29: איך הופכים 30 אלף פניות חודשיות לשירות לקוחות לתובנות מעשיות (English)

Hi, Eran. Hey, Leo. Hi, everyone. You're at Startup for Startup, the podcast on which we at monday.com openly share knowledge, experience and actionable insights among startups. Customer success is the one team in the company that receives feedback from every shape and size of account, whether they're coming from trial, paying large or small. We receive roughly 30,000 tickets a month, covering many different needs and questions, and we answer each ticket as if it were our only user, with empathy and empowerment. However, at some point we realized that this wasn't enough, that we could use this information to improve the platform, reduce our ticket load and to empower our users even more. So essentially what we want is to share with the rest of our teams what users say to us about the features we release, pain points they might have, to show trends or correlations between what our dev and product do and what users are saying as a result. So the question is, how can we turn 30,000 tickets a month into insights that everyone in the company can work with? Today's guest is Efi Arman, head of customer success Israel here at monday.com. Efi will walk us through what we call voice of customer and share with us how this challenge of turning support tickets into actionable insights has been addressed. Hi, Efi. Hey, Efi. Hello, hello. All right, voice of customer. So it's a title that's thrown around a lot in this field, in SaaS and startups. What is voice of customer is the big question. For us, it's giving the voice to our customer. We have a product that we believe is for the people. And how do you build a product for the people? You listen to the people and let the people build the product. So kind of to bring back why this all started is how do you do that? How do you take what a person needs and build something? You know, I can say I know what I need to make my day work successfully and that's Aks and Iran can say he needs something else and that's why. And how do you measure that and how do you take it and actually. Right, and one of the ways we do that is just by the way our product is, which is very easy to set and people can just go and customize it to their own needs. But you're talking about a whole different level of. Exactly, so I'm not saying build a board the way you work. I'm not saying write an update that works for you. I'm saying let's actually go in the roots and the core of the product and conquer the pain points and basically bridge the gap between the company and the customer. We have a lot of channels that our customers can speak to us and we had that since day one that the product was built but we didn't do anything with it. We just kind of, like most companies, we said thank you for your feedback. We'll send this to the appropriate person and it went to the air. We're not a company that wants to be like that. We don't, we're not, we care about our users. You're here because you use it and we actually also use it and how do we say that? Right, and that's really the point that it's not just for the customer anymore, it's also for us. Exactly, but most importantly, like how do you make a product that works for everybody? You listen and you do something with it. So we had to take a step back and say, how do we take this information and do something? So the voice of customer started and actually started with Iran. So let's hear your voice then. Yeah, I think both as a founder and as somebody that's working on the product, I feel hearing feedback from customers is one of the most important things. It's surely very important when you start a company and when you start with the product, but it's never ending. Like you have to get feedback all the time. So when we started a company, it was very easy because back in the days we would even go and meet customers, go to their offices and get feedback on how they're using the product. Then we scaled a little bit. So I started reading tickets on like tickets that we got from customers. So I read- Sorry, emails, by the way. Yeah, tickets or emails. So whenever somebody sent us a feedback, I went and read that back in the days. I used to tag NPS tickets, but I feel most importantly, I felt like I was connected to the kind of vibe and feedback from our customers. But as we scaled- And as you were part of the people actually building the product, you could just use this feedback and directly implement it into what you were working on, right? Yeah. Which is really the essence of what we're talking about here again, because I think there are two parts and let's just make sure everyone understands it. One thing is making our customers happy. Right, and-

And listening to them for the sake of listening to them. But the next thing is making sure that our roadmap and the things we build next is really connected and aligned with what they need. It's actually making them happy. Which is the real way to make them happy, right? There's the idea of like, I'm going to put a smile on your face. Like I said something nice. I answered your email in less than six minutes. I gave you the right answer. But there's the like, getting you like genuinely happy where when you see you asked for something, you received it. It's kind of like Christmas every day. So, you know, obviously, you know, one point that's important to emphasize is that people might suggest features, but we really care about their problems. So it's not taking, you know, the feedback that they give as is, but trying to understand kind of what are the problems they're having. You know, as we scale the company, I think two things happen. One is that, you know, I have less time now to go over tickets, obviously, but I'm not the only consumer. So there's a lot of product managers that need to get feedback from customers. But also the volume of tickets is unbelievable. So we mentioned we have right now 30,000 tickets. Like no way, you know, a human being can go through all these tickets to get insights because, you know, limited in how much you can go through. So I feel, you know, it wasn't a problem when we started, but it became a problem as we scaled. And yet it remained one of the most important things for us as a company to get that feedback. Without that, you just feel detached from your customers. And again, you mentioned that you need this information for our product managers. And if you can elaborate on that. Yeah. It's not just for our customer success team to get this feedback. I think what's important is customer success is taking the information. We're kind of the middle person. Some of the information is for us. It's knowing how to write an article better or if we need to create an article or a webinar or something like that. But we're actually taking it deeper than that and influencing the product with it. So I'm not going to a meeting and saying, hey, product managers, I really would like it if you do something. I'm actually going and saying, listen, our customers are, X amount of customers want this. It's actually making them happy. Which is the real way to make them happy, right? There's the idea of like, I'm going to put a smile on your face. Like I said something nice. I answered your email in less than six minutes. I gave you the right answer. But there's the like, getting you like genuinely happy where when you see you asked for something, you received it. It's kind of like Christmas every day. So, you know, obviously, you know, one point that's important to emphasize is that people might suggest features, but we really care about their problems. So it's not taking, you know, the feedback that they give as is, but trying to understand kind of what are the problems they're having. You know, as we scale the company, I think two things happen. One is that, you know, I have less time now to go over tickets, obviously, but I'm not the only consumer. So there's a lot of product managers that need to get feedback from customers. But also the volume of tickets is unbelievable. So we mentioned we have right now 30,000 tickets. Like no way, you know, a human being can go through all these tickets to get insights because, you know, limited in how much you can go through. So I feel, you know, it wasn't a problem when we started, but it became a problem as we scaled. And yet it remained one of the most important things for us as a company to get that feedback. Without that, you just feel detached from your customers. And again, you mentioned that you need this information for our product managers. And if you can elaborate on that. Yeah. It's not just for our customer success team to get this feedback. I think what's important is customer success is taking the information. We're kind of the middle person. Some of the information is for us. It's knowing how to write an article better or if we need to create an article or a webinar or something like that. But we're actually taking it deeper than that and influencing the product with it. So I'm not going to a meeting and saying, hey, product managers, I really would like it if you do something. I'm actually going and saying, listen, our customers are, x amount of customers want this. And usually we're not going with a feature, a specific feature. It's like Yvonne said, it's a problem. So we're not going through each ticket and saying someone wants a checkbox. It's not about the checkbox. It's what does the checkbox mean? So someone must want some sort of to-do list. So how do you build a to-do list? Now we have to go deeper and say. And what would be the problem behind a to-do list if you take it all the way? Exactly. So what is a to-do list? It's someone is missing what they have to do when they come in or what they, what needs to be done in a project. It's taking a big picture and breaking it down into smaller pieces. So how do you do that and still fit our product? Do we want to, do we want to build a feature? Do we want to maybe enhance a feature? Do we want to say it doesn't work in our feature, but you can integrate with something else. And so that's kind of where it then goes into product and our hands are off. And it's now products, information and not customer success. Yeah. One thing that's important for me to say is that it's not only product managers. It's like people in R&D, it's designers that need to get that feedback. Sales team need to get that feedback. So everybody in the company, marketing, obviously. And the way I look at customer success is that I feel, obviously they need to make our customers happy, but a big part of it is communicating that feedback to the company. Because today they're the face of the company. They're bridging the company to our customers and vice versa. I think we're the ambassador of both sides. We're the company's ambassador and we're the customer ambassador. And it's up to us to decide when do you do which. There's sometimes people that come to us and they want something and we know, listen, it's just never going to happen. It doesn't fit our DNA, but it's on us to now help that person still succeed. Okay, so we understand why we want to do that. I think we explained that. And then it became a challenge, right? It was easy, as Ahan said, in the beginning. What was the challenge? Of scaling. Challenges came from a lot of different places. Like Ahan said, he was the one who started doing that. He was going through tickets and analyzing what's happening. We also started NPS, the Net Promoter Score, which is a way, it's basically a simple question, would you refer this product to a friend? Followed by, depending on what you give, you score it 1 to 10. Sorry, I'm saying this quickly. You score it 1 to 10. Depending on what score you give, you get a follow-up question. And the follow-up question is basically, that's the meat for us to find out what is it that we need to do. So traditionally, customer success is measured by NPS. In Monday, we're not measuring.

It's actually something that the company is measured by. And we decided it's actually a way to understand what customers want and let's use it for the product. And kind of going back to the scaling, also this was something that Eran did when each NPS score he went one by one and tagged what they wanted. In terms of their qualitative answer. So we were able to say was it positive or negative and then what was the topic that they were discussing. So is it billing? Is it a feature? Is it satisfaction? I think at some point we had about 20 or 30 tags. It's now grown to about 90. I just want to give you a specific example so people can relate to it. So just imagine somebody filled in an NPS survey and then we asked them how can we do better. So he might write something along the lines of I love the product. I think the user experience is very good and it's easy to use. However, I found the billing to be very tricky or hard to understand. So let's analyze that. So the customer talked about the user experience, the ease of use and billing, but a different connotation. So ease of use and user experience in a positive way and billing in a negative way. So at the end of the day, you're going to take this ticket and going to add three tags to it. So two positive tags, which we color in green, which is user experience and ease of use, and another one which is billing, which we'll color in red. Because the ticket might be like a very positive sentiment, but the details are very important. And something like that, that's kind of an example of where it affected product. But there's also a time where maybe they'll say they give a 10 and then we say, what can we improve? They say, my team just isn't using the product. So now that's customer success information. So now what can we do for this account to get their team on? You have someone who loves the product. So if you love the product so much, but your team's not using it, how can we help bridge that gap? So we'll put a negative adoption. Even though a 10 is the best you can get, it's amazing, but we actually have a negative coming from it. And that's something we can keep internally in customer success to improve. And each NPS score that is given comes into the customer success email feed and we respond to everybody. How do you do that with 30,000 tickets a month? So some of it is automated. If the answer looks like they just put a score with no text, they just send the number, we don't answer them actually with a human. There's an automated email from the Effie bot, actually. There are people now in the company who think I am a robot. I am real. So that is automated. We don't actually answer every single NPS ticket that comes in. There was a time that we did when we were not 30,000 tickets a month. Today, now we answer the ones that actually have information, like someone wrote something. And if someone writes something, they deserve an answer. And the point of us answering is that we can dig deeper and see that it's not just a red or a green tag, it's actually now we can get into the details and see what is your problem or what's your positive. Positive information is also very useful. If someone writes in that, like, I love the product and my team is adapting and it's amazing and I love it, I want to know how to do you get your team to adapt to this product because I can use that for me to continue and make other people adapt to it. And going back to the rest of the team, then product people can also use this information to make maybe some changes into the product. Yeah, but I want to be devil's advocate here because, you know, it sounds good, but it has a lot of problems with it. Which we're going to be dealing with right now. So why isn't it ideal? Why doesn't it just happen the way we just described it? A lot of different reasons. I'll go with the first one. You can't promise, you can't make everybody's wish come true. I can't be, you know, the fairy godmother who makes everybody happy. But how do I choose? How do I decide what is the biggest pain and what's the smallest pain? For example, billing has always been our number one. For negative tags. For negative, yeah, but not positive. It hasn't always been our number one thing we want to conquer. So sometimes it's the understanding. Is it pricing? What's billing? Especially on billing, just people find our billing confusing. The way we process? The way we present it because we have like buckets of users as opposed to price per user. So we have like a five people bucket, a ten people bucket, which works for a lot of users. But some people don't get like why are they like four users.

have to pay for five. So we created a lot of friction with our customers, and we're now in the process of changing that to price per customer. But also that statement of that's what it is, took a long time for us to understand that actually was the problem. A lot of the responses of billing issues was not that it's confusing, but they would say it's too expensive, or it doesn't fit my needs. But what does that mean? Like, what does that mean? And it took months, months, I think years actually, to really understand what were they complaining about. Today we aren't such an expensive tool in the market. If you look at our competitors pricing wise, in some cases we're cheaper than a lot of them. But there is, there's a problem. Yeah, but by paying for a 10 users account and I have six people on my team, then it might sound expensive at that point. But from our side, we couldn't get that. We didn't understand what that meant. And even still with going through the tickets and reading the tickets, and it still wasn't that clear. I remember that when we started to do tag those NPS tickets, we were very excited from the results because it was the first time that we kind of quantified feedback that we got from customers, both in positive and negative effect. So I remember the first time we processed that into a chart and it was amazing to see. First of all, like to see the negative tags or what are people complaining about. It was great insight. We got a lot of people complaining about billing. We got people complaining about our mobile app and other stuff. And it was great because we felt, you know, this is the thing that we need to fix and it helped us prioritize what we need to do. But also on the positive tags, we got a lot of good insights. So people mentioned that we're easy to use and that they like to use the interface and the tool is intuitive. So we actually took those messages and incorporated them into our marketing and homepage because this is how people describe the positive stuff about Monday. Right. But wait, you started saying before that this is not straightforward. So all I'm saying is that when we started, the immediate result was amazing because it was the first time we got an insight into that. But then, you know, moving forward a few months later, first of all, this charge remains static. So it's very high level, you know, it's billing. As Etika said, it's super high level. It's very hard to take action. So you might see that, you know, people mentioned the mobile app in negative text. So you might say, OK, I'm going to fix that. But, you know, a few months back. What does it mean to fix that? Yeah. What does it mean? And like we knew that, like it didn't help us. Right. It was very high level. It was very, you know, pointless in direction, but not really helping us in the day to day actions. And again, in a way, I felt detached from what happening because I got like a better picture, like with more details. But still, you know, it was very high level. Yeah, you couldn't drill down and understand. Yeah. I think also like forget about just the high level. Again, scaling happened. So now we're talking about where we were getting maybe 70 NPS responses in a week. We're now getting 70 in a few hours. So how do you tag everything? It's not automatic. We don't want it to be automatic because we want to actually feel the touch it and understand it. So Yvonne was doing it. How can he doesn't have time to do that? So got handed off to customer success and we had one person doing it. Now one person can't manage it. So two people. But now do we have to like how many people do we want to put on this? And let it alone. This is only to get high level information. So we're not even getting to we're not getting much meat out of this. We're getting kind of the packaged meat. But we didn't open it. We didn't look at it. We didn't weigh it. We don't know what's going on. So the first problem we had was that it was too high level information. What else? So I feel the second problem that we had was related to the interface of the NPS. So what we did is after three months from when you sign up, we would ask you to rate our product using NPS. And then we kind of ask you for feedback and rated it. But what it kind of created is the fact that the only feedback that we got was from people that signed up three months ago and are using the platform. And we asked ourselves, you know, what about the feedback from people doing a trial period? Right? Like there's so many people trying out the product, not succeeding with it. We didn't get any feedback from that. And what about people using the product for over a year? You know, it might have a different set of problems. So we felt like the input that we got from NPS was very one dimensional and didn't represent the whole set of customers that we had in our platform. So not only that we felt it was high level, but only we felt it didn't represent our customer base.

It also represented a group of people that were in our favor. Because if you made it to the three-month mark, you were the smaller percentage that didn't churn either in trial or in your first month of using it. And that's where the biggest drop in our users are. So now we're getting people who, they might have complaints, but they weren't enough to cancel their plan. We analyzed people that were paying for the platform after three months. What about, you know… Those that never made it. Yeah, the ones that didn't succeed with the platform. The one that was frustrated from something during the trial period. This is the feedback that we want to get. So we felt like it's very biased towards people that are going to succeed with the platform. So too high level, too specific of an audience, and a very easy kind of feedback for us to digest, but not necessarily realistic. Right. I think in general, when information is easy to get, it's not the information you want. You want to get, you always want what you can't have. And that's exactly what Agan is talking about. Like, we want to hear from the trial users who didn't pay. And originally we were thinking, we'll never know because they didn't pay. But then there was that aha moment where we're like, we actually do know what they're saying because they write us. Our largest number of tickets come from trial users. We were not just, so we were just not tagging them. That's the point. We were doing nothing. It's not even about tagging. We weren't even thinking about it. It was like not, it wasn't a thought in our heads. We were thinking, answer. So NPS was good. NPS was good. But it just wasn't enough. It, let's say it was one piece of the puzzle. What's the next piece? The next piece was let's use our platform that we're talking to the users. So we went into Zendesk. That's what we use. And Zendesk has the ability to actually go and create tags on tickets. You can do it in a bunch of ways. You can do different levels of tagging. You can have only one level of tagging. That's actually where we started with one level. So we ended up kind of in the same boat as NPS with the two high level information. I think we had at the time exactly the same tags that we had as NPS. Except now we open the audience. So now we are hearing from trial users. We're hearing from users who've been with us for one year, two years, three years. We have more people, which is great. But now we have a lot more information that's actually not so useful. So we have a ton of… Noise. Yeah, it was noise, which was still amazing. It's more than what we had before. But we hit the next wall. So we ended up with, again, so many high level, so much high level data. And we didn't know what to do with it. So we know there's a billing problem. Okay, what's the billing problem? I can now say a little bit more because I spoke to the person. So we came with a little bit more abstract information, but you can't measure abstract information. I can't say, I can't go to Ivana and say, change the pricing because someone said they don't understand it. First of all, you need to understand that it wasn't an easy change for us. Because I don't know if you remember from previous episodes where we discussed that we answer tickets under 10 minutes. To ask everybody on the team to tag each one of those tickets, you know, at first people, like, it was hard for them to accept this. I'll be the one to raise my hand to be the first one who said, no way am I doing this. I thought to myself, you want me to answer every ticket in less than 10 minutes? And some tickets are paragraphs. Someone wants me to, like, help them fix their managing their team, and I have to do it in less than 10 minutes? It's our main KPI for your team, right? It is the, yeah. And you want me to do this in 10 minutes, and now you want me to think about what they wanted, and I have to go through the dropdown and look for it, and that alone is a minute, two minutes. Changes the whole way your team works. You're changing, yeah, it's a whole new mindset. So that was problem number one, but I think one of the things that we explained to the team and the team realized that in the long run, it's going to reduce the amount of tickets. So I think it gave a lot of motivation. Why? Can you explain that? Basically, the point that we explained is that once we understand the problems, we can fix them, and then you won't get those tickets in the future. But I feel we introduced a new problem, because, you know, if you go back to NPS, so it was very high level, because we have a very small amount of tags, so we thought, you know, let's make this more granular, add more, you know, another level of complexity in those tags, which introduced a new problem, which is every person is different. If you ask two people to tag the same feedback, you might get different results. Basically, what happened is customer success now became translators, and it's not as simple as translating from English to Hebrew. It's translating an idea, a feeling. So I would think that this person is really trying… Let's take an example.

Someone writes in, I'll go back to the billing, where someone writes in, the billing is too expensive. Now I'm talking to you and you're answering and so I say, can you explain that further, what do you mean? And the person comes back and says, well, I'm four users and you want me to pay for five. So I take it as they don't like the pricing structure. But the person next to me might say, oh, they just want a discount so they're only paying for four people. So is it too expensive? Is it confusing? Everybody interprets that differently. Was it free text for your team to complete? Yeah, we work very little with macros and it's all you communicate. So how many tags did we end up having in this process? I think once we opened up the high level and then we opened the low level, which we didn't dive into yet, but basically had roughly 200 tags, I think it was. Yeah. What do you do with 200 tags in this process? So again, I mean, the thought process was to make this more granular, which… Right, but the result was different. Yeah. Also, like a lot of people, you know, this is a simple case, what Effie mentioned. Sometimes you get a ticket on something that you don't have a tag for. Often. Often. What do you do? So a lot of people, you know, they want a tag, so they choose this generic tag. So one of the tags that somebody created, I'm not sure who, was customization. And a lot of people felt, you know, this is very generic. It falls under a lot of categories. So people started tagging customization. Every once in two weeks, we have a meeting where the CS team is presenting, you know, trends in tags. And all of a sudden they say… You have a new trend. Listen, over the past month, the trend of customization has gone up by 30%. What do you do with that? First of all, I ask, what the fuck is customization, you know? And the truth is, when he asked that, we looked at each other and said, what the fuck is customization? Like, it meant nothing to us. And we realized… What is it? That it was, it depends on who you ask. When I saw customization, to me it meant that someone didn't feel like they can customize the board to their needs. When I asked someone sitting next to me, they said, oh, customization means that they want to make the background of the screen a different color. So it was UI or UX or some design related. And some other people said it's, they want to be able to integrate with something. So it's not quite integrations because it's not exactly what they meant, but they want to customize the platform to what they want. Like they want to build it as their own. So now we realize we're speaking three different languages on one tag. And we're going to the product decision makers and saying, change customization. And it's like saying, it's like going to the supermarket and saying like, you don't have food. There's tons of food. It's just, I don't have the food that you want. So what does that mean? Another problem that was presented as part of us taking all the tickets that we got from them and making this more low level is that it was very hard to analyze trends. I remember Galit like doing a meeting where- Who's Galit? Galit's the person that now- Masters tags. Yeah, customer. We basically realized that in customer success, we need someone who's in charge of setting the tone for voice of customer. What is this tag mean? What does it do? And who's analyzing this information? So we need someone who, alongside answering tickets, is only focused on voice of customer. And that's Galit. And she also runs the meetings with product. Yeah. So just imagine, so she had given this presentation and she want to show trends on tags. So first of all, to show absolute numbers is very hard. Because I mean, she might show a graph where the feedback on billing has gone up, but I would argue and say also the amount of signups has gone up or the amount of paying users. So is it rising higher than the amount of signups that we're getting? Is it in line with the- Yeah, with our growth. I mean, is it an indication that it's more of a problem or is it less of a problem because we scaled faster than that? So we could be improving on that actually, you're saying. So absolute numbers don't work. So then we tried percentages. Also very hard. Percentages out of what? We tried to do percentages out of tickets. Very bad. Because what ends up happening is that, you know, if billing was 10% and it's gone down to 8%, is it because we introduced another problem which created more tags? So billing became a lesser proportion of all the tickets? Like we might introduce a new problem in the product and it might look- It's such a painful one that now people, yeah. And it might look like billing improved. So it was very hard for us to analyze, are we getting better with tags? And again, the reason to analyze is just to have like action items and just to know what's the more- Yeah, exactly what we discussed about NPS. It was very high level, very hard to move the needle. We thought that once we get more granular, like we can track like more specific tags and see specific low levels. So that we can improve product in the other side. I mean-

At the end of the day, you want this in order to improve. I think what was also really hard is we would go to this meeting and you're excited because you have all this data and numbers and graphs and percentages and all this information. And remember, customer success is usually a pretty abstract team. We speak in words. We don't speak in numbers. And now we're like, finally, we're coming with numbers. And you walk into the meeting and you show these numbers and then you end up with nothing. Not because you didn't do anything, not because you don't have information, but because you're measuring something that is almost impossible to measure. And I remember every meeting, Galit would leave there feeling defeated. In fairness, I also felt defeated because I thought we got it and then we realized we didn't get it. And you have to go back to square one and say, okay, we're going to lose data because we're going to shift our tags. We're going to cut tags. We're going to create tags, whatever it is. And every time you change something, you lose what you had before because I can't go. I don't have time to go back to 30,000 tickets and retag them because that doesn't make sense. So everything always pushed another month. So now we have information, but it takes about a month to actually measure something and mean something. And we worked the month before and that's garbage. Now we're going to do it again. And at my next meeting, I'm going to go in and say, listen, I don't really have concrete information, but this is the direction and looks like this. And again, we're abstract and frustrated. But you do actually have gold. It's just how do you dig and get to it and make it clean and worth something? Yeah. I mean, I do want to say that we also got a lot of good stuff from those taggings. I mean, we talked about the problems, but the fact that we got such a volume of tags allowed us to break down and do more granular analysis, which you can do. Not about specific tags, but for instance, we could separate people on the pro plan from the standard plan from the basic plan, which was big enough groups that we can see. And we actually saw that it was pretty interesting to see the problems that they have in the pro tier is very different from the problems they have in the basic tier and then the standard tier. So I got a lot of better insights, I would say. Went even deeper than that. It was admins versus regular users. It was trial versus paying. So the volume was very beneficial, but not enough to track specific tags, I would say. It even allowed you to understand what questions you'd like to ask yourself, but it didn't give you the answers, right? Yeah. I think I'm going to also go on the positive here. It started for customer success. It started giving us, it reminded us the importance we have in the company, which was also enough to say we can impact the company. And though each meeting we felt a little defeated because we're coming with not enough data or maybe sometimes wrong data or wrong analysis. But you do actually have gold. It's just how do you dig and get to it and make it clean and worth something? Yeah. I mean, I do want to say that we also got a lot of good stuff from those taggings. I mean, we talked about the problems, but the fact that we got such a volume of tags allowed us to break down and do more granular analysis, which you can do. Not about specific tags, but for instance, we could separate people on the pro plan from the standard plan from the basic plan, which was big enough groups that we can see. And we actually saw that it was pretty interesting to see the problems that they have in the pro tier is very different from the problems they have in the basic tier and then the standard tier. So I got a lot of better insights, I would say. Went even deeper than that. It was admins versus regular users. It was trial versus paying. So the volume was very beneficial, but not enough to track specific tags, I would say. It even allowed you to understand what questions you'd like to ask yourself, but it didn't give you the answers, right? Yeah. I think I'm going to also go on the positive here. It started for customer success. It started giving us… It reminded us the importance we have in the company, which was also enough to say we can impact the company. And though each meeting we felt a little defeated because we're coming with not enough data or maybe sometimes wrong data or wrong analysis of data, we still would leave that meeting and be like, shit, we really have an impact. We can push things forward. We're gaining confidence. Although I have to say that it doesn't sound like you had the wrong analysis. Everyone together, you were in the process of finding out what's the best way to use data. It's not like you were doing a bad job. It's everyone together on the first time using this data. But when you go into the meeting, you're super high on like, look at all this stuff. I have 20 graphs and I broke down this and I broke down that. A lot of work. You put in like, Galit would work a month straight on just this one presentation. And then you go in and you forget that it's still amazing information, but you didn't hit the target. And you feel like, oh man, I didn't hit the target. I mean, listen, the CS has done an amazing job and Galit has been amazing in creating this presentation and analysis. But what they felt in customer success and what they felt in R&D is that at the end of the day, it didn't make an impact. So we would go out of that meeting and everybody felt the same way that what should we do now? We got this information. There's no action item we can take in order to solve this. How do we translate it into actionable insights? So take us to the next step because we didn't stop there. So we brainstormed with product managers on what information do you find impactful? So two things came from this. One, we realized we have to change the way we work with this data and how we can analyze it and build with it. And the other is how can we impact and move things forward. So I'll go first in how we changed our process. And what we did is we sat with the actual product managers and said, okay, you're the ones thinking of building a feature. What do you need from us?

make that a successful feature. And we sat and we gave what we're capable of doing, which is basically tagging. It's analyzing the data and we can do high level and low level. And we realized that we have to go feature oriented. So now someone's working on a feature, let's say, for example, the timeline. Timeline is basically like a calendar view where you can see all your board information on one timeline where you can see what happens when and when a task is starting and when a task is ending. And people wanted more. They had a lot of negative… It was a new feature we presented. It was actually an existing feature that now we were going to work on improving it. And when we first released this feature, we thought, wow, this is killing every problem. And it actually opened up the door for a lot more problems. And so what we did is we went into our Zendesk and we created a high level tag, which was just timeline. So now we know there's an issue with timeline. But then we went on the low level and we said, let's break down what timeline can be. So for example, we had people asking for yearly view. But then we also had people asking for hourly view. And currently it was shown in daily or weekly. Sorry, weekly or monthly. So people wanted it either zoomed in or zoomed out. And so we created those tags. And then we were able after… Actually, after like a week or two, we had enough information to say, listen, more people are asking for a yearly view. So now the product manager can say in their head, let's focus on yearly view maybe. I guess that's what went on their head. I don't sit with the product manager. But actually we can see impact there. Because now if they create the yearly view, which now they have, what we should see is the amount of tickets coming in negative yearly view should almost disappear. So now we're translating customers and speaking product. Directly into product. And we're not thinking customer success. We're thinking customer product. That's it. That's where our minds are. So you could be on the same conversation with a customer still tagging for other needs but not when you're thinking product, you tag it product. Exactly. And also what came from it is now we have a KPI. Our KPI is to kill tags. That's what we want to do. What do you mean? We want to say if today there's a tag for timeline yearly view, next month we don't want to have that anymore. So we're now killing tickets and killing tags. So one, we're helping the whole customer success. We're helping the whole product. But we also have a new KPI that we can measure the success of voice of customer. So we're now working just for the sake of working. We're working and we're measuring our own work. So we're measuring tickets. We're measuring success of product. And we're measuring our work as voice of customer. So tags now become a more dynamic thing rather than just something that sits there. And so I can say to Galit, I want you to kill as many tags as possible. So it's her mission to go into this product meeting and give actionable items. To say we don't have an issue now with this broad topic of billing. We have an issue with price per user. So now product can go and work on pricing. And we know that in however much time it takes to develop this product, like for instance pricing, it doesn't happen overnight. You have to start measuring it and understand what effects. Do you change everybody? There's so much behind it. But we know that it will take a certain amount of time and we can see that tag disappear. And so Galit is now measured on tags and not on just this idea of coming with a graph. The second thing that we changed, which like all good things happen because of a crisis we had, was actually a couple months ago. So we made a big scale up in January. We doubled our marketing spend. And as a result, the amount of tickets that we got, customer success, was almost doubled. And I remember… Overnight? It was overnight, but yeah, like over the course of a week. It felt overnight. Think about there was Christmas week where it's quiet. And in customer success, we think of it as like vacation because you're talking about a fraction of your normal amount of work. And then boom, like January hits and double them out. Yeah, so I remember one specific meeting where a few days later, so Tom, which is the head of customer success, went through Zoom because he's based in New York and basically said the CS team is like crushing down. Like we can't handle the volume. We managed to keep it under 10 minutes, but now like this month is going to be about 20 minutes. Just so you understand, it's never happened to us as a company. We're always under 10 minutes. Ever since we got to 10 minutes. Ever since we set this KPI, we've always been under 10 minutes. This is the first time we're over not 10 minutes, 20 minutes. I was on maternity leave. Like I was basically a month and a half into my maternity leave.

The point of maternity where you're now starting to maybe enjoy not being at work and not resenting your life for not being there. And I remember Tom coming to me and saying, like, I don't know what to do. Like, this is insane. We've never had this. And I opened our dashboard because you can open it from anywhere. And I saw 20-something minutes and I was like, oh, my God, I failed my other child. And then I looked at how many users we have and what's going on. And there was a problem. We can't scale our team as fast as we scale our marketing because it's just not. And we don't want to. Yeah. So, I mean, one part of it was we didn't scale fast enough with the amount of people, but also, you know, it's not scalable. So I remember we sat down also with people from R&D. And basically we said, look, it's a company effort we need to do now. It's not just a customer success problem. And let's think about, you know, how to reduce the amount of tickets. Because one way to reduce the amount of time is not to get tickets, right? So I remember Tom saying, you know, just out of passion, look, just last month we got 800 tickets about people asking which tier they're on during the trial period. And we replied to them they were on pro. It doesn't make sense to me that we don't fix this. And we said, look, it's 10 minutes to fix this. Like, we get it. In terms of fixing it in the product, you mean? Yeah, just, you know, mention somewhere. Just providing this information somewhere else. It's a sentence that you have to put someplace on the product. Yeah, you know, again, it's one of those small problems that as you scale it becomes a big problem. What's helpful about that was it was very specific. Like, we felt this is… This is something you can do something about. Yeah, we could leave the room and have a specific action item of adding information about which tier you're on inside the product. It will take about 15 minutes to do so. And as a result, it would reduce 800 tickets for customer success. How many problems like that did we have? So this is what we asked Tom. And then Galit went on and basically sat with the team. And they mapped about 20 problems, 20-25 problems. I think what the mind shift here was that the customer success actually acted as product managers in a sense. Because they analyzed the problem and they suggested a solution. And sometimes we took the solution, sometimes not. But, you know, the essence was they thought about, you know, why is this a problem? So we took 20-25 problems that were easy to fix in terms of like just adding something to the product. Expecting to see many, many tags and many tickets just disappearing, right? Yeah, but I think one of the best things about this list of problems is that next to each one of those problems, we counted the amount of tickets that we got last month. So when a developer sees, you know, this problem generated 2,000 tickets, this problem generated 5,000 tickets, you say, you know, I can fix this and reduce 5,000 people asking a question. And if you think about this more, how many people actually open a ticket? I would say, you know, 10-15%. What about the rest of the people that got frustrated because of something and didn't open a ticket? They're kind of speaking for a larger crowd. Because the average person doesn't open a ticket because they're frustrated. The average person just gets frustrated and shuts it down. So we have to take this information and say we're holding the key to what probably many, many more people are thinking. And so like what Rahan said, when you see there was 5,000 tickets, probably it's not 5,000 people, it's probably 20,000 people. I would even say 50,000 people. Probably, I just wanted to scare everybody. One out of ten. It's another invented. I'm with it. So basically just to summarize that, so the CSTM a day later came because of this crisis with a list of 20 things to do. And we said, listen, there's gold in here, like stuff that's going to, the product is going to reduce the amount of tickets. And we scheduled a hackathon a few days later where the entire R&D team basically went through each one of those items. Together with the CS team. Together with the CS team and fixed all of them. And what's cool about it is because it was very specific, we could track that like a week later or two weeks later. The tickets on those specific topics were gone. So let's just touch on that example. So people didn't know which tier they're on. We added an indication in the pricing dialogue. And then we saw that two weeks later nobody asked about which tier I'm on. So, you know, great victory. So we said let's do this on an ongoing basis. I think also what comes from it is from a manager's perspective is it's empowering your team. Like there was an energy that comes from the team when they see like I can put all my complaints on a piece of paper. It's not on paper. It's on Monday. But I can see something. And now the whole company cares. So that means that next time I see a problem, I'm not speaking to empty space. Someone's listening. And it makes the team actually want to tag the tickets. Like it goes back to that problem that we had of adding extra time on working on a ticket.

But now it means something. It means that when I tag something, it's going to go to a board that now product is looking at or R&D is looking at, and they're going to change something. So I'm actually physically having an impact on our product. And it makes you want to do it. It makes you less annoyed that you're taking the time to tag a ticket. It makes you less annoyed that you have to send that extra message to somebody and maybe create another ticket for the moment, like, at that moment, but you're actually killing so many more tickets, and you're making your work easier. More meaningful. Yeah, easier and meaningful. And it's an amazing feeling as a manager to see that, like, your team now is really impacting a company. So now we're good, but looking into the very near future, we're scaling all the time. What are the challenges that are just about to come? So I think there's a couple challenges. One is it's something that has to constantly evolve. So that feeling of being defeated when it wasn't good enough has to kind of go to the garbage and understand that when you scale, the product is evolving, the needs of users evolve, the users evolve, and this voice of customer has to evolve with it. So it has to be known that, like, on a weekly basis, we're working on this and not that in three months from now we're going to have another hackathon because we have tickets coming. It can't be that hackathons are a problem. Yeah, a crisis, there's always going to be a crisis. Let's leave that for something else. So I think those are our biggest challenges. If there's anyone listening to this episode now, what would you give them as a piece of advice? Start this right away. As long as you have one customer, you have a voice, but you have to give to the product and work on it. And don't wait until you have 30,000 tickets a month. Do it when you have one, and there's a process built in, and make the customer success team or whatever team it is that speaks to your customers, make their voice powerful from day one because as long as you're selling something, you have to sell to the people, and you have to measure it right. When people think about scaling the customer success team, I think that there's three ways that you can do it. One is to recruit more people to customer success. The second one is to improve the process of how you handle tickets, automations, bots, interfaces in Zendesk. And the third one is reducing the amount of tickets. I think from what I saw is that a lot of people, and also we were very focused on efficiency and scaling the amount of people. But the best ROI you can get is by not getting those tickets. I would say our goal is to get the amount of tickets to zero. That's the best customer success we can give, meaning that people won't open tickets. Not because they don't want to, just because we solve every problem within the product. You can open a ticket and say, wow, you guys are amazing. I'll take it. Yeah. So it's not only helping the customer success team to scale, but also improving the company and the product as a whole. So going forward, we put a lot of our focus on reducing the amount of tickets. And I feel this is a great way to think about scaling the customer success team and scaling the company in general, just to solve as many problems as possible inside the product. Because the last thing you want is that, let's say, you got a building problem, is to create a very efficient bot in Zendesk to answer those tickets in no time and feel this is not an issue. You want to address those problems. The root problem. I think I'll add to that also with scaling your team. We made a decision very early on that we don't want a support team of 80, 90, 100 people just answering tickets. We want to bring in people who are more than that. I'll take a small team who can think big than take a huge team that thinks small. And we want to keep it that way. So if we want to keep it that way, we have to grow in other ways, and that's killing tickets. And that's not a bot. We do have EffieBot. I'm not against EffieBot. I'll keep her. But otherwise, yes. But it's for a very specific… It's very, very specific. Thank you, Effie. Thank you. For joining us today. Thank you, Lil. Thank you, Elan. Thank you for listening. Start up, start up, start up. Start up, start up, start up.

Oh-oh-oh.