In this episode of Mastering CS: Candid Leader Insights, Irina Cismas sits down with Hana Gruzman, CS Operations and Enablement Lead at Agora, a real estate investment management platform that helps firms manage investor relationships, fundraising, and reporting in one centralized system.
Hana shares how she grew from CSM to Team Lead to CSOps within the same company, and what she learned at each stage about building structure without losing the human element. She explains how Agora scaled its CS department from 4 to 30+ people, why splitting CS into specialized mini-teams was a turning point, and how she approaches enablement in a product that requires deep domain knowledge.
She also opens up about the limits of health scores, how AI is reshaping the sales-to-CS handoff, and what practical skills CSMs should start building today if they want to move into operations one day.
What You’ll Learn
- What CS Operations and Enablement actually means in practice at a fast-growing company
- How Agora scaled its CS department from 4 to 30+ people without losing alignment
- Why splitting CS into specialized mini-teams was the key to improving quality and efficiency
- How to build a health score system — and why it can still mislead you
- How Agora is using AI to improve the sales-to-CS handoff and churn prediction
- What enablement looks like when your product is highly complex and domain-heavy
- What skills to start building now if you want to move from CSM into CSOps
Key Insights & Takeaways
Structure becomes critical at a specific inflection point. For Agora, the shift from 12 to 20+ people made it clear that everyone doing things their own way was no longer sustainable — consistency had to be built in.
Splitting CS into mini-teams by function unlocks better quality. Separating onboarding, support, and CS allowed each function to attract different skill sets and focus on what actually matters to each role.
Health scores alone don’t predict churn. A client with a perfect health score can still leave. The human pulse — what CSMs observe in conversations — is often a more reliable signal than usage data.
Use the data you already have before adding new tools. Agora found that AI applied to existing transcripts and internal data delivered more value than reaching for yet another external tool.
Enablement is not just training — it is a system. Quarterly self-assessments, monthly sessions, product release walkthroughs, and post-session tests keep knowledge sharp and measurable across the team.
CSOps cannot fix everything. Internal prioritization and collaboration across a growing team will always have friction — the role is about reducing it, not eliminating it.
Getting things done is what sets CSOps candidates apart. Initiative, data literacy, and the ability to spot and improve a broken process are the skills that move CSMs into operations roles.
Podcast transcript
Intro
Irina (0:05 – 0:32)
Welcome to Mastering CS Candidate Leader Insights, the podcast where we dive into the world of customer success with industry leaders. I’m your host, Irina Cismas, and today I’m joined by Hana Gruzman, CS Operations and Enablement Lead at Agora, a real estate investment management platform that helps firms manage investor relationships, fundraising, and reporting in one centralized system. Hana, I’m really happy to have you here.
Thanks for joining.
Hana (0:33 – 0:35)
Thank you for inviting me. It’s great to be here.
What CS Operations and Enablement Actually Means
Irina (0:36 – 0:48)
You’ve moved from CSM to Team Lead and now into CSOps at Agora. When someone asks you, ” What do you do today?” What do you usually tell them?
Hana (0:49 – 1:29)
It’s hard to explain because my title has a lot of words that most people don’t understand, but basically I build the processes for the department, like how CSMs should work, where they manage all their work, how they should interact with customers. I’m also in charge of enablement, so I need to know everything in the product really well and then teach all the CSMs the product. I also now have a solutions engineer that I work with for the more complex use cases, because our system has a lot of calculations and financial things. So basically it comes down to processes and knowing the product on a really high level.
The First Problem: Efficiency at Scale
Irina (1:30 – 1:37)
And when you first stepped into the CSOps part, what was the first problem that jumped out at you?
Hana (1:38 – 2:21)
The main thing that I try to solve to this day is around efficiency and helping the CSMs as the company grows. Because of AI and everything, everyone expects you to be more efficient. So I basically try to see what takes up the most of their time and how to help with that.
The first thing I did in the role, for example, was look at a certain type of customer that’s quite small and only uses one part of the system. So we moved that to be under support. Things like that, looking at what’s taking up most of their time and how you can help them be more efficient, while still keeping them at a good level, making sure that they do their work, and holding up the adoption and helping the clients with the regular things that CSMs need to do.
When Structure Became Non-Negotiable
Irina (2:22 – 2:41)
And as the team grows and everyone somehow starts to work differently, at what point did you realize that you needed more structure in the CS team? What was the point?
Hana (2:42 – 3:36)
When I started, there were only four CSMs and now there are over 30 people in the department, which is crazy.
Getting everyone to work in the same way is difficult because everyone has the way that they like to manage tasks and the way that they write emails. When I first stepped into the role and we started getting batches of new employees that we needed to train, that’s when we realized we need something more structured. They all need to work in a certain way.
I would say when we went from around 12 people to 20 plus, we realized we need more structured processes and for everyone to be doing the same thing. We very much trust the employees and that everyone knows how to work in the best way, but still people need to work in the same way. And we have two different teams, one in New York and one in Tel Aviv, and we need to make sure that each team works in the same way.
Splitting CS Into Specialized Mini-Teams
Irina (3:36 – 4:06)
And what were the first processes that you put in place while the team was growing?
Hana (4:07 – 5:44)
What we realized is that when we started as four CSMs, we were doing everything: onboarding, support, and CS. We realized that we need to make mini departments in our CS department that focus on each thing.
And it really worked. When we focused on something and we put people in that role, it got better. The first thing we did was split onboarding and CS. When I was a CSM, I was doing everything, spreadsheets and everything, I was just guessing. So we built an onboarding department that focused on onboarding. We brought onboarding specialists that have different skills than CSMs. They’re more detail oriented, good in Excel, more data driven. Whereas CSMs are more relationship and upsell focused.
Once you start doing that split, it also changes the people that you bring to the team. We then did the same thing with support. CSMs would spend a lot of time just answering how-to tickets and regular questions. So we built a support team. We also have additional services other than the SaaS, like tax and bookkeeping, so there’s a team that works only on that.
When you have something that A, takes up a lot of time, and B, needs a different set of skills, building it out as a different group in the department really helps. And then it’s about how they all work together: how does onboarding work with CS, how does support work with CS. Now we also have a solutions engineer that comes in for more complex use cases. So basically as you grow, you do need to split out, but still make sure that everyone knows how to work well with each other.
Health Scores and What Data Can’t Tell You
Irina (5:45 – 6:10)
Let’s also talk about some metrics. When you look at CS through the operational lens, what metric turned out to be misleading or not as helpful as people thought initially? What did you start with when you first put the structure in place?
What did you first track and how did the tracking system evolve while the team was scaling?
Hana (6:11 – 7:43)
One of the first things we realized we need to track is a health score. We built a health score based on usage through Mixpanel. We built a formula based on what we think is good usage, like if they use the waterfalls, reporting, or features that we feel really keep people in the system. There’s a score from one to a hundred, which is very easy to understand.
But it doesn’t always predict if people are going to churn or not, because someone could be using the system because they’re dependent on it, but they don’t like specific features or they want to leave or they’re very unhappy. You can’t always rely only on data and numbers. You need the human element.
So we also have something called the pulse, which is the CSM’s assessment based on what they think. Someone could have a hundred health score but their pulse is very risky because they’ve expressed frustration and want to leave. Numbers can only tell you so much. When we look back at our churns, we really saw that the prediction was based more on the pulse and not on the health score. Someone with a hundred health score can still churn.
We also track upsells. As a company grows, you can’t only sell to new clients, you need to also upsell within your own client base. That’s a skill we’ve been developing and something we need to monitor, because a company can’t only grow from outside business but also from inside business.
Tools vs. Process: How to Know the Difference
Irina (7:44 – 8:06)
You know what I’m curious? I often see that CS teams respond to problems by adding more tools. What’s a moment where you realized that we wouldn’t solve the issue or how did you approach the tooling part when you started adding structure in place?
Hana (8:07 – 9:41)
We definitely always get approached about new tools and people want to add new tools. But what we’re realizing now, also with AI, is that there’s so much data that’s being underutilized.
For example, Avoma has an AI feature called Ask Avoma, which we recently started using. It basically has all the transcripts of all the history with the client, even from sales, even before they became clients. So before looking at outside tools, we have so much data inside the system that we’re not using because nobody knows how to approach it.
I know AI is just a buzzword and I also don’t really believe in something until I can see that it actually gives me a good solution. But for example, what we did with AI was take all the churns from January, put them into Claude, and asked it to do a thorough analysis of why people churned, what the main reasons were, and whether we could predict it using the pulse and the health score. Then we looked at the current at-risk companies and asked, based on past experience, can we predict whether these clients will churn and what can we do to prevent it. We also realized there were some issues with onboarding that could be improved.
So I think the right approach is to utilize the data you already have with AI tools first, and then if something is still missing and you actually need more tools, look outside and see what tool can help you solve that problem.
Irina (9:42 – 9:49)
So how do you now test whether a problem needs a tool or a process change? How do you know?
Hana (9:49 – 10:42)
First of all, if something is reoccurring, like if we have one person that’s always doing the same mistake, then you know it has to do with people not understanding what they need to do. Like we had this whole process on HubSpot to do upsells and everyone suffered from it and did everything wrong. That’s when we realized this process is not working, everyone’s complaining about it.
So if it’s reoccurring with one person or with the whole team, you know something in the process doesn’t make sense. And I would say you need a tool when you don’t have access.
A lot of my job is to make everything accessible for the CSMs. I want them to have everything as easy and within reach as they need, whether it’s tasks from the product team or questions people ask them.
So if I can’t make something accessible, then maybe I need a tool. But if everything is accessible and they’re still doing it wrong, then there’s an issue with the process.
What Enablement Really Looks Like at Agora
Irina (10:44 – 10:56)
Tell me what enablement actually means at Agora, because it means different things in different organizations.
Hana (10:56 – 13:14)
In Agora specifically, enablement means having knowledge. Our product is very complex, with a lot of domains like reporting, waterfall calculations, distributions, managing investors, managing entities. There’s a lot of things that you need to know and keep in your head.
Enablement is first of all making sure that all the CSMs, but also onboarding and the support team, knows what they need to know within the product. Every two weeks we have a product release, so I sit with the whole team and go through it with them and show them how to teach clients about it.
Every quarter I also do a self-assessment where they vote on the domains that they feel they need help with. I take all that data and build a plan for the quarter. People are very busy, so I try not to have too many sessions, just once a month based on what the self-assessment showed. For example, a lot of people said the waterfall was a bit complex for them, so I did a thorough breakdown session on it.
I also do a short test on Google Forms after each session because sometimes people can just sit in a session and stare at the ceiling. I want to know that what I taught them actually went through. And you can see it later in the data too. The more knowledge a CSM has about a feature, the more they relate it to the client and the more the client uses it. Over time, usage of the waterfall feature has gone up because CSMs feel more comfortable with it.
But enablement is not only about the software itself. It’s about everything they need to know, how to do things in the system, how to work with other teams, how the tax and bookkeeping service works. I do sessions, one-on-ones where needed, and cover new product releases. I see myself as the person that needs to know things, so if I don’t know something I go to product or whoever owns it and ask them. I really want to make things accessible for the CSMs and make sure they have all the knowledge they need.
Building Bridges: Sales Handoff and Product Collaboration
Irina (13:15 – 14:02)
You mentioned that internally, the team is structured between CS onboarding and support. Those are the three mini teams that stay under the CS organization. And I assume that you have internal processes because you mentioned about the alignment between the three of them.
How does the CS team interact with sales and product? What processes did you build to make those handoffs work?
Hana (14:03 – 16:08)
Sales and CS handoff is something that comes up with every company. Currently we have a Slack channel where we get all the information about the clients. But what we’ve built recently that works really well is using Avoma AI.
Sales has so much information about clients that up until now we didn’t really get. We just got their meeting summary, and not everybody works in the same way. Some salespeople are very thorough and send a really good summary, and some would just send two sentences and we needed to do more digging on our own.
With Ask Avoma, you can ask a prompt and get a summary of the client, their pain points, everything that has happened from the actual transcripts of the sales calls. You can ask whatever you want: what about their documents, what are they onboarding from, did they have any deadlines. You can really utilize things that have already been said by the client. We’ve now implemented this as our new process for getting as much information as possible about clients before they come in.
When a new client joins, they get an onboarding specialist that helps them onboard all their historical data and they get a CSM. That’s where they need to work together. Even though we have different roles, we’re still very much one team. We all sit together, have lunch together, and we have people transferring from onboarding to CS and the other way around. They meet with the client, work closely, and have weeklies where they discuss their mutual clients.
The handoff from sales is being enhanced by AI, and internally we still work as one team with smaller teams within. We have a CS all hands, weekly onboarding and CS meetings, and support also works closely with us. When you sit together and work together all day, you stay very much aligned.
Irina (16:09 – 16:21)
And how’s the relationship with the product team? How does a product develop the product based on the CS feedback or how does it work? What’s the process?
Hana (16:21 – 17:58)
Yes. I’m sure this relationship is a complex relationship in every company. Yes.
The product definitely, I will say, listens to our needs. We have a request board through Monday where we have an application where we input every request that the client gives. And then we have also, as I said, because our product is quite complex, we have on the CS team, we have a domain expert that works closely with each product.
So for example, we have the CRM. So we have on our team, the CS, someone who’s an expert in the CRM and they work closely with the product of the CRM. And if I need something, I’m like, Hey, Yard, can you push this and this request?
And then she meets with their product. So that’s a good way because as I said, like there’s so much knowledge that you need to keep in your head. So we’ve really split it up in an equal way so that every CS who’s a senior, who’s come to, who’s been at the company for years and really has like knowledge is coupled up with a product.
So that’s one way. If we want, we have a lot of things in the CRM, we would go through our CS and they would work with the product. That’s one way.
The second way is through just product requests. And then for strategic clients, we have a different way of working where we have big clients that have a lot of requests. So we would usually just summarize them in a meeting, meet with the head of product and listen, Hey, this and this client wants these requests.
Can we push them? Can we, let’s say we definitely, we do our part. We push these requests on product as much as we can.
Of course they have their own prioritization and they have a lot of, a lot of other factors that we’re not aware of. Also sales comes to them. And also a lot of, a lot of other factors ask for product development, but I will say that specifically at Agora, like a lot, most, a lot of the product is built by client requests.
What CSOps Can and Can’t Fix
Irina (17:59 – 18:08)
What do people expect CSOps to fix that it actually can’t?
Hana (18:08 – 19:40)
So, yeah, as we said, everyone needs to work the same way, which usually they do. But sometimes like, I feel like I’m like the sheriff of the CS force them to do things like a product, just for example, a product manager wants a list of clients that were interested in sending checks because right now they’re working on the check feature. And then they send a group message and nobody answers them.
And then they’re like, Hey Hana, why is nobody answering me? I need, and then I need to go and start chasing the CS. So I do feel like sometimes like people expect to, everyone’s very busy.
Everyone has their own tasks. And then when you rely on other people, sometimes you need to activate them to activate them, which I do usually, but in the end, I’m not like, I’m going to sit on someone’s head and be like, why didn’t you send the check? So I will say just like the collaboration in a, in a company that’s growing and everyone has like a lot of tasks all the time.
So the internal prioritization of other people, I can’t control. I try to help. I try to book processes, but in the end, everyone just works in the way that they work.
But definitely trying to improve even that process. So another, for example, I feel like I’m like marketing a Voma here, but I was just been using it a lot lately. So for example, for these things, I told them, Hey, why don’t you ask a Voma who has been interested in checks?
And then again, it looks through all the tasks, transcripts, and it can, and it can say these and these companies have mentioned that they’d like to send checks through Agora and then they can get that data. So again, I try to take things off the CSM’s plate by giving solutions also inside the CS and also to other teams like product, sales, marketing. So I would say those are things that I’m trying to help with.
Irina (19:41 – 19:49)
You’ve seen CS from a few different angles. If I link to one system or process you put in place, what would break first?
Hana (19:50 – 19:51)
What would break first?
The Process That Holds Everything Together
Irina (19:51 – 19:59)
If you removed one process you’ve built, what would break first?
Hana (20:01 – 21:38)
Okay. Let me think about it. I would say the, okay, we have, we have an early adoption process.
So basically after clients finish onboarding, they go into a stage called early adoption where they have training with the CSM’s to use the system. The CSM’s meet with them. Usually it’ll be weekly and they train them on their different domains.
And then once the client has reached a health score that passes a certain threshold, then they move on to ongoing because it basically tells us that the client has adopted the system, they’re using it, and then they don’t need specific training. They can just go on to use the system. So I would say if we stop doing that, it would hurt the adoption because our product is very well made.
There’s a lot of domains, but it does sometimes need explaining because each client has their own different use case. Like no two clients are the same. The way that they manage their investors and the way that they manage their, that their investor investors is different.
So the CSM really needs to meet them and be like, Hey, you should do this and you should go about it this way. So I would say if there was no training process, we would see it in the decline of adoption of clients. So that’s one of the processes I think works well.
We monitor it. We’re able to see who does trainings. Also through Avona, we have a scorecard with that.
Basically, like I put in some thresholds of what a training session should look like. For example, did the CSM mentioned the value of the, of the feature? Did the CSM, did the CSM start with an opener?
Did the CSM ask correct questions? Like all these things that should be in a training session, and then that’s being monitored and it gives an actual grade after the session. And then we can see that those training sessions work well.
So I would say that’s something that, that has worked well. And if we were to remove it, it would be unfortunate.
How to Move From CSM Into CSOps
Irina (21:40 – 21:51)
The last question is for CSM listening, who’s curious about moving into the CSOps one day. What practical skills should they start building now?
Hana (21:52 – 23:13)
Wow. That’s a good question. I feel like even when I started as a CSM and there were four people, I was always like doing things like anytime something needed to be done on top of my CSM work with all my clients, I would just do it.
First of all, to just be a person that does things like don’t need to, of course you need to ask questions, what can be done, but like that’s already will set you apart from a lot of other people. So anytime, even like small things, a presentation needs to be made because you’re meeting with the CEO. So just do the presentation and come to your boss and be like, I made this presentation for the meeting or something needs to be analyzed.
Like someone asks, why did we have so many churns last month? And then you can look into it and use data. So if you can show that you’re technical so that you can use data that you get shit done, sorry, that you get things done because that’s people that usually get promoted.
And I would also say there’s a strategic element to my role. So as I said, like, how do we, how do we be more efficient? How do we reduce time that’s wasted on things?
Like really need to think of new processes. So if you can see something that’s not working very well, and you can think to yourself, okay, if we only did this and this, we could improve it. If we only had more data, if we only changed this process, if we only talked to sales and then actually do it, I think those are the ways to show that you have additional skills that could be used as CSOps.
Irina (23:14 – 23:36)
Hana, thank you for the conversation. It was great hearing what CSOperation really looks like in practice. And how do you think about building systems that actually support the team?
And to everyone listening, thanks for tuning in. Until next time, stay curious, keep learning and mastering customer success.
Hana (23:37 – 23:38)
Thank you.
