Customer Support Strategies, Metrics, and AI | Jovana Kandic

59

---

Joana Kandic, Product Manager at cake.com, discusses customer support strategies, metrics, and future trends.

Jovana is a seasoned professional with a strong focus on customer support and success. She excels in managing all facets of customer service, from developing strategies to overseeing performance metrics. Currently, she is leading a team of product managers, driving innovation, and leveraging data to achieve business growth. In her previous role, Jovana directed customer support efforts, establishing policies, metrics, and quality assurance programs. Jovana also recruited, trained, and led teams while implementing systems to gather and analyze user feedback.

Takeaways

• Align customer support with business goals and focus on exceptional support
• Measure metrics such as first reply time, resolution time, and agent performance
• Utilize AI to analyze feedback and assist support teams
• Collaborate with other departments and prioritize communication
• Take care of support agents’ well-being and monitor team retention

Connect with Jovana

• Website: www.cake.com
• LinkedIn: https://www.linkedin.com/in/jovana-kandic/ 

Connect with Vit

• Website: https://vitlyoshin.com/ 
• LinkedIn: https://www.linkedin.com/in/vit-lyoshin/ 
• X: https://twitter.com/vitlyoshin 
• Other: https://linktr.ee/vitlyoshin 

Timestamps

(00:00) Intro

(04:55) Customer Service Strategy

(16:20) Measuring Support Team and Customer Satisfaction

(27:45) Collecting Feedback and Analysis

(32:27) Trends and AI Tools for Support Team

(43:13) Collaboration and Communication Across Departments

(47:28) Advice from Jovana

Transcript (Edited by Vit Lyoshin for better readability)

Vit Lyoshin (00:02.076)

Hello everybody, welcome to the Vit Lyoshin Podcast.

Today’s guest on the show is Jovana Kandic. She works as a product manager at cake.com.

Welcome Jovana.

 

Jovana Kandic (00:16.994)

Thank you.

 

Vit Lyoshin (00:19.21)

All right, so just a couple of words about you. You started as a customer support manager, and then you grew to director of customer support, and now you’re product manager lead. So it’s quite a journey, and I bet now you know a lot about your customers as a result of that journey. I invited you to talk about customer support in general, metrics and analytics around that, and future trends and how AI can help in this area.

Before we jump to the topics, if you would like to mention a few words about what Cake.com is, how your journey went, and how you end up in the role that you are in.

 

Jovana Kandic (01:09.281)

Yeah, sure. First of all, thank you for having me on your podcast.

cake.com is actually a software company. We have three products, Clockify, Pumble, and Plaky. Those are basically productivity tools. And I’ve been with the company for, I think about five, more than five years. I actually started off as a customer support rep, basically from the trenches of customer support.

I’ve listened to your previous guest, Elena. She said like she had 17 different jobs across her career. I feel like my career trajectory is completely opposite of that. I’ve been with one company for the whole time. So yeah, that was funny for me. But it is quite straightforward being with just one company. But I’ve evolved into a managerial role quite quickly. Those beginnings were a bit formative for me. I’ve learned to listen to user feedback quite quickly and how important it is for product success. When I took more of a managerial role, I focused more on growing the support team, analyzing metrics, improving the processes we had in the support team and strategies.

So everything in order to just improve the quality of our support. This experience was really important for me to shape the way I look at the product and our own customers. So I had the opportunity to work really closely to communicate with them regularly, to really get to know their needs and their challenges with the product and in their workflows, eventually.

So transitioning into product management was quite natural for me. It just felt like the right decision to make. Yeah, I currently lead a team of product managers across all three products, actually, at Cake. And I try to leverage that experience with customers to drive our product development.

 

Vit Lyoshin (03:25.25)

Great. Yeah, that looks like a natural journey. And actually, I started my journey as a customer support rep as well. And then it went in different directions at a certain time. But I think, this is probably the perfect path to become a product manager because now you know exactly what people are using and how they’re using it. And you’ve been there for a long time. So you also grew alone as a company grows.

Okay, so let’s start. And first question I would like to ask you is about the overall customer support strategy, how you start building that strategy, and how you align it with business goals.

 

Jovana Kandic (04:14.561)

It kind of rolls off of what you previously said. I grew with the company, so a lot of things that were relevant back then at the beginning are quite different now that we have grown as a company. At cake.com our focus was just on trying to solve customer problems in any way possible. And I’ll be honest, at that point, there was no discussion about the strategy or anything like that. We were just focusing on serving our customers in the best way possible. As the team grew a little bit, we definitely needed a more structured approach. I remember the company being really straightforward with me and in their intention to have world-class support. And I knew from the beginning that, okay, we need to have exceptional support.

What does that actually mean for us? How to define what’s exceptional support. So I started off with that. If you say what’s important for the customers, you have to analyze what they feel is the best customer support. So the first thing I did was basically analyze the data. At that point, we mostly had tickets from a support team. In the beginning, that’s the early stage of developing our first product, Clockify, we had a few thousand tickets, so it’s not a big amount of data, but still pretty relevant. And I went manually through all that. I basically analyzed what were the most common compliments on our support team and what were the most disappointing cases.

And what I learned was that the one thing that struck out was how fast our support team responses. The most common compliment was always, oh my god, you really answer quickly. I didn’t expect to have my issue or question resolved or answered in such a short period of time. On the other hand, we had disappointing experiences where agents couldn’t really figure out what’s the customer’s issue. So if troubleshooting went in a different direction, if there was a misunderstanding at the beginning of resolving the case and early in the communications. So these insights really helped in forming our goals for the support team. We already knew based on those insights what would be the most important goals for us to resolve in the future period. So basically that meant that we needed to be more responsive, and more available to the customers because those were the things that we wanted to double down. So if they already feel this is great, let’s double down on that because we knew we could cover more channels or even have a whole day like 24/7 support time. We knew that we needed to enhance the knowledge of support agents so we could minimize those cases where basically support agents go with troubleshooting in a different direction or don’t have the tools that they need to actually resolve or diagnose those issues that customers have. This kind of rolled off to improve our consistency.

As the team grew, we had 10 agents that are all different kinds of people. They talk in a different voice, they solve issues or understand issues in a really different way. So we need to make it a little bit more consistent. So when a customer contacts us once or twice or multiple times, they get the same level of support from us. That was really important to make. 

And after that, basically, we talked about, okay, now we got the whole coverage and we are responsive. Maybe we should contact them before they contact us. So proactive support and that’s where our success team idea, to develop a success team was brought.

So at first, we decided it would be best to prioritize those first goals like responsiveness and availability. We first had only email support, that was the first step. But quickly we introduced both chat support and phone support. Lots of support teams don’t really like receiving calls. Especially when you’re troubleshooting software, you have different kinds of people contacting you, it’s really hard sometimes to diagnose issues when you don’t even see the screen of the customer and you don’t get any time to actually process the information they get you. They’re just on the phone, they want their solution right away and it creates a little bit of stress. But for us, it meant that we were available on another channel and that this was the way that people wanted to contact us. Fine, we can do it, we want to be there wherever our customers are, and if it’s the phone, then it’s fine.

We also went parallel with that. We worked on making our team a little bit more knowledgeable, like upskilling them. This consisted of two ways. One way was actually developing some internal tools to help them diagnose and troubleshoot certain issues faster. This was nothing special, just a quick troubleshooting page so you can send the data quickly about your device or stuff like that so that the agents can do that faster. The other part of that was actually training the agents about the product, about the processes, about soft skills, and basically focusing on that.

So we introduced a whole new level of training, both for onboarding new support agents, because we grew quite quickly, from maybe 5 agents to 35 in a period of a few months, you know? So that was a really challenging thing to do.

We made a whole new training program, and new materials, and the whole onboarding process last for a couple of weeks. So it’s not something that you just come and start working with customers right away. You just need to learn the product first. This was all tied to that goal of being consistent with customers, trying to find that voice. We were never too formal with our customers. Never too strict about it. We always wanted to have this friendly tone in our support. So there had to be some alignment between agents regarding that. But this is the part of soft skills that they had training on. And I think that worked really well, investing that time at the beginning of their work, like a month or three weeks, depending on the time.

Agents themselves, to set them up, to onboard them quickly, and then they can actually contribute to the whole team really well. 

And lastly, of course, we formed that success team because we figured, okay, there are clients that request different kinds of attention. There are large companies that need to have personalized onboarding and different kinds of setups. This all required way too much time for regular agents to do. So we figured this is the need we need to cover with a different team, maybe to even proactively reach out to those customers. Some people will never call for support, they’ll just struggle on their own and give up if they don’t see it works. So we figured it’s best for another team to cover that. And this is what we really did.

An important part of all that was to actually revise the metrics we had, because we now set up certain goals for our teams. In support, you can track everything. There are a lot of things that you can track, but not everything relates to your actual business goals. So we had to revise the metrics we actually track in order to be more aligned with the business goals we need.

 

Vit Lyoshin (13:45.354)

There is a lot that you covered. And I really like streamlining and making the customer support sound and feel the same, regardless of how many times you contact, because that experience with the brand is very important. And if it’s different every single time, it’s kind of annoying and you don’t know what you’re going to get. One day you get a sleepy person, another day you get a happy person or whatever, right?

 

Jovana Kandic (14:17.143)

We had a couple of challenges there because our support team covered all three apps at the point. So you have this content switching. One customer calls you about one product, and the other customer calls you about a different product. It could be a bit challenging. And that part actually helped the agents to keep the same voice across products. That was also important for us to cover but yeah one day you get a happy person the next day you get someone who’s like I don’t want to do this today.

 

Vit Lyoshin (14:51.834)

Yeah, exactly. You covered a lot of goals and strategy, how you figured it out, and exceptional experience that kind of reminds me of Jeff Bezos with Amazon, that he’s preaching all the time and saying that the company has to provide exceptional customer support. And if you do that, your business will grow and it will be, you know, since people have an amazing experience, they will come back all the time.

So what are some of the measures specifically for the customer support team that you installed versus how you measure customer satisfaction from the other end?

 

Jovana Kandic (15:35.881)

I would say that measuring the success of your team strongly relates to customer satisfaction because if your team is successful, but your customers are not, something is wrong. Something does not work quite well there. But as I said earlier, we tried to revise those performance metrics to align with goals more. So this is what we focused on tracking.

We had the first goal to be more available and more responsive. So in line with that, we had to monitor our coverage. So the team had different shifts and schedules. And we had to arrange this to allow the agents to be available across all channels at basically all times. This was really challenging because you never have like abundance of people. You don’t have an unlimited number of agents. You always have some issues with staffing, holidays, vacations, and stuff like that. So that was a difficult part also to cover. But we managed and this was important to keep that consistent so the customers always knew they can reach us at any channel, at any point. 

As for responsiveness, I think, the best way to measure this is by measuring the first reply time. We basically measured the average and median response time. That’s the time when the customer actually initiates contact, but regardless of the channel, how long it takes us to respond at first. For emails, we always had this target of one hour, this is what we are happy with. Usually, it’s actually way shorter than that. It doesn’t take one hour for us to respond, but this is with something that we started with and said, okay, like this is something that we can cover with the resources we have. And this is something that our customers respond to really well when they see that we answer so quickly. So let’s try to maintain this target.

And I think that It’s really important not to measure just average time. I said because you always have those outliers that can skew the data significantly. I think it’s also important to check how many tickets are actually responded to within the target. So if you have like 80% of your tickets responded within the target, what happens with the 20%? At what time do they get a response? And what kind of issues are those? Do you need to train your staff more on those tickets? Or it’s something else. Maybe it was a spike for a different reason. So this is the important metric for the responsiveness part. That was our goal.

Of course, you need to track resolution time, to see how long it takes for your team to actually resolve a customer issue or answer it. This is a tricky one. I know that it’s not always up to the support team to resolve the issue. Sometimes it’s basically just an issue in the platform, in the product, and it goes to another team for development to be resolved. You may keep the ticket open or on hold until it’s resolved so you can follow up with the customer. That also can make an issue on your analytics dashboard. It could skew the data and you might not get the right result. Whether you close it, whether it stays open, it’s a bit difficult. But you can still use those data to see, okay, these types of issues took the longest or these types of issues were most common and stuff like that. So it’s still really valuable to track that metric as well.

We also track per agent, the stuff that the number of tickets saw. You can see which agents perform. I wouldn’t say better, it’s just more efficient, maybe than other agents. And you can see the score of their quality. This was something that we did.

It was a process of a few months that we said, okay, we have this all quantitative data like we have the time, resolution time. What about the actual quality? Like, how do we measure this? And it’s not an easy thing to do because every ticket, every case, every request from the customer is a bit different. So it’s not easy to measure the quality. But we actually tested this for a month or two and we finally came up with the criteria that worked for us, and we made some kind of internal scorecard. This internal scorecard is basically a set of criteria that you have a quality assurance manager go through some random tickets per agent and just look through and see if that tone of voice or the structure of communication is followed through, if the comprehension was at the level that we require, if the troubleshooting was appropriate to the type of case, and if they followed through the process after resolving the case, like categorizing the ticket properly, following up with the customer, and stuff like that. Those were just a set of criteria that we used, and weighted in the scorecard. And, basically used it to measure the quality of our tickets. Of course, the most important one is whether the problem is actually resolved or just answered and nobody knows what happens with the customer afterwards.

But these were the most important metrics we followed to measure the success of our team. For customer satisfaction, like you mentioned how that corresponds to customer satisfaction.

Of course, we use the customer satisfaction survey and the score for that. It’s quite basic, just a way for customers to actually provide feedback on each case we resolve. And it’s our way to get those comments about what’s missing out, what’s negative, what’s positive in our approach. But it’s not the only one that you can use.

We at cake.com actually use only customer satisfaction score, but you can use a lot of different metrics to measure satisfaction. I think a better option if you can implement it in your system could be that you measure customer effort score. This is a really neat metric for support teams where you can actually see, okay, you’ve resolved this issue for the customer, but did it take a lot of effort for them? They had to provide you or spoon-feed you with the information, or you could have maybe found that out on your own somehow. So this is a great metric. I think the implementation can be a bit tricky if you are sending already a customer satisfaction survey. So oftentimes you’ve got to choose between the one because you don’t want to overwhelm the customer with that. You don’t want to send too many surveys. You would probably want to choose between one. So there’s a balance you need to do, but if you are able to mix it and able to figure out which times to send one and which times to send the other, I think that can work really well for the support teams and to see if the customers are actually satisfied because they may be respond to your product overall, not to the actual ticket case. They may say, okay, this was a great interaction, but because I love your product, not because I love the resolution of the solution. So satisfaction score can be a bit inadequate in that sense. Effort score can compensate in that way, but if you’re able to mix it, that probably works the best.

And there are a couple of other metrics that are really great for your team, as well as for customer satisfaction. You can measure average handling time. It’s pretty good for the support team and their performance, but also for the performance of your product. If you can see that certain cases take too long for agents to resolve, then that might be an indicator that there’s an issue with the product area itself. It might not be, but it’s worth looking into. And it’s a great way for agents to actually collaborate together. So if you are looking at average handling time, this is basically how you set it up. Like you can set, you already probably measure or track types of cases. It could be onboarding, it could be billing, it could be some product area. You have an email coming in or a call or a chat and you can tag it with certain teams. And once you’ve done that, you can see how long it takes for an agent or the whole team to resolve, on average, those types of cases. And if you see that onboarding is taking way longer, you may look into your onboarding. Maybe it’s not as intuitive as you think. You can see the differences between agents. If one agent is performing better than the other in one area, then they can share tips and mentor each other with that. I think that’s a really good thing that you can add to your team and it could turn out great.

 

Vit Lyoshin (26:26.314)

That’s great. And I assume you also have some sort of visualization for all these metrics and you review them periodically, right? And you mentioned also managers doing some of these reviews and picking and choosing certain cases and things like that. Yeah, that’s great. There’s a lot to measure with specifics to these issues that customers send you.

Collecting feedback from people just when they feel like maybe just overall experience they want to share or they want a suggestion to send your way. What are some of the best practices for collecting this feedback and processing it? Does it go through the customer support team or it go directly to the product team? How do you do that?

 

Jovana Kandic (27:16.893)

Yeah, I think I went through various stages of collecting feedback. So for startups it could be in parts, it’s a bit easier. In parts, it’s a bit difficult. When I say this, I mean, when you’re in your startup, you get less feedback. You can monitor all that manually. If you’re a product manager, if you’re a support team member, depending, some startups don’t even have a product manager per se. It’s all a team effort. And this part could be a bit easier. As you grow, it’s more difficult to monitor all that. You’ve got many, many sources.

For us, at first, it was just a feedback form. It was just a single form that customers themselves could basically fill out and let us know if they wanted a new feature or improvement, or any kind of feedback. The support team also filled out the same form instead of a customer if they realized that a certain issue was actually an improvement request even though the customer didn’t explicitly say so. So this was a great way to capture those sources but as you grew you got many reviews on software review platforms you get feedback from exit surveys, like if you have an exit survey when somebody deletes the account or cancels the subscription, you get internal ideas. If you have a big team, then everyone has their own idea. You got sales, you got support, you got marketing. Everybody wants to chip in with their own interests in that. So it’s hard to manage all that. It’s hard to keep a pulse on every feedback. So you would probably look into automating stuff like that.

For us, the challenging part was to centralize all that feedback. It was difficult to organize it in a way that you could see everything in one place. And we did this by leveraging APIs and webhooks in order to basically get everything into one central repository. And then product managers can swift through. But in the early days, basically, product managers were a part of our Zendesk support platform. They were a part of it, they could see the emails directly and they could browse through them manually. So that was in the early stages. They were definitely involved with actually seeing the client-customer communication as well.

Once we got everything in one place, then it was a problem on how to actually get valuable insight because you got all of this data, but you need to organize it, you need to categorize it in order to see the teams, the issues that are common. Because if you have in a day over 100 requests that go through, then it’s really difficult to see the broader picture.

Therefore our issue was to actually organize it properly. So it’s a bit more useful for product managers to do and I think the easiest part for us there because we had our own tool, project management tool. We kind of feed everything from the central database to our actual ticketing tool which we continue developing for there. So when product managers figure out, okay, these are the issues that we need to solve, these are the opportunities that lie behind that feedback, then we can basically just create a ticket to our own tool and just move on the development process from there. But yeah, it wasn’t easy to have everything and not miss any feedback.

 

Vit Lyoshin (31:26.198)

Yeah, okay. So let’s switch gears a little bit to the future and see if we can mention some of the trends that you see with some new tools that come out that help. You mention something about automation so maybe something that you use already or you just see in the SaaS industry in general what companies are using

 

Jovana Kandic (31:53.197)

For us, this is something that is early for us. This is something we’re testing out, but we are trying to basically make an AI model to prompt that feedback. For starters, we did this with reviews. This is the information you can find on reviews platforms like Kaptera, G2 or stuff like that, Trustpilot, wherever. And you can feed all that into a model, and you can prompt, and you can ask the model, okay, what was the most commonly mentioned, the pros and cons of our app in the previous, like couple of months. This helps, and in the AI, I can summarize this for you. This helps product managers quickly get an overview without actually having to read all of the reviews. Of course, you can go straight to the source. If you find something interesting, you want to see the actual review because models are not perfect and things can slip. But this is something that we are testing right out. And I think with time, it will be even better. The model itself will be trained a bit better and it’s gonna work for us really well. I think this is like the new hot topic for everyone. Like if AI is going to replace the support teams or we had like that’s a bit of deja vu. We had that with chatbots, the initial chatbots. Everybody thought that this is going to be the new first line of support. I think that it’s a bit different with AI. I think it’s more useful for companies.

I noticed the other day, that Intercom actually changed their whole strategy to an AI-first platform. And they actually, I think introduced their AI co-pilot for agents and the AI chatbot for customers. I think the usefulness of AI will actually be in helping support teams be more efficient and productive. I think this is a really neat thing. I think this is something that people should be excited about and see how it works. I see a lot of companies going in that direction. So I mean, I tend to believe that this part of chatbots is not ideal because people will always like to talk to other people, but still, AI can be leveraged to help support teams to find tickets, find the history of customer communication easier to summarize stuff like that, to write notes quicker, to pull out knowledge-based articles or write knowledge-based articles much faster. This all can contribute really well to the productivity of the support team. So I think this is going to be the way companies will go further.

 

Vit Lyoshin (35:05.666)

Yeah, I see that some people are scared a little bit that we’re not going to have customer support people anymore. We’re not going to have developers. We’re not going to have QA engineers anymore and things like that. But I think what you just said is what I also believe is that those AI plugins and models will help become more efficient. Yes, it may be to reduce the workforce in general, so instead of having 100 support agents you may need only 50 but the quality will grow dramatically

So it’s not like everybody’s gonna be gone and the job disappears completely. You just will have the best talent there and with the help of these tools now to solve customer problems more quickly and more efficiently and things like that, I think if you train these models on existing tickets and how they were solved. That’s really the benefit of this. And people can really say, okay, here we go. I know how to solve this problem. Other than just relying on them to remember the steps and things like that.

Because I remember my days of this. And you know, something you solved a couple of months ago, you may not already remember how you solved it. And you have to remember everything and go through the same trouble.

 

Jovana Kandic (36:31.785)

Yeah, of course, I totally agree with that.

 

Vit Lyoshin (36:33.85)

And then maybe eventually AI can also learn from this and provide some simple answers to people how to do this with instructions. Maybe even like you mentioned, write articles simply about how to reset your password or how to, you know, which technologies we are compatible with, which ones we don’t, and things like that.

I think there are some opportunities.

 

Jovana Kandic (37:04.501)

And what you said about people being scared about AI, like taking over jobs for customer support people. I think this happens with every innovation, every time. This is probably the most disruptive thing that happened in the customer service industry for quite some time. So I understand the relevance of it, but if you look through history, every innovation made that change for jobs and people.

But it didn’t make it worse, it made their work a bit easier and that’s something to be happy about.

 

Vit Lyoshin (37:40.678)

Yeah, exactly. Before people used to carry everything on their backs. And then they adopted horses and donkeys. And now we have cars and trains and ships. So it becomes more efficient. You can do more with fewer people. And here we are. We still have like 8 billion people on this planet. And everybody, well not everybody, but most of them have jobs and stuff.

So, I think in the end, everybody’s winning from this breakthrough. There are other concerns about AI, which I’m not gonna get to right now, but I’m sure people know those.

So, yeah, so when picking and choosing these tools, which ones to deploy to your organization, are there any methods or any priorities that you set for your team specifically, which ones to install what will benefit my team versus what will benefit the customer? If you can speak to that a little bit.

 

Jovana Kandic (38:46.489)

Well, I think it’s more about the questions that you need to ask yourself as a company. You mentioned issues with AI and that already brought to my attention about security, privacy or certain data. You won’t be able to use some tools because of that. Like, okay, if you are training a model in a closed environment, that’s fine. But some tools will not be available for you if you have a strict policy for certain data that you wanna have in there. So that’s one thing that you’ll definitely want to consider when implementing any kind of tool. And what you said about seeing what helps your customers, what helps you, like as I said earlier when we added another channel to our support team, that was more like, this helps our customer. It might not help our agents really well, but it turned out great for both parties. Our customers were happy to have another channel. Our agents learned a new skill that is quite good for them.

I think when you consider implementing any kind of tool, you need to think about if you’re going to use it internally or externally. Like if this tool is going to be helpful for your agents only and they can only see it. This is the thing about AI. You can use it to help your agents. They can use it to summarize their answers, to make their tone, fix it, or more friendly or more elaborate, depending on what you need. Or you can use AI externally. You can use it as a frontline. So this is what you want to consider. What’s your end goal here? Do you want to try to leverage AI for internal or external purposes? Then you need to think about integrations. Do you need to integrate data with something? Do you need to integrate those tools with the tools you already use? How will they integrate? Is it costly? Is it easy to implement? Is it something that you need to maintain throughout your work

 And this is something that I think it’s really important for teams to plan out in advance before they jump into implementing any new technology. And lastly, is it a good investment for you? Because as I mentioned, like Intercom, I think they price their AI assistant, they price it for $1 per resolution, per resolved case. For some, this might be a really good deal. For some, it may be really, really difficult to calculate in the future, to see if this is a good investment. They might feel that this is really expensive for them, especially if it’s a startup and they feel like this could ramp up quite quickly.

It all depends. So bottom line is this new technology actually a good investment for your company or not?

 

Vit Lyoshin (42:08.598)

Yeah, that makes sense. 

So let’s talk about collaboration between different departments. When customer support receives their request and they are stuck, they can’t fix it. It’s a true bug or whatever, or maybe sales receive some request or feedback in the email. How do you streamline this? How do you figure out the collaboration between different departments and how do you synchronize all that and process?

 

Jovana Kandic (42:40.129)

Yeah. For us, it’s actually, again, we have this upside because one of our products, Pumble, is a team communication tool. And it’s really neat that we have that kind of channel that we can just forward the communication to a certain department or certain team.

We also have Plaky, which is a project management tool, where we actually create every bug or report. So if QA or support reports any bug, it goes to our internal tools where we have a lot of control. We know how to set everything up. We can streamline our workflow really well. And this proved to be really efficient for us.

I think that the problem arises when If you get a lot of support tickets and they are not organized properly, and you can lose touch of priorities like you can lose touch of what should be solved first.

Clockify our tool that was the oldest tool we have, we have millions of users on this tool. So it’s different to see how we solve problems in Clockify versus how we solve problems in Pumble or Plaky because they’re in different stages of maturity. So we experienced this problem with Clockify where you get a lot of requests in a short period of time and then you have to struggle on how to prioritize this. In the beginning, it’s easier. You just always try to push forward everything. We are building a new feature. Let’s try to fix this issue with that new feature. But eventually, it doesn’t scale well.

So you need to have a different system. We have a board in Plaky where we note down all of the bugs. Then our product managers prioritize those bugs and send them to a different board where developers actually continue to work on them. So that’s the process we take. But I think it’s important to note that if teams end up working in silos like everybody is working on their own thing, then this creates a lot of miscommunication, a lot of repeated work that is simply not efficient as much. Definitely, you need to pay attention to that part if you have multiple developer teams and more people working on a single project.

 

Vit Lyoshin (45:20.502)

Yeah, so you’re basically solving these problems by building your own tools. The short answer is, which is great. You know, you have the use cases that you solve, build the tool, and then sell it to other people. So maybe you will come up with a fourth product at some point with AI or something, I don’t know.

 

Jovana Kandic (45:43.557)

No, don’t give them ideas. I mean, it’s pretty neat. Even Clockify went from our own needs. Clockify is a time tracker. The company used to work on projects and wanted to actually track time on how many projects, and how much time is spent on those projects. And they just figure out, okay, we don’t have to pay for this other tool. We can build a time tracker ourselves for our needs. And it just kind of went from there, which is also cool that what you mentioned, we actually build products that actually solve our own issues. And we hope that this solves issues for a lot of other people as well.

 

Vit Lyoshin (46:27.886)

Right. Okay, so we’re coming to the end here, and the last question I would like to ask you, or not really a question, but ask you what would be your advice for support teams, for product managers on like maybe a few highlights or few most important things about customer support in general, about the teams, anything like that you can provide your advice on.

 

Jovana Kandic (46:57.393)

I’ll try to structure this in certain segments for support agents, for reps. My biggest advice is to take care of yourself. Being a support agent can be really emotionally demanding. People tend to forget this. They are caught up with their everyday work and they easily burn out. I just want to remind them that it’s okay to take a break after a while and just pause for a bit.

For product managers, it’s similar, but for them, I would advise that they listen to every other team to include them early in the process to make them not just feel heard, to actually hear them out. This is all to enhance the collaboration between them.

If you grow quickly, then it’s easy for teams to drift apart and everybody’s doing their own thing, but you’re actually all building one product or even multiple products. It doesn’t matter. You all have a common goal. So product managers are the glue for that. So they need to pay attention to that. For support managers, I would advise keeping a pulse on their retention metric for their support members for their team members to see if you’re this relates back to your question about metrics for the successful support team if your retention is high for the team members in the support team I think that then you’re doing a good job and your team is successful because they want to stay they want to work even in a high highly demanding job such as being a customer support rep

 

Vit Lyoshin (48:45.322)

Thank you very much. Those are great pieces of advice And especially in this time right now when the job market is not the best, it’s important to take care of people and pay attention to their stress levels and not overwork and things like that.

Okay, awesome. Thank you very much for your time today. And it’s been a great conversation. I hope we can talk more in the future.

 

Jovana Kandic (49:18.841)

Thank you, it was a pleasure talking to you today.

 

Vit Lyoshin (49:22.466)

Thanks, bye.

 

Jovana Kandic (49:24.045)

Bye.