Turning Conversations into Intelligence: Protecting Renewals Before It’s Too Late

Yeah, so in addition to those fun facts, I am Adam. I'm one of the solution architects and senior Gainsight admins at Rockwell. I've been at Rockwell for the last three years and in customer success for the last 11 years now in various roles, in customer success as a CSM, managing CSMs, CSOps, and as a Gainsight admin now as well and really passionate about CS. I'm Dave Newman.

I've worked at Plex Systems who was acquired by Rockwell for 16 years. I started out in the professional services world doing implementations of our software, moved into support, and then over the last six years have helped build our customer success team. Also I'm very disappointed. Adam had promised me he was gonna dress up his short round from Temple of Doom and here we are.

Yeah, we were supposed to coordinate and then I dropped the ball. So I guess one more thing to complete the look, but this won't last. It's way too hot. Amazing.

Anyway, what we're gonna start out is I'm gonna, well, I'm actually going to take a poll first to get an idea of where you all stand with your AI journey. In the question it specifies their case AI but really if you're using any AI to help with renewals, you know, feel free to answer with those. And while we're waiting for that, I just wanted to explain like the whip. I do not own a whip.

I borrowed it from a yoga teacher. She's a very intense teacher. And when we get started, I'm gonna go through a little background of Plex, the company, some of the changes we made, and basically our journey in detecting churn, essentially. And then Adam will take over and kind of talk about some of the solutions we put in place.

Awesome. Okay. So it looks like most folks are preorganized and collecting information in AI adoption today. Cool.

Yeah. That's about where we expected. So all right. So let's get started.

So Plex systems is an ERP system that was built for manufacturing from a manufacturing company. So we started out basically an IT department that was building solutions to run the business. And what they built eventually was this comprehensive system that they decided that they would be able to sell to other manufacturers. One of the differences that Plex had versus our competitors was that it was built from the shop floor up.

So it was very good at like the MES aspects of manufacturing. But at the top floor type activities, it was really geared towards small to medium sized manufacturing. It's a comprehensive system. If you purchase Plex, you purchased everything.

There wasn't a breakdown or pieces. You just got the whole thing. And then we had a SaaS model where one of our differentiators as a business was that when we signed a new customer, we didn't force them to change business processes to match the software or to work with software. We would basically take their specs and build our software to match their process.

And the result was we had almost evangelical loyalty. NPS scores off the charts. And we basically didn't ever have to worry about churn. It just didn't happen.

But there were challenges that went along with that. It didn't take very long to realize that if we did everything for everybody, we were going to break the system. We had one code base. And we got to the point where we had to get control of the changes that were made.

And so we had that challenge to deal with. And then the second challenge that we had to deal with was we wanted to move up market. And when we were selling to small to mid sized manufacturers, it was very easy. We had software that worked well.

When they got into more complex situations, like hybrid systems where it would be Plex running a shop floor and SAP running the rest of the business, we ran into struggles. And we needed to resolve that. So what we did was we created a product-centric approach. But along with that came some challenges that we hadn't experienced before.

Reduced flexibility for customers. Now we had to actually have them change business processes to match software. We had to deal with the modular complexity. We had to take a system that was highly integrated and break it into pieces and build the integration for those pieces to connect to other business systems that customers are using.

Sorry, I need some water. Always prepared as the inventor, right? And so from the sales standpoint, we had to change our process from selling a fully RP system to being able to approach sales of the land and expand strategy. So we go in with a smaller piece of the product.

And then the idea was we'd prove it out, show the value, and then expand. And then finally, the thing that we hadn't had to deal with before suddenly became a reality. Turn risk is much more feasible when you're selling pieces of a system rather than a full system. Just to give you an idea, we were a very sticky software.

If I had to use an analogy, have you ever had a Jolly Rancher and you bite into it and then you try to open your mouth and it pulls the fillings out? That's how sticky we were previously. So part of the process was we had to kind of re-envision what customer success was because we didn't really have a customer success team. We had an inside sales team and their job was mostly relationship building.

And the expansion opportunities didn't really exist because if you bought Flex, you bought everything. With the modular approach, we had this new pathway where we had to show value along that path so that as they did expand, those new sales and new expansion were easier. And ideally, this is the way that we wanted things to look, where these set milestones and the path was easy and clear. The reality looked a little more like that, mostly for our team, not for the customer.

And what we realized pretty quickly was there was a lot of information we needed to be successful in helping those customers achieve their goals and reach the... and for us to understand how they are performing. One of the first things we needed to understand is the adoption of the product. Again, previously, if you bought the product, we weren't too concerned about what you used in it.

It was just if you used it in your business running, that was great. We didn't have a good idea of how our customers were performing against peers. We suddenly had this need to understand and recognize expansion opportunities that we didn't have to in the past. The churn understanding became critical to our business because for the first time we started seeing it.

And then, of course, the big thing from a few years ago, segmentation strategy, reworking that. And then the other piece of it was 70% of our customers from an account standpoint were long-tail customers. About 30% actually had a name CSM on them. So some of the milestones over the last few years is we really started building our new customer success team in 2019.

We were, again, we were wanting to do it in a value-based CS strategy. We had the luxury of having an auto-renew system, so we never had to have CSMs dealing with the renewals and all the pain that goes along with that. We implemented GainSight in 2019, and we started to expand the team. We started adding some CSMs.

Some of those inside salespeople moved over to our team and became those CSMs. And then we had further expansion in 2020. I came aboard in 2020 for the team with the task of implementing our digital strategy. And then, yeah, when I say we expanded CS operations, we added me.

And then we did some preliminary things. We added some implementation health scoring. I'm sorry, we implemented health scoring. So used the scorecards in GainSight, had the team using that.

And then we started to do, like, a first pass at churn tracking, so we had better understanding when we were doing analysis. We created some reason codes. And that did help us. We got to develop some decent reporting.

Then we moved into kind of the business intelligence phase, and this is where we were trying to get into the adoption and usage tracking. So we added GainSight PX, and we started monitoring our customers' use system. We added an industry benchmarking exercise, where because we have all the data in our system, we went out and actually started pulling and calculating how our customers were performing rather than having to rely on them to tell us. So we could see what their on-time shipping was.

We partnered with APQC to bring in industry benchmarking. That then gave us some metrics to tell which customers were performing well or not. And then we would use that with the adoption monitoring to help make suggestions for customers and how to improve their performance. Implementation health tracking was another big piece.

When we had partners implement our software, we had very little visibility into how those implementations were going. So we used GainSight to start tracking that as well. And then in 2024, we had a big segmentation strategy. We had previously, or segmentation previously, was built completely on ARR.

So if you're a large customer, you got a CSM. If not, you didn't. What that ignored was things like expansion opportunities with the land and expand. We looked at things like the customer complexity, the offering complexity, like what parts of flex do they own?

Is it easy to maintain or is it difficult? We looked at things like regional differences and then the industries that our customers were performing in. And then finally, 2025, we did our first implementation of staircase AI. And we started kind of poke around at it, started using it a little bit.

But the big project I worked on was redefining reason codes again, the initial set didn't give us the insights we were needing. Let's see. And then we also were working on developing the mitigation strategies that were tied to those reason codes for those reason codes that were controllable. And then we did a full reworking of our churn reporting and the meetings and things that go along with it to talk about when we did recognize that there was a risk, how we were going to mitigate it, and then reporting it out to the rest of the organization.

So before I get to the results of that, we have another poll just to kind of get some more insight of where you guys stand. And the things that we had done, they did give us really good insights on what has already occurred. But we were still struggling with, and we'll see this in just a second. Okay.

That was the first one, I think. I don't think that one is our poll question. Yeah, it looks like it's saying 211 is the room. We're, is it 105 here?

Yeah. It'd be a different. Sorry, we won't worry about the poll for now. It was just kind of just a little add-on.

So what we what we quickly realized was that, there we go. Hold on. When we go back to, oh, there we go. That's more like the Indiana Jones questions we were anticipating.

Perfect. Yeah, so it is a themed question, of course. So that's interesting. It looks like some signals, for most of you in the room, some of you are seeing some clues in your torchlight.

So some signals are being captured today, but it's not yet consistent for you. And looks like, you know, some of the early adopters are maybe already all the way down to the bottom, already holding the idle conversation signals reliably driving action, not surprising, just given where we are. This all being fairly new, things with staircase and predictive actions through AI. But yeah, makes sense.

Okay. Switch back to the other slide. So then what we found was, like, when there was a CSM on the account, we usually would get pretty good insights in seeing churn early enough to react to it. But there were things that were still an issue.

So human connections can create blind spots. So there's when you build good relationships with people, sometimes you don't want to share bad news with the person that you're working with. So we had a little bit of that. Our health scoring was very manual.

It was subjective and manual. There were some automated components to it that were objective, but for the most part it was a manual health scoring. And then there was inconsistent approaches. One CSM to the next would handle the way they figured out that risk.

And then there was a big problem of scattered input. Since we started out as such a small company and we retain customers for so long, people would have, like, inside contacts within Plex. So rather than reaching out to the CS team, they might be going to sales, product, engineering, support. All these places might be getting things that we should be able to see and detect and do something about, but we weren't seeing it.

So the result was we were still getting churn that was blindsiding us. So based on that, we're going to have Adam take over on how we resolved that. Yeah, thank you. So how do we take the business context from signal to action?

So for us, it's all powered by Staircase. Surprise, surprise. It's for us really the insights engine behind a lot of the things that we're building and a lot of things that we're doing right now. So for us in our business, we've integrated Staircase across almost all of our customer-facing touchpoints.

So all of the teams where they may be interacting directly with customers, we want Staircase to be integrated and connected and collecting conversational data to help us understand the whole picture that the customer is experiencing from working with both Rockwell and Plex and our other business units. And so that allows us to provide real-time end-to-end view of the customer's experience. And the best part of this as a systems admin guy, we can do this automatically without any manual user input or adoption required. And so we've implemented this within our sales team, our customer success team, our professional services teams, and our customer care and technical support teams, with the idea again being that anybody that is touching or communicating directly with a customer, we want those conversations being collected and recorded through Staircase so that it can run its analysis.

All the analysts you may have heard about or learned from it can run all of its magic from all of that. So for our business, this looks like our Outlook emails and calendar events, our Microsoft Teams transcribe meetings, and our support tickets and chat. The Plex business today is using Salesforce cases for that. Today on the Rockwell side, we're using another tool that will be migrating into Salesforce cases in the future.

And so this allows us again to get that full picture. And what we do with that then is really nothing revolutionary, nothing really too innovative, but we've allowed, we take that information from Staircase as soon as it detects a churn risk signal and integrate it into Gainsight. And then from that trigger a risk CTA that gets assigned to the CSM. And basically, like I said, integrates it directly from that turn risk signal detected by Staircase completely out of the box.

And I was chatting with some people last night and asking like, do you have to train the models? Do you have to configure or do anything for Staircase? No, the beauty of this tool is that a lot of this is doing it right out of the box, just by learning and absorbing and ingesting all of your conversational data. It's learning, you know, what for your business is going to be considered a churn risk, what has happened in the past, and all that.

So that's really beautiful to be able to have it kind of right out of the box as an admin. And so once those risk CTAs are then created automatically and assigned to the CSMs in Gainsight, we're also adding them as a human in loop to basically go into the CTA and validate that the AI got it right. So we've got a Boolean field in there just to say yes, no. Does the CSM agree that this is a legitimate AI risk?

And from there, then they have to also add in a few manual fields that are required. Things like how much revenue is at risk? What is the churn reason code or main reason? And what's the sub reason for it?

And then if there is for the Plex business, an expected like cancellation date or churn date, where we may have that. And once we have the reason codes populated, we'll also the rule actually goes back and runs a playbook. So that it's going to go back and basically apply a playbook based on the reason code. We're still developing the playbooks, but that's kind of where we are right now in this journey is the churn risk CTAs are created and firing.

And so as CSMs are populating that, we'll be able to go back and basically learn from the AI models, you know, what are the most effective strategies for our business for Plex to be able to run those playbooks against as well. And so this automated solution of taking the data from staircase and then driving it into a gainsite risk CTA ensures that no customer signal is missed or overlooked when they're explicitly telling us language that is deemed by the AI to be a churn risk. And yeah, so those again, the CSM validation steps trigger that automation and allows us to keep that human in the loop. Here's a couple screenshots of what that actually looks like.

We do have multiple tenants of gainsite here at Rockwell, Plex being one of them. And so this is what it looks like for their environment, where we've got the Boolean field, the CSM can say that, yes, I agree, this is at risk, I can include the at risk reason, the at risk sub reason, and the at risk amount in there as well. And so here is for anybody who likes process maps, here's kind of a visual of what that looks like from start to finish. And the cool thing about this is for Plex, anyways, we weren't starting from scratch, they already had some churn forecasting, some horizon nutrition processes that were already in place.

So it allowed us to be able to add AI and technology on top of existing processes and enhance it and improve the process overall. So some of these, the start items are new items that we've introduced in the last few months, where staircase again is the identification, it's the insights engine for us, helping us to just listen to those conversations, listen to the communications for those signals, and be that insights engine that drives everything and the human action off of it. So basically, if there's risk data entry, you know, then that's when a playbook can be fired. If then there is a an amount, the actual how much revenue is at risk, that also then gets added into the forecasting.

And it actually also CSMs can put in like a confidence level to say how confident are they that this amount is going to churn. Once it's over 75%, it's actually going to automatically create a churn opportunity in Salesforce for us, just again, using rules engine and gain site, that allows us to then turn this into something that's in Salesforce for more legitimate, you know, forecasting, reporting, financial reporting on the business on the commercial side of things. So again, we're taking it from a multi, it's a multi system solution that we've taken here from staircase through gain site and into Salesforce. And that allows us to get everything from the actual communication point all the way into forecasting data that our finance teams on the plex side can then use for their forecasting and budgeting conversations.

Some other really cool new processes that we're going to be introducing after churn has occurred, really important to make sure that we've captured the full context from those customers, really understanding why they chose to leave, you know, what we could do differently in the future, and really capture rich context that may not have been captured through the initial conversations, just by simply asking more probing questions or curiosity questions, things that open up open ended questions that open up more dialogue and, and feedback. And so we're introducing some post churn customer interviews as well. It's not going to be done by the CS team, because obviously they have a prior relationship, perhaps prior biases, or other things that may impact the ability to get a full picture. So we're having a third party and internal group that is going to be able to help us to conduct these interviews with customers and kind of play, you know, of course, they're still employed by by Rockwell and the team, but they are kind of more of a neutral party and trying to collect a bit more context that we can then document to have a fuller picture of everything that may have happened for customers that are interested in doing that with us.

And because of the power of tools like staircase, you know, Gainsite, all the AI stuff that they're doing, and Salesforce as well, we're going to be able to also do deeper horizon, or sorry, churn analysis data, churn analysis reporting so that we can actually better understand from all of the data that we've collected, you know, what's really going into what's causing churn at kind of the root cause level. And so it is both a systems improvement, technology improvement, AI improvements that we've been able to do, but it's also human and process improvements as well. Time for the third poll. So what's the trickiest renewal risk signal to spot before the boulder starts rolling?

A couple of results starting to come in here. Interesting. Okay. It's really switching.

Stakeholder drift champion disappears. Okay. It looks like now those are the top two. So value gap hints.

So in that case, anyways, tools like staircase are going to be able to help if the customer is communicating it. So I think, for our CSMs, this is also an enablement and training opportunity to make sure now that we know these tools are there and they're listening and they're looking for these signals. Of course, you have to have the customer actually express them to you or give them, allow the AI to detect them. So it's also a human training opportunity and making sure that we have humans asking the right questions, asking the powerful, sparking the powerful conversations so that AI can pick up on these things as well.

It's really interesting to see those two coming. Yeah. The stakeholder drift was interesting to see because that is something when we first started seeing churn and kind of offering these smaller point solutions that were easier to leave. There were fewer people that we had as contacts and fewer people using it.

And if somebody left, it was really easy for them just to leave the solution altogether. So. Number two for the last couple of wins. I think our last slide is just a little, it's going to be a little bit about lessons learned.

I will say the one thing is staircase AI depends on conversations, right? So like without those, if your customers aren't engaged, it's not a tool that's going to be extremely helpful in detecting churn. Can we switch back? So basically, tip one was expanding the reach of staircase AI to the rest of the business.

Don't limit it to just your own customer success team. I'm going to skip the second one for a second because I want to focus on a little bit more at the end. Establishing all that work we did up front to establish the processes and the structure and the strategy, all this stuff makes the AI piece more effective. So like spending the time there is worthwhile.

Starting with human evaluation, of course, we did get a lot of false positives to begin with, and that's something that over time we're going to continue to work out and make better. But establishing a robust digital strategy is the key. If we want to get insights on customer sentiment through a conversation agent, we have to have those customers engaged. So that's something that we are we have a decent start to it.

We are really pushing it. That's going to be a big focus over the next year, is getting that to be a much more engaging digital strategy. That way, we're encouraging people, we'll get that information, and we'll get better decisions out of it. Just to add to that, what that really means is, you know, think of your light touch, customers, your long tail.

They are customers who are typically also, they may be less engaged with your CS team, either because you've made the economical decision that you can't support them as much with a high touch experience, or because maybe they're just not a big spend and they're not super invested in your area or your products and so on, for any reason like that. And so naturally, there are less conversations happening. So for a model like Staircase, where it's listening for a conversation, there's the absence of a conversation. So you have to kind of introduce new ways where you're actually engaging these customers and getting a conversation started.

And so that's what, you know, David's really mentioning there, like looking also not just in terms of your communication places like with CS and the customer, but also like where else maybe in the community in with your support teams in, you know, of course, in their product usage, if you've got telemetry data as well. These are the places where you have to have that whole picture to really understand it. What Staircase is really doing is listening for communications data, conversational data, and then using that to analyze and listen for turn risks, sentiment risks, cancellation notices, you know, personnel turnover, commercial conversations, is listening for those types of signals. But again, that assumes that there is a lot of engagement and conversation already happening.

So if it's not through your CS team, you got to look to other avenues and make sure you've got that full picture wherever your customers may be going and engaging with you. Make sure you've got data to look at that as well. So again, maybe it is none of those places. And it really is just telemetry data.

And that's all you've got. And that's what you got to use then. Yep. And like the key with any of those things is whatever you are producing, like make sure it's going to the right people at the customer.

You don't want to blast somebody in finance with a quality announcement or something. So just make it something that when they get it, it's valuable to them so that you're not unsubscribing. It's getting them to start asking questions back. So yeah, that's where we get into like the usage and benchmarking data, kind of showing them where they perform, showing them where they're falling behind industry peers, or even flex peers.

And then automated business reviews over time will be continually evolved to be better, giving them better information. All right. Awesome. So that's it.

Looks like we've got 12 minutes left, I think, for some questions. Katie, what do you think? Yeah, we'll do about 10 minutes. So there's a little bit of time there, but you have a lot of questions.

So awesome. Yeah, bring them on. Thank you so much. That was great.

I love the first question. So can you share an example of a churn that blindsided you and how changes you've made would have helped in avoiding that churn if it had been in place at the time? All the initial examples, I'm thinking of the ones that blindsided us were things like mergers and acquisitions where they were being very protective of that information. And we didn't find out until too late.

That's not one that this is likely to capture. So I'm trying to think of another example. I can think of many, not necessarily specific to the Plex part of our business, but across both staircase tenants that we're looking at. Part of evaluating staircase and evaluating your implementation of it is to know, like Dave mentioned, the accuracy, going in and checking, like, did it get it right?

And so one of the things I've done is looking at a lot of the customers that we've known to be who have had a close-loss opportunity or churned, either a part of their business or maybe even the whole logo in some cases, and then going into staircase in those examples and seeing what did staircase see and would it have been helpful? Did it get it right? Did it provide any insights? And the interesting thing that I found is in almost every single example, staircases lit up like a Christmas tree.

It's all, you know, reds almost across the board in a lot of cases. Constant recurring themes of extremely negative sentiment, maybe personnel changes at different points. For anybody that does have staircase and has seen the, like, journey map, a little widget, that's the one that I use the most for looking at these kind of things because it visualizes all of these signals that you've got over a timeline to be able to see, you know, when these moments are happening. And so it's interesting because there is some kind of a correlation there, looking at a lot of that.

So, you know, to the question there, the changes that we've made by getting these things, you know, churn risks in place, oftentimes, you know, this is the starting point of this process for us, right? Oftentimes the churn risk when it's detected, it may be still too late to intervene properly, but this is really our first signal that we're getting, that's a clear signal. And then what we want to do is be able to eventually work backwards to intervene earlier so that it becomes a non-event or a non-issue for that customer, right? But right now what we have is kind of the highest priority, the most urgent signals that customers are sending us.

We're going to start there and work our way backwards into, you know, extremely negative sentiment and eventually other signals that help us understand when that moment that customer is really headed down the wrong path. Awesome. I was going to add as well with that, that if you are not a staircase customer, but you are considering it, we can actually put in data from the past. So, we always show, you can show value pretty quickly.

You would input data from a customer that did churn, and then you would run that through staircase to see kind of what staircase would have picked up on. So, just an idea there. Do you, with the question, the first one, do you want to answer that? If not, I'm happy to answer that.

Oh, sure. I can, yeah. So, I would say, I don't know a percentage offhand, but... I actually read the...

Oh, yeah. Good call. Sorry. How often does staircase get it wrong and CSMs are manually overriding the risk generated?

Is there a way to retrain the out-of-the-box ideology to avoid that? If so... So, I can say that initially we would have had a lot of hits that were false, and it was because of some technical debt we have with our engineering team and the influence that cases have on the sentiment. So, Adam did some work on configuring the AI to segregate that data into a separate bucket, and since then we have had fewer false positives come up.

Essentially, specifically, it was overclocking and flagging extremely negative sentiment that was happening during support experiences. And if you can imagine, when you're reaching out to support, you're not having a great time. You're having a bad... Something's broken.

There's an issue. There's something going on. Of course, it's generally going to be a negative sentiment. So, what we did is created a separate signal that breaks out, like, CS-related sentiment and then support-related sentiment.

This helps us to be able to know when we want to know or when we want CSMs to be in the loop or notified or alerted or be taking action. And yeah, so it just helps us to be able to break that out a little bit more. And then the other point of it, yes, definitely you can retrain it. So, in any moment that there is a signal detected in staircase or information that's flagged by the AI, you can definitely go in and train it.

And so, any of the signals that gets detected on a communication, you can go in there and hit the little, like, delete button, and it just gives you a little pop-up box to say, like, why? So, you can provide context back to the AI model to help it train and understand, like, what's going on? Why did it get wrong? And how often?

I have seen some examples where it maybe just got the context wrong, didn't fully understand everything that was happening, or, you know, it was kind of a one-off for that customer. But I would say the vast majority of it that I've observed and heard from others is that it's getting the churn risk signal stuff correct. And that's encouraging, it's helpful for us to know that, you know, it's mostly right. And of course, as we train it and give it feedback, it's only going to improve.

So, right out of the box from day one, I would say it got it very close to right, like 80, 90% correct. Absolutely. And it's also churn risk is also friction. So, I try to help my customers understand that a churn risk doesn't necessarily mean that they are churning.

So, to enable... Oh, sorry, I missed one. Do you work with many European customers? If so, how do you handle GDPR compliance with transcriptions and emails going into the system?

Yes, this is a big one. It's a fun one. I'll answer just because the Plex team is primarily North American based, but our Rockwell team that has a separate staircase tenant, they do work globally. We've deployed staircase globally.

And so, this is definitely a big one. It took us... We deployed initially to North America, and then it took us a few months, three, four months to work through EU compliance, legal. We have an AI governance council.

It took us months of work going back and forth with them and with Gainsight and staircase teams to make sure that we were giving them the confidence that we would be in compliance. We were, and staircase is in compliance with GDPR, but we had to demonstrate that to our business and to ourselves. And so, we did that. And the other thing that's happening right now, if you're thinking about deploying staircase to Europe, make sure you understand any workers councils that exist, and you make sure you work through them first and foremost.

We are kind of working backwards now and going through one in Germany, where they've kind of clawed it back. And we have to now go through... In Germany, they've got full co-determination rights, the whole thing. But we're...

So, we're working through that now. So, definitely, there's a lot of considerations in the EU. Switzerland is also another place that's really cautious in thinking about AI and privacy and so on. So, definitely, if you're deploying any AI solution, including staircase in there, definitely think through that.

And the other point I would make in thinking about rolling things out is making sure everybody's aware of what this is, what's it doing, and making sure everybody feels comfortable. It can be unnerving or unsettling if your people are not used to having a tool like staircase monitoring their communications or looking at things. Even though the intention is good, it's good for our business, you got to make sure still that your people feel comfortable with why we're doing it, what it's going to give them, and how it's valuable for our business. I think we have time for only one more.

Can you see those? Do you want to pick one off of there? I don't want to pick the last one. Sure.

What do you think? They're all really good questions. I'll let you choose. Arthur, I'm going to choose...

Oh, it's gone. Oh, there it is. Okay. Does your team have any plans to use staircase?

Is it gone? Oh, they're top. Does your team have any plans to use staircase for growth signals, CTAs? If so, how, and what will those signals be?

Yes, excellent question, Arthur. I know you. It's not a plant, but thanks for the question. So, yes, definitely.

So, we've been working very closely with Brady, the product manager for Staircase, for a while now, and they've recently deployed the expansion agent as well, which is really exciting and a really cool new feature that they have. So, they were always listening for expansion signals, but they're doing it more with the analysts and the agents now. So, they give you a full breakdown of everything that's available. So, it breaks down by product and opportunity level, basically, today.

And so, within that agent, you can see how many opportunities are in flight or in conversation right now, which is great for all those salespeople who don't like documenting their opportunities until they feel it's real. This is now a chance to be like, "Here's what's actually happening right now, what's really in conversation right now." And it's great for CS to have better visibility to conversations that are happening across their account as well, so that they can be aware of and maybe an inform or consult or guide the salesperson and say, "Hey, there's some landmines that you've got to be aware of. Stay away from these types of conversations or these people or just be aware of some context." So, yes, definitely we're thinking about that. And we don't have that implemented yet for any kind of CTA process yet, but we're excited to continue exploring it and seeing what we can do with it.

Yeah, and the detect, I'll just add to that. I do a lot of our CSQL analysis or expansion analysis, and I was comparing it just for fun, like what did we have officially recorded versus what did that capture, and it captured nearly everything that we had. The communication that we were doing was already getting caught before the expansion piece you were talking about. Yeah, and so to that point, for Arthur, for other people, if you're thinking about deploying Staircase, also think about which teams are going to be the right ones to have it be deployed against.

You'll have to work with your IT teams or your groups, your security, privacy, whoever. Make sure that it is, for us, it's a set of permissions that have to go on to the users' email and Teams accounts, so that it's all captured from the backend, from the server and from their communications. And so just thinking through who are the right teams, if you've got solutions, consultants, implementation teams, onboarding groups, think of all the groups where you would want this type of tool to be deployed. Fantastic, thank you so much.