI. No hype, no promises there.

So we've got lots of these customers. I don't have time to talk about all of them. But that's one of our favorites. Okay.

That brings us to our next poll. So get your slide already. The question is, after this session, where does agentic AI sit on your radar for onboarding? Is it something that you're already doing?

Something that's a high priority that you're working on? You're interested, but not quite sure of the ROI. Or maybe you're not convinced at all and you have more questions than answers. Oh, there we go.

Okay. So already exploring building is winning. And that's good. Regardless of where you are on the adoption curve, the thing that I want to reiterate that's important to take away is that the time for getting started building your strategy and rolling out whatever you're going to roll out is now the teams that are leading the way in 2026 are going to continue to widen the gap over the teams and the leaders that are laggard in adopting the technology.

Now, the question is, while we're still in an AI hype bubble, how do you pick a platform? How do you know, we all know that eventually the bubble will burst and a lot of the players will consolidate out and there'll be a few left that are the real value providers. How do you identify who those vendors are and how do you choose to pick? So we want to try to provide some assistance in that selection process.

And so we've built an RFP template. You can scan the QR code and you can get it. Or you can talk to Angelica over here and she'll give you a little card or you can stop by our booth. By the way, stopping by our booth has another benefit.

We have really delicious cookies there and Legos to build race cars. So there's a lot of reasons to go to the booth. Anyway, we've built this RFP template and it's built around five areas that we think are the critical things you should hold any vendor accountable to when you're exploring platforms. So the first is, is this solution built for all customers?

All right. The days of having and have-nots in our customer bases where some customers got benefits and others didn't, that's the characterization of past errors. In the modern era, every customer from your top dollar customer to your longest tail customer should get the benefits of your agentic AI solution. Make sure that whoever you're considering is taking your whole customer base into account.

You don't want to be in a position where you have to have multiple tools for multiple parts of your customer base. Make sure it's inclusive to all the stakeholders. Those ops people, they're really important. Your IT people, they matter.

Your teams, of course, you always think of them, but your customers as well. For a long period of time, we've bought and deployed technology that provided benefit internally to our teams and cost savings to our own P&L, and we just sort of took for granted that those benefits would somehow indirectly iner to our customers. In the agentic era, you should have, certainly you should have control and choices to allow you to determine exactly when there are interactions with your customers, but your customers should have the option to take advantage of direct benefits from the technology as it matures. Make sure that whatever platform you're considering is going to be available to all of the stakeholders.

Make sure it looks at all the lifecycle stages. We all could recite the typical customer journey, right? It starts with the handoff. It goes through onboarding and implementation into adoption.

You have a renewal cycle. Hopefully you have expansion. Sometimes you have churn. If I had another hour in another breakout room, I'd give a whole presentation on my views of where lifecycle stage is going, but I think the hard boxes with handoffs that we have today, I think that's going to be a thing of the past, and we're going to see that modern technology and ever increasing expectations from customers is going to lead to more of a continuous lifecycle with a consistent experience across all of it.

And so you may not need to be in a rush to totally blow up the way you manage your customer journey, but when you're looking to partner with a vendor who's going to bring you into the future of CS, you definitely want to make sure they have a strong opinion and an understanding of where the world is going and where lifecycle stages are going, and then it aligns with your view and that they're going to bring you to where you want to be. Safety and control. So AI is not a fully mature technology. Anyone who's used it knows that.

It's only as safe as the context that you give it and the controls that you put around it and the way you allow it to access data and what you allow it to do. You probably should stay away from platforms that someone just vibe coded up a demo in a weekend and then started charging for it. You want to find a vendor, and there are several of them, that have a deep understanding of how to deploy agentic AI in a safe and controlled way where you get to stay in the loop when you need to be, and the AI only can be autonomous when you want it to be. Safety and control, it's very important.

And then we talked about the real proof points. Think about what the ROI is that you want. Do you want accelerated expansion revenue? Do you want increased retention rates?

Or do you want cost savings? Or a little bit of all of them? And then look for real customer proof points that are using the technology you're considering getting involved with and make sure there are customers, real customers and real businesses, getting the real ROI that looks like the same ROI you want. And that's what we recommend you work with.

So that is the end of my presentation. Thanks for coming and thanks for staying. There's nothing worse than being in a presentation where everybody walks out. And so please stop by our booth.

Our head of CS is there. Our head of product is there. And we've got some sales as marketing people as well. I'll take some questions now.

And then any that we don't get time to, you can come find me at the booth and I'll answer the rest. Thank you, Frank. I also loved how you put all of your tenure in eras. And it really made me think about what the next era will be after this whole AI era.

I'm hoping it's retirement for me. You and me both. You and me both, man. I keep waiting on my husband to make more money.

But has not happened yet. Okay. We have about 10 minutes for questions. If you all have anything else that you want to ask, feel free to see Frank at his booth.

Number one, how did you initially make the case to test out a CS team? And how do you justify the cost of the CSM team now? Very differently. So when I say that it was a hallway conversation at HubSpot, it really was.

I was walking down the hallway in our offices in Cambridge and I bumped into the CEO. And he's like, man, we've tried so many things. Nothing seems to be working. What do we have to do to get this journey done?

I'm like, I don't know. Maybe we should build a CSM team. And he said, well, how many people do you think you need? And I'm like, I don't know.

10? He's like, well, I'll give you budget for nine and you're the 10th. That was how I justified it back then. Today, it would be a lot harder.

I think the way to justify the cost of CSM team or any post sales investment today is you need to make the case that you're going to move a number that the person who's approving the budget actually cares about and that it's and you're going to achieve a result in a reasonable timeframe. The days where we could justify investment through soft metrics or improvements to the churn rate 18 months from now, that's a really hard sell today. That's why I really like the onboarding metrics. If you could get your customers fully through onboarding in half the time or when they get through onboarding, they're actually all the way at value.

And then you can translate that value into accelerated expansion cycles or increased retention in the short term. We have another customer who's not on a slide here. They measured their 60 day retention rate and they improved their 60 day retention rate by 76%. It's a B2C business.

Hard dollar metrics are the way to justify investment in CS today. Thank you. With AI agents, do you think we're at the point of CSMs headcount fully being replaced by AI agents or do you think that it's more of these AI agents are taking parts of a CSM job but we're not at the point where you can say, "Okay, we're slashing the CSM headcount by half." Or we expect them to take on two times as many accounts. I think that's a personal decision for each business.

I think today the state of technology is such that you'd have to have a pretty cold heart with respect to your attitude towards your customers to wipe out your CSM team and just throw agents at your customers for everything. Technology is not quite there yet. Now, will it one day, maybe not all that far in the future, be good enough to do that? It might be.

But the question is, what's the wisdom in that? Is that being penny wise and pound foolish? If what we do is we say we can use AI as the ultimate cost saver, well, once you've wiped out most or all of the team and replaced them with agents and saved that cost, where do you go from there? The AI isn't going to come up with the next great innovation.

The AI isn't going to figure out how to build a deeper relationship with the customers. You still need people. I don't know, maybe I'm an optimist, but I don't think the role of CSM is going away. I think the right balance here is to deploy the technology, whether it's AI today or whatever it is tomorrow, deploy the technology alongside the team, encourage the team to embrace change and give up the things that the technology can do better, and then use their creative skills to go deeper in the things that are higher value.

Yeah, you can get some cost savings too along the way, but I think a balanced approach is the way of the future and it will ultimately be those teams that deploy the technology alongside their teams that figure out the next innovation and the next best practice. I love that. I mean, you can't replace relationships for me as a CSM at GainSight. These events I absolutely love because I can finally see my customers in person and give them a hug and it's, yeah, you can't.

An agent's never going to replace me. I'm irreplaceable, right? Okay, let's see. In the agentic era, how do you envision how we onboard customers in the next two to three years versus historic approaches?

Faster is the first word that comes to mind. Think about the first thing that happens when you sign up a new customer. You sign up a new customer and there's some kind of a handoff from the sales team to the onboarding team. Maybe it's a quick meeting, most likely it's an email introduction, and then what happens?

Then everybody is playing calendar Tetris. You're trying to get a bunch of people from your team and a bunch of people from the customer's team to find time on a calendar. So much momentum is lost there. Then you have a kickoff call and in your kickoff call, you know, the first 20 minutes of the hour-long call is spent.

Everybody introducing themselves and then you rehash all the context that the salesperson could have given to you anyway and that's more time wasted. It can be two or three weeks before you really get going. I think that in the agentic era there can be parallel paths. You're always going to want to have personal introductions in a kickoff meeting.

But there's no reason why agents can't interact with your customers beforehand and keep the momentum going. There's plenty of things that can get done before the kickoff call. There's plenty of things that agents can do to help in between meetings, in between email responses from the CSM. And so, you know, we've been talking about time to value when we talk about onboarding for so long.

But for a lot of companies, it's not really time to value. It's time to finish this project. Time to get all the boxes on this checklist done. And maybe the customer has value and maybe the customer hasn't got to value.

How many times have you talked to a frustrated CSM on your team and they complain about how customers are getting thrown over to them after onboarding, but onboarding isn't really complete? I think one of the big things that changes when agentic technology is applied to onboarding is the onboarding is get done faster and they get more complete. Faster time to value for real. The 20-minute introductions that hit home.

Okay, last question. Are you seeing agentic customer journeys moving right deeper into the customer journey, such as renewal or expansion? Yes. Yes, we are.

We have customers that are seeing measurable acceleration of their expansion cycle. I'm embarrassed to say I can't think of the customer's name, but I know that we have one customer that has shortened their expansion cycle by a full three months. And three months may not seem like a lot, but what month is it now? It's May?

Yeah, right. So it's almost the end of Q2 for all of us on a calendar fiscal year. What if you could have all of your Q3 revenue, expansion revenue, accelerated into Q2? Who would like that?

So I think that that three months is a lot. And yes, we do have some customers that are moving right. Now, just in full disclosure, we have some customers where the politics of moving from one stage of the customer journey to another is a big thing to overcome. So this is definitely organization-specific, how far you can take the technology through all of the different parts of the customer journey.

But we do have some who are moving right, and they're seeing a payoff. Perfect. Well, thank you, Frank. You've been absolutely amazing.

For the rest of you, lunch is going to start in a little bit, and then it goes for about two hours until 2.15. If you haven't checked out his booth in the expo hall, please do that. And there's puppies out there, so feel free to go play with them. And thank you all for your time.

Thank you.