Episode Transcript
[00:00:00] Speaker A: What kind of formats for, you know, educating hub work the best, like workshops or the office hours or maybe some guidelines. What, what's your experience on the formats?
[00:00:11] Speaker B: We had a lot of different formats and for example we had regular office hours, we had lectures, we had workshops. But with the workshop, for example, I didn't guide them myself. So here I don't have a lot of experience. But speaking about lectures, it was different formats again it was like pure vendor sessions. It was sessions with for example, somebody speak about the, the vendor speak. And plus it could be a case from the inside from the company and it could be the third format when we started making the dashboard to understand if you need the session or not.
And this dashboard thing we just started doing the last year and it was kind of interesting experience because the person can understand directly if he need to come. And most of the times they coming because they see that for example exactly this optimization we need right now, we see this main metrics, we see this main KPIs which is influenced by this metric, by this optimization and they are coming. And for me this format of the lectures and internal case was one of the most of the, one of the favorites. The most of the people joined it because they saw that somebody already did it and maybe they can apply it too because they see that this case is alive. This is not something like artificial, this is something real. And we explaining them like how they can do it, how another team did it, how they find out it. And then one of the vendors explained more into details like what else you can do, how you can understand how to do it and what is suitable for you. And most of the times people joining this. But speaking about for example finops office hours and just office hours in general, I did it kind of weekly, but it was not the best for me. It wasn't work as the best format because teams still continue texting me and continue asking me and sometimes of course I send them to this office hours like come join me on Monday and I will explain you everything.
But at the end it's better to leave something like a channel for them to communicate with you than just saying like join exactly this day, only this time I am available.
I think it's not the type which I think is working very well. But yeah, lectures was one of the best things.
[00:03:00] Speaker A: That's cool. I totally agree that like case studies and you know, seeing people already doing that is one of the best resources in FinOps. I tend to like focus on that, on the newsletter itself and I think it's More realistic when I see you know the different companies like Adidas, Alando, all these big companies like doing like a major thing, it's like yeah, they did it so maybe you can do it. And they posted like GitHub code even or things like that. So examples and then office hours. Yeah, I think you know it's better to have like a Slack channel or a team channel and you know, drop questions over there and then from those questions like you can maybe leverage and do a guideline or maybe do something that can be shared to our an FAQ section maybe.
So I think those are very good, good formats that you mentioned and you know, starting from you know the formats. So have you been able to like on board or embed sorry this education into like processes of the company. So for example, when you are doing the onboard or when you are doing you know, certain updates, how have you been able to do things like that?
[00:04:13] Speaker B: We decided that we don't want to put and to push our finops education everywhere and we just focus on the few key processes where it can be suitable. And I tried to keep it like, like inside the central team and of course we had some processes where we did these meetings like every time for example budgeting. Every time before budgeting started I always did a session where I explained what to do, how to do who have to participate. Like all this information. I teach them how they can forecast themselves. And it was something what take apart with the process.
Sometimes it was but mostly we try to do our own schedule and for example we try to do this education session once in two months, one in three months like gathering this data, this optimization ideas about which we can tell to the teams.
But we had this phenops we had this meeting for the practitioners and which was something like an onboarding and it was my main onboarding process. And when sometimes it's a lot of changes in the company when you need to onboard a lot of people but not just the people who is just who was just hired but people who take a part in on this phenops practitioners activity. For example somebody left and they find a new person who will replace it in the finops activity. And when it's a lot of people I did this onboarding. Yeah but of course all the finOps process were supported with the documentation and yeah, so long story short, we try to keep this education activity based on our finOps processes and that's the main thing here because sometimes you can try to put it somewhere and to drawing any activity from other teams but we still Was afraid that it could be a blocker. It could be a blocker for something. Because when you put a lot of activities into one process, it's becoming longer. You can try, you can do your best, you can do everything, but it always will be longer. So that's why finops practitioners and education became mandatory. And I think it has to be mandatory if you want to grow the maturity of the finops in your company.
But it's still the thing which FINOP central team decide when you have to do this education meeting and which pro before which protest you have to educate your people more.
I think here it's like this.
[00:07:01] Speaker A: Okay. No that's, that's an interesting approach. I've seen like, like mandatory. Everyone does it to like nobody does anything. So I think that's a hybrid approach. It. It's a good stuff. And you know what have you seen like as a motivation for people to to learn finup so you know, you have the, the mandatory things but you know, for people that you know want to learn more about it, what what are having you know the incentives or, or tips that you have seen on, on your side.
[00:07:33] Speaker B: It's so different because people is so different.
Everybody's motivated with so different things. Forgot for example, somebody is crazy self motivated. Somebody is motiv and when the manager is pushing he will start doing something.
Somebody is crazy about the process and they like the process so much and they like to, to learn to educate and to do all the stuff and but the thing is that people like to see the result and people like to see how they influence something.
Like one person came to me and say once that like I want to do finops because I want to see the result of something because I want to see how my which all the power which I spent influence the money influence something influence the product.
And I think this is the thing which can motivate most of the people. But to be fair things goes much faster when all the skippy eyes becoming mandatory and after that people splitting for two types.
First people who is doing a bare minimum and they will continue doing bare minimum if you will show them the results. They don't care. But and another part of the people who, who, who you can, whom you can motivate and who want to do more.
And both of them again you have to show them, you have to visualize their results.
They have to show them the KPIs which can motivate them. For example how they're decreasing the waste, how they decreasing the cost and how much they save. This is what can motivate them. And the very good sign which I saw once, but not once I think more some people started setting their year goals with the finops goals.
And so this is like this year goals which you set up for HR with your manager. They started putting finops there and of course it motivates them too. So people see people can be motivated when they see that it's something valuable for them, that their actions make a value for them. And for the, for the, for the company, it's much less to be honest, because it's always about us.
And so yeah, from my point of view, something like this, yeah, that's, that's interesting.
[00:10:15] Speaker A: I think what you mentioned about the goals, like I haven't thought about it because I do like have a similar goal system on, on my company. And, and I think that would be interesting like especially. Well, of course for the Phenoms people, but also for, for engineering like, you know, reducing cost or making cost more efficient for existing deployments, I think it would be a very interesting goal.
And also for development like, you know, I'm able to do, try to do the same architectures in a more efficient approach. So that would be super, super interesting. And you know, okay, so we've talked about like how you explain to spinoffs and how you make people learn it and how you can educate. But so let's say that someone like from the different Personas, let's say, okay, I get it now, I pass the education that you have gave me and they want to start applying fino. So what do you think are the first steps that they need to do?
And start with.
[00:11:16] Speaker B: I want to think about this question from the two points of view. What will do my engineer engine when I will talk to him and when he will go through the education process and through this FinOps champions process or whatever and what the manager will do. I think from this part because I think that we everybody know what the phenops practitioner like professional phenops practitioner will do because it's much more difficult. I don't think that we will not have time today to say all the steps which, which the person have to do when they want to start the phen. It's a lot. But if we speak about when you educate somebody and when you want to see that they achieve the result.
With my engineers, first he has to go and check the tagging like if everything is tagged. And the second thing he has to check his waist dashboard. This is like in my experience what my engineers have to do. But again of course it's like your way how you will set up your finops processes your communication with engineers and whatever.
But I think here is the few things which you can make more self sufficient for them for the engineers which engineers can do. And this is everything what they can touch with their arms area. And most of the time this is simple way simple optimization. This is a structured optimization which is suiting for everything. This is right sizing. This is waste. This is for example checking out which is 3 type you can use etc gravitons for RDS and so on. This is simple activity with the budgeting.
And yeah this is like an examples which. Which. Which can be done. And for my engineer he will go and he will check the tagging and waste dashboard like how much waste? Because he already will understand that waste is super bad and simple applicable thing which you can do first and that the tagging that is like our strategy. This is our long term activity and he is responsible for that because not we are doing not finops team doing Terraform scripts. And this is something what the team is doing and that's their area.
And speaking about the manager, if we talk to the manager or to the owner of the branch or whatever. From my point of view but this is only my point of view that he have to go and set up his alerting system like you have to set up how much he have to spend daily based on the average or based on the maximum for example and his monthly monthly letting system. I think this is like first steps for people whom you educate and who talk to you.
As a phenops central team.
[00:14:29] Speaker A: I think that you. I was curious about the different Persona approach and how do you do it with engineering. What one is one thing and I think I started with tagging and the waste. Yeah, I think waste and tagging are one of the concept that I have seen more people involved into and more people wanting to learn. And the budgeting is also of course and it's more like a managerial kind of thingy and it's fairly easy to start with. And you know we have started with. We have thought about the things to start and the first steps. What have you seen are the first mistakes that people do like when they are starting and they do the best approach but still make some mistakes when they are starting doing finops on their own.
[00:15:10] Speaker B: I think that people want to achieve a quick wins and it's not always the thing like again if we'll go back to check your tagging. Tagging is not a quick win at all. But it is a thing which you have to take care about at first and when you're starting doing this quick wins, quick optimizations without understanding what your business want to do. And if you really need to do this exact optimization right now, this is the mistake from my point of view which you can make and you have to be aware of it because quick wins is not always win win. This is just quick and not quick is not all the quick things is good, but sometimes, of course, sometimes it can be exactly in the eye, but sometimes it can be directly wrong thing.
[00:15:59] Speaker A: That's a good quote.
I think that sometimes we think about the quick wins and yeah, this is fairly easy and this is what I need to do. But sometimes it probably generates an issue in the long term or it's something that you need to make a quick win every week and it's not sustainable and maybe you need to autumn automated or things like that.
So yeah, that's interesting. I can tell like one of my managers when I started doing Finops and all of this, one of the advices was like, if your hourly rate is more than the cost that you are saving, you probably better use your hourly rate to develop the business and not save money. What I mean by that is if you are doing experiments and enjoying a small project and the time that you are thinking about doing a quick win optimization that is not, let's say replicable or automated, then if your hourly rate is more, then you are wasting company's money instead of doing development.
So you need to think about that, right? So it's something also to deliver to engineers because we like I come from the engineering side and it's like we think a lot about testing and doing the latest thing and the latest trendy thing and we forget about like the cost to the company like in the hourly rate, right? So it's like you need to think about what you are doing and you need to think about, you know, is this worth for the company?
[00:17:29] Speaker B: Yeah, because it's always about the collaboration and I think that anything ops, Most of the FinOps activity have to be done with the collaboration. And so if you are this like when you're doing this first steps, you cannot do these first steps without support. Even if you are, even if you are first phenoms practitioner in the company, if you just was hired, you still need to start with the communication, you need to talk to the stakeholders, you need to understand what is your, you have to have this list, what is your perfect first steps. And quick wins have to be there, of course but you have to communicate with people because not all these optimizations applicable directly everywhere.
[00:18:12] Speaker A: And yeah, no, I think you say right and now that you're talking about communication, like what had been the main problem or problems that you have seen in the communication from the FinOps team when you are trying to build awareness and communicate about the practice?
[00:18:34] Speaker B: You know, I don't know a lot of mistakes because people like to speak about success, they don't like to speak about mistakes. And when you are search for something or when you are listening to the podcast or when you are watching this, I don't know, synopsis, sessions or presentations, people saying you like feel free to speak about the problems. I remember once they told me like if you're doing a presentation speak about the problems, it's okay, but at the end you don't want to, you don't want to speak about that. And that's why the I don't know a lot of mistake from other people. But I know the mistake which I did once at the beginning of all this finops communication stuff and trying to onboard people on the knobs board and it was not to focus on the people who is not interested from your point of view, who is like not suiting or something.
And I remember I started going to the managers and saying them like give me one practitioner, I will educate him. Just choose this one person, not more, I need only one.
And people started providing me two person, three person and I'm like no, I'm not going to onboard them, I need only one and this. And at the end it shows that these people they want to learn because the manager choose them because they want to educate themselves and they want to participate in something and they crazy interested. But at the end, at the beginning I said that I don't need them but at the end this is like extra brain in your FinOps group. This is extra person who can even teach you something, who even can find something interesting for your company for your FinOps activity. And at the end I onboard them all. It was just him that like it was one main phenops practitioner and two like sub phenops practitioners and the first person was responsible and others was just goes went to the education sessions, they participate, they ask the questions like they were there, they were there all the time. And I think yeah, it's a mistake when you think that you know exactly which people you want and you not involve others, but you always have to be open.
[00:21:06] Speaker A: That's very interesting. I think you don't have enough resources anytime in Finops especially in the human side, as more people that you can get involved, the better. Even though it may be trickier for you that you can feel like a bit of a botany. Like I have to teach too many people at the same time and all of that. I think that. But, you know, having as much as you can and having as many people involved as you can.
[00:21:31] Speaker B: You know why it was like this? Because it was a major process. It wasn't what it was not like only just join us and that's it. It was the my responsibility how much people I onboard, how they meet their KPIs.
Because I need this list of the responsible person. I just needed to be sure that they're gonna.
Because it was my annual KPIs, it was my activity. I have to be sure about these people.
And at that time I didn't even think that maybe I have to onboard somebody else. I think it was about the mandatory. About the mandatory of the process. Yeah, I said wrong at the beginning. That's because it was a mandatory, because it was necessary. These people have real responsibilities. Not just like something in the air. No, for real. You have to track your KPIs, you have to track your budget, you have to track your waste. You have to do all of that. And I'm tracking how you track it and I'm tracking how you're good in it and how you perform. And that's why I decided that it had to be strict. But at the end we still found this responsible persons. But I was able. I created another group which I called like education group. And they just joined us and took part in the processes in education.
[00:22:57] Speaker A: That's definitely like an interesting approach. And I think what you mentioned is very interesting from the finop side because I remember reading a blog post from I think Ellie Rao, she was in Disney and now I think she's working in Finaut. And she mentioned about the burnout of FinOps. Right. So because sometimes like FinOps are involved in too many things. They have too many tasks and it's, you know, since it's a bit. It's a process that is involving in too many factors.
I don't know if you have suffered like not finops or not, but you know, you have feel like a bottleneck and how you've been able to solve, like if that happened, like have you been able to solve to, you know, try and delegate or try to leave this task to more people?
[00:23:44] Speaker B: We delegated.
Yeah, it was delegated. And what I said about this mandatory thing about this Mandatory phenops practitioners. The more teams you have, the more difficult for you to be customized central team because you need to support everyone.
And like at the end it was 120 teams or something and Finop's team was overwhelmed. It was a lot to do and the self sufficiency helped helps you a lot. And this is how to not to be a bottleneck again at the topic of the discussion. Educate them, provide them tool set, provide them documentation, support them and just let them do their stuff stuff because they in. In some things they so close to their numbers, they're so close to their resources.
Sometimes when we speak about budgeting I don't remember if I talk about it or not but how it works right now and how it I think have to work like you are as a FINOP central team. You are doing main canvas like your main forecast and the team can put their details, they add their details exactly in their forecast and that's how you can achieve the best detailed and the most tried forecast. After providing opportunity for our teams to budget themselves. Of course of course I control it. Of course I do analysis of the numbers. Of course we had our KPI for example is it how much is bigger than our forecast? How much is lower than our forecast was a few KPIs at the end we get like if the pre the year before it was only 75% of the teams was inside our forecast, inside the budget at the end of the year next year when teams become more self sufficient. When team started budgeting themselves it was already 95.
So it plus 20 years of 20% of the budgeting accuracy. And I think it's a great. It's it just. And we achieve. We started spending three times less time as a central finops team for this budgeting process and plus we get it 20 better.
I think it's real numbers and this is like we were just bottleneck and we make the process better because we just did the clear processes, documentation tools, et cetera.
[00:26:40] Speaker A: Yeah, I think what you mentioned is key. Right. So being able to create like a self sufficient thing, self serve thing like being able to define the process so people can follow it, try to optimize it as much as you can. I think in general good businesses are built upon good processes and once you do a good process that impacts a lot of the company. What you mentioned three times less and an accuracy of 95% is key.
You probably cannot have an AI doing even better than that. So that are great damage especially for budgeting and forecasting which are probably the most difficult things that you can do in finops, you know, knowing how much are you going to spend. Right.
[00:27:32] Speaker B: Yeah, it is, it's. Yeah, it is. But yeah, I'm really proud of that what we did. And yeah of course we did very good forecasting. It was my colleagues who, my colleague who did it. Very crazy good. But anyway, I think that it's forecasting plus teams, plus process plus tools. It's not, it's.
It's in the connection with.
[00:27:55] Speaker A: Oh, that's, that's great. Yeah, I think it's, it's a wonderful result what, what you said and you know, looking at the. A bit about more the. The human part, I don't know how you have seen like the, the evolution of the culture in the, in the company around FinOps and how you have seen like from when you started in the, in the practice like maybe in your early experiences until now that you may have like seen like a lot of situations and a lot of changes. How you have seen the culture evolving like in the different companies and how you have seen it growing. Maybe in your case.
[00:28:29] Speaker B: I can say how I see how the culture growing. In my case I think that this is about how people doing and when and why what is their reasoning and it's always changing and then more major is becoming then high is ownership.
That for example, I can count it on the amount on teams with the finops practitioners. I can understand it on how many again how many teams was able to do their budget without my support. I can measure it on the for example tools usage, how many people start using the tools, checking the dashboards, how often who is joining the meetings. Like this is the level of the interest in the company and when the interest is growing then you can understand that the culture is growing. Of course it could be sometimes a toxic culture or whatever.
Yeah like people just saying finops without understanding what is it.
But I think when you start showing them KPIs or where to pay attention on it's becoming more better. For example, this year I saw how much more directors joined the process and how they joined it that their vocabulary become more professional.
They started understanding what exactly to do. They started providing the ideas how to make it better. For example, when I did my products researches inside the company they already came with ideas because they started understand the framework and this is for me how the culture change without the saving. Because everybody understands how to measure it with a saving. But this is more about the human communication. Because when people, when people join, when you see that people speak about finops already more professional. When you see that people, without you pushing them to check the dashboard, they. Checking the dashboard and when they come up with ideas, I think for me it's a good sign of growing.
[00:30:57] Speaker A: Oh, definitely. And I think you wrote like a point that I haven't thought about or seen. Is that to measure it right? So, because the culture sometimes is more, let's say more human, more abstract, more things like, yeah, it's growing or it's not growing or we are in the situation, we have a better or worse culture and it's like what is a better or worse culture? But what you mentioned about mentioning it and seeing like how many people join, how many people are talking in the events, how many people are, you know, proposing ideas, checking the dashboards, all of that, like having KPIs for the finops culture, I think it's something that I have never thought and it's a wonderful, a wonderful idea. We'll probably need to think about defining the best practice about like what are the KPIs that are useful over there? And you already like shared a lot of them.
[00:31:45] Speaker B: So yeah, to be honest, everything what I said right now, part of it can be counted and it can be, yeah, it can be a quantity, quantity. And for example, you can count how much people check your dashboard. And it was one of my KPIs, which I check all the time and when I make the changes. And for example, when I did the education sessions, when I speak about something, what you can find on the dashboard, I checked after if people started searching it, if I explain it right, if I show them where to check and if they go and check. So yeah, a lot of the things you, you can, you can count, you can set up this logging and monitoring and it will work for you. You can just do it in a good technical way.
[00:32:28] Speaker A: So yeah, yeah, definitely. It's just another KPI and you can set it up the same that you do, you know, with the cost per user or things like that. You can do it for, you know, how many clicks, how many visits a dashboard has instead of a website or the checkout cart or whatsoever. You can do it for the dashboard.
So, yeah, I think that's a wonderful thing to end up the conversation. I think it was very insightful and a lot of tips and tricks about how to build a FinOps practice and culture in your company from scratch. So thanks a lot for being here today with with me and looking forward to having you next time.
[00:33:08] Speaker B: Yeah, thank you for inviting me. It was great to talk.
[00:33:12] Speaker A: Yeah. Pleasure to talk and see you in the next episode. Bye.
[00:33:15] Speaker B: Bye. Bye bye.
