Organizational Network Analysis
44:15 Share

You're listening to Data Skeptic, Graphs and Networks, a podcast exploring how the graph data structure has an impact in science, industry, and elsewhere. So welcome to another installment of Data Skeptic, Graphs and Networks. Today we're getting more into core network science, I believe, or at least how to visualize them. In this audio podcast, we won't be doing a whole lot of visualization, but we'll talk about a tool called OrgXO for visualizing and what it's good for.

Asaf, you put me in touch with Gabriel and he actually has some nice things to say about you in this interview and some of the help you did. How did the two of you meet? Actually, Gabriel turned to me because of my network science podcast, Netflix.

He was very proud of his software and he wanted to share. And he did, and I saw it, and it was, I have to say, pretty awesome. So it's neat. It's like a network analysis tool. I guess his users will, for a company of a certain size, I can't remember if he says 50 or more, but something like that,

Not just to see that typical tree of how we all report up to the CEO, but also to see the cross-functional people and the true connections of an organization. What Gabriel does is organizational network analysis. He looks at the organization as a network, and that allows him to analyze the, I'll call it in quotes, the real organization, right? How the organization really works.

What's interesting, except for all the ONA thing, I think what's interesting is Gabriel touched on one of, I think, the biggest problems in applying network science in the industry. And that's trying to convince people to check their networks or check their networks hidden in their data. They are ubiquitous, but also very transparent. You hardly think of them on a day-to-day basis, unless, of course, you're a network analyst.

Another part that was interesting to me is how he gets his data. I thought maybe, oh, you're going to import this from some system, but it's through a survey process. The employees self-report who they interact with the most, which ought to be a pretty good source of ground truth, I think.

You can do it actively, like surveying the employees, but also you can do it passively by linking your system to an API and could be emails, chat, messengers or something like that and analyze it. But it's the organization's choice.

So I've got a personal anecdote that's kind of interesting. Do you know what Active Directory is? Okay. It's a Microsoft product that is like a Rolodex for a corporation, basically. And it's surprisingly rich in metadata to the point where usually IT organizations manage it. And it has an API. And I was at a job quite some time ago with a little extra time on my hands.

And I wrote a crawler that used Active Directory and pulled down a copy of my whole company's organization, which was about 2,000 people.

And then I would diff it every week. And I got advance notice before a big company meeting of layoffs because certain people just disappeared from the tree all of a sudden. I should have figured out how to inside train with that information, but I never did. It's okay. After you get arrested for going over GDPR restrictions, I guess you'll have much time in jail to think about what you could do with all this data. Right.

I don't know if I broke the law. It was quite some time ago. I think I got statute of limitations on me. Definitely pre-GDPR, pre the California one too. Probably. But who knows? Anyway, Active Directory, interesting tool. But a better data source, I think, is the survey process because then you get, you know, you could also maybe look at email. I was thinking who emails each other, but that cuts out like water cooler discussions and meetings and stuff like that. So self-reporting is a pretty good measurement, I believe.

One last point I wanted to ask you about. Are you familiar with, I believe, Gabriel calls it the Brontosaurus distribution? The Brontosaurus, yeah. It's a metaphor. Speaking of metaphors, I have another one, but it's a metaphor I think Anderson used in his book, The Long Tail, in I think it was 2004, when he talked about the advantages of selling products that are not popular, that are niche.

The idea behind it was if you try to sell popular items, you'll have lots of competition by big companies and so on, and you wouldn't have a chance. But if you're trying to sell something online, now you can because of the drop shipment and so on. You couldn't have done it before. Now there's internet. You can sell niche products and make an advantage from the long tail of products that

less people are buying, but there are lots of products like this. So that was his idea. And I think he painted it as like a large Brachiosaurus.

Yeah. You've got the head, the body and the tail that all have different shapes that do kind of elegantly describe an organization to me. While there are people who interact with lots of people who would, I guess, be in the head, you have, you know, 50 employees you interact with. That's probably a valuable team member. But you also need sometimes that one engineer who is grumpy and just solves problems in his office alone. And, you know, there's a balance of all these types of interactions in a company. Yeah.

Yeah, usually our engineers think so. Anything else on Gabriel? Yeah, actually, speaking of metaphors, he was using... Oh, yeah, the other one. Yeah, he was using metaphors like fragility and black swans and so on. And then I remembered it's from the book Black Swan by Nassim Taleb. He uses the same metaphors.

If you don't know the book, most of the book by Taleb is, I think, dedicated to Taleb's own genius. But once you pass it, you find that he talks about the problem of seeing the world from a normal distribution perspective. What happens is you miss out the black swans, that is the extreme cases that you can see in a long-term distribution, but not in a normal distribution.

like orders of magnitude from your regular standard deviation. The example from the book is if cities were normally distributed, then New York shouldn't have existed, right? Because it's too large for a normal standard deviation. So I encourage people to go. I will have a link in the show notes. Look at some of the nice images that come out of OrgXO that we're going to talk about in the interview.

I think, you know, picture's worth a thousand words sometimes, but we have a good discussion about the project. Well, let's jump right into the interview then. I'm Gabriel Petrescu and I'm the founder and managing partner of Evos Innovation, which is a company based in Romania, Bucharest. We created a platform which is called OrgXO, orgxo.com.

that is used more or less like an MRI machine for organizations, and it is designed to serve non-data scientists to understand their organization as a network of people rather than a formal hierarchy or formal processes.

Well, an MRI is a detailed observational tool. It's also one you only tend to get when you have a problem. Do people find your tool when they have an issue or are they more looking forward? Our clients are not necessarily those that wait until it is too late to do the MRI.

Probably a preventive approach in using MRI might be better. But sometimes we have clients that are using them when things are tough for them because nothing else worked or worked.

They may think or they intuit that network science or the approach of looking at the organization as a network rather than the usual Newtonian type of organization that is hierarchical. And so they have this intuition that this may provide them some insights that will help the organization improve. There are other clients, for example, that are using OrgXO.

like an MRI, before, for example, before they acquire an organization in an M&A. And when they acquire an organization, this is probably the moment that is the best to understand how the organization functions, especially because you as the acquirer, you're not necessarily acquainted with all the realities or all the history within the organization. And this may be new to you.

to you as an investor. So what types of insights do users typically get from the tool? First, they discover who are the real influencers beyond those that they think.

They are, because not always those that you put on certain positions are the only influencers or the real influencers within an organization. So finding who are the key people is important. Another thing is find where are the bottlenecks and see where there are some constraints within the organization, which sometimes is not due to persons themselves or the way they

Their skills or the way they behave is sometimes due to organizational design, and this is interesting to discover. Also, they discover the vulnerabilities in the form of how organization or the collaboration with the organization is distributed among the people.

And it is kind of, you know, those aha moments when they discover that, for example, that the organization stays or relies too much on very few people, like, I don't know, less than 5% of the people do more than 20% of the work. And, you know, 85% of the organization is somewhere not really engaged. And this is sometimes shocking and frustrating.

The shock is not the fact that they discover that, it's that this is the effect of the policies they have implemented over the years. Also, they can discover some clusters of collaboration that sometimes are surprising, that are cross-departmental or cross-sector.

groups and this is very useful in figuring out whether the processes that they probably paid very high in terms of money, the processes really are reflected in the way that people interact.

in reality. What are the sources of data that you're able to use to do this sort of analysis? What does a client bring you? We can collect data actively or passively, meaning that we have a questionnaire. I mean, this is the most, I would say, reliable because it's the most actionable way of collecting the data.

based on a questionnaire that is sent to all the employees within the organization. So we do not do sampling because if you're looking for organization of 150 people, you simply cannot do sampling. I mean, maybe if it's thousands and thousands and 10,000 people, sampling may make sense.

But in smaller organizations, like, I mean, the sweet spot for us is somewhere between 100 and 500 people, organizations. We do even more, 3,000, 5,000. But when you're looking at a large organization, you're not getting all the, let's say, the actionable insights out of it. We collect through questionnaire.

It's a very reliable questionnaire we developed over the years and improved it and simplified it because to complicate questionnaires is very easy. Almost all the questionnaires on the market are too complicated and people hate them. So this was one of our focus, how to make it simple.

But we also collect data from systems through APIs. But this data collected passively from integration with the systems of the client, first of all, requires a lot to tailor data

For each client, you need to do a certain, let's say, tailoring process in order to get the right information. Because even if they are using the same systems, even, for example, if they're using MS365, the way they use it within that ecosystem is always different. So you need to adapt to it.

The passive data comes on top of the active data we collect through questionnaires and it's helpful because it provides a kind of a double check of, for example, frequencies of interaction, whether this is what they stated within the questionnaire. So if you want to describe an organization as a network, you have to develop nodes and edges. What are the nodes and edges in your graph?

First of all, we do not call them edges. I know that in data science, it's called edges. I hope that nobody is going to kill me because I'm using another language. But if I go to managers or to CEOs and tell them edges, they will probably look shocked at me. And most of them might not understand what I'm talking about. So nodes is pretty clear. A node may be a person, maybe, I don't know, an AI person.

maybe different things, but in our case it's people. But the edges or the connections, the inlinks or the outlinks, we call them links. I mean, links, we discovered that in our dialogue with CEOs and managers, we discovered that links makes a lot of sense to them. And not necessarily connections because connections has different connotations, especially in different languages. So we call them links. This is the solution we found recently

as acceptable for non-data scientists. The bubbles or the nodes represent the people, as I said, and the links represent the collaboration between those people. And this comes from the way we designed the questionnaire, because what we ask is basically, who do you collaborate with to do your job? So every person nominates all those colleagues from their organization with whom they collaborate with.

either with a high frequency or low. Of course, we collect also the frequency whether it's daily or weekly or monthly. We also collect data about how much do they depend on their colleagues. And this dependence, for example, is very difficult to collect passively from the systems. Systems do not know whether you depend on someone to do your job or not unless you do a very complicated analysis.

Another thing that we collect through the questionnaire is how do each of them evaluate their colleagues or in fact not their colleagues, the collaboration with their colleagues. So if the two of us are having an excellent collaboration or sometimes we have some problems then this is some information that we collect through the questionnaire as well. We ask the respondents

to nominate who are the top five people they collaborate with or they rely on the most in doing their job. Okay, so who are among those 20, 30, 50 people I collaborate with in my work? Who are the top five people I rely on the most? As soon as we close the questionnaire for a company, for example, we provide access to the results instantaneously.

So I imagine one piece of that presentation is the network you've built. So this, as we're looking at one on the screen that listeners can't see now, but we'll have a link to an article in the show notes where they can see some examples. You know, it's a series of people who are the nodes here, and I believe they're colored by department. And then we've got those links in between them. So a really interesting summary of an organization. But at this point, once an executive takes this in, what are their next questions?

So when they see this image you just described, some of them said, "What is this? This is like spaghetti with Bolognese." So I don't get anything out of it. Of course, this is the idea of the MRI. So you can look at it from different angles and try to dissect it in various ways or look at it from various angles. So for example, just by clicking

one button here, which is a hierarchy, you can see how the organization is looking like as a hierarchy. So as soon as I show them the hierarchy, the Class K hierarchy of their organization, they immediately recognize it and they say, "Oh yeah, now I know it."

And it's the sort of family tree version of the organization with, I presume, the CEO at the root and the entry-level people in the leaves? I mean, the CEO is on the top, not at the roots. Okay. All right. Another terminology. It depends on how you look at it. Roots are usually at the base, you know, it's underground. Yeah.

The CEO is always, he puts himself on the top. Put the CEO at the top, he or she is at the top. It's a very Newtonian approach to looking at the organization. It shows you, it's important because this is the way we designed organization for centuries. The whole, I think, nature is built hierarchically.

The first aha moment probably for them is when we just click, for example, what is called a hierarchy. Okay, so a hierarchy is the hierarchy together with all the connections. So, you know, if you look at the hierarchy of an organization, it shows you how dense it is, in fact, and how much we ignore it.

from the connections that are happening within the organization. And they start to think very attentively when I show them this, because it really shows them how much they miss just by looking at the organization as a hierarchy, as a simple hierarchy.

On the other hand, if you're looking at the organizations as a network, you can also look how siloed is the organization in their functions. For example, if you just look at the organization as how the departments are connected with one another. And if you see that the departments, I don't know, for example, from the technical departments with those from sales departments are literally separated from one another.

And this comes probably from the homophily, most probably. Homophily, for example, the technical department, they like to work with one another because they speak the same language, they speak the same, they use the same terminology and so on. And data scientists are also prone to this approach of homophily. They like to speak only those that...

understands them or speak the same language, which is normal. But in the end, this creates a lot of tensions within the organization exactly at the borders, at the crossroads between the various departments. We had, just to give you an example, which is

Not necessarily an anecdote, but the idea is that when the CEO saw the silos in his organization the first time, he was like, he said, oh, I did this stupidity.

And I said, can you explain, please? We want you to understand what he meant. And he said, because they were fighting between the technical departments and the sales departments, I decided that they should go through only two intermediaries. So this would be the bridges between the two departments. And then he said, and now things started to move even slower.

And we have even more problems because now there is a fight between these guys I nominated and each of the other teams. So I have two problems now, not only one.

In the end, they discovered that the network approach helped them to understand that the processes need to be rather fluid and adapt to the realities rather than imposed just for the sake of solving internal conflicts between people. Are there schools of thought about what a healthy organization should look like? Of course, there may be schools of thought.

It's very difficult to come up with a universal answer that would fit all organizations. Each organization is different. Two people are different. When you put them working together, of course, the result is going to be different from just the sum of the parts. So organizations are different and they change all the time. Probably the biggest challenge

Insight I came across in 2024, the end of 2024, is when we did some recurring studies on organizations. So we discovered, for example, in some of the organizations, they change a lot in nine months, one year time. They really change a lot.

But there are some other type of organizations, which we surprisingly discovered, that are very robust. I mean, they stay almost the same, even if the people changed. So you have, for example, an organization of 197 people. They were in 2023. They grew to 2020.

So there was an increase in the number of people, but they also changed some 40% of the employees within the organization. But if you compare the two organizations, they really look very much alike.

Even when you start dissecting it into, let's say, distribution analysis, all sorts of analysis, you can see that there is some kind of DNA, organizational DNA, that is very robust.

My instinct would have been to think that as an organization grows, they have to change their structure. I'm not quite sure why, but it seems like the structure of a small business wouldn't suit a large one. Do you find that to be true? Oh, definitely. Sometimes this happens. For example, we had an organization that did this. We did it for them. The baseline was done in 2021.

They were having some problems with the technical department because they thought at the time that the head of the technical department is the problem because he's not capable. He's a very good technician, but he's not very good as a manager. But then doing Orgexo, they discovered that the guy was having basically 41 direct reports underneath him.

When you put 41 direct managers to a manager... That's too many, in my opinion, to be a good manager. Yeah. So it's impossible. You're going to kill that person. So the problem was not necessarily the person in charge. It was the organizational design.

After one year, we studied them. After one year, what happens? So they still didn't learn the lesson. So they kept putting the same number of people and said one responsible for the technical. In the end, after another iteration we did after one year and a half, they finally got it. And they started to delegate and nominate some, you know, responsibles for certain areas and delegate to them.

We had a case in Romania.

One United Properties. They allow me to use their name. They have real estate development. They have the investments in real estate. Let's say in 2023, when we look at them, they were looking very, how should I say, siloed. The various departments were very evident. You know, architecture versus sales versus development or legal. In spite of the fact that they grew from 197 to 220 people,

within one year the organization was looking very much alike. So of course it's not identical, it cannot be, but still there are some things that you can, if you start dissecting them, look at the various departments, you can see they remain very much the same. The beauty stays also in the way the organization is distributed. So

Because I learned from one of my, I would say mentors. He doesn't know that, but it's Asaf Shapira. I learned a lot from his podcast and he's very, very important, at least to me. So the long tail distribution, like he was giving that story with Brachiosaurus. I don't know if you ever heard about the Brachiosaurus analogy.

Or share it for listeners who haven't. The idea is that all netters are long-tail distributed. And to understand this, long-tail is looking like, let's say, the distribution is a brachiosaurus with a very long tail and a tall head and neck and a small body, usually. So the idea is that if...

networks are distributed in such a way, it also reveals how fragile or robust is the system. Imagine that you have in the top of the distribution, you have a very few number of people, sometimes 5, 7, 10% that are doing most of the connectivity within the organization, most of the collaboration as a matter of fact.

And then you have a very long tail of people that are less engaged. More fragile networks, the long tail is really...

long and it also shows how disengaged are the people. Okay, so in the long tail, for example, at the bottom of the tail, you have some people that have been nominated by only one or two of their colleagues. And if this is the case for a lot of them, it means that you either miss a lot of potential that is coming from usually from working together or that people are simply disengaged. And why the people are disengaged?

It's usually because leaders, especially those that are doing MBAs or economics or finance, they are fascinated by efficiency. So they are put into our positions and they immediately start cutting things and making efficiencies.

Those efficiencies usually, usually not always, translate into transforming the people into cogs, into a machine. If the people feel that they are just a simple cog into a machine, they are not going to be engaged. And then you can invest how much money you want, you know, getting people engaged. You're not going to solve this issue.

If you tell them that all they need to do is interact with two colleagues or one colleague when they go to work, they wouldn't care less about the future of the company. I mean, it's becoming transactional. Work is becoming transactional. And this is probably one of the problems. The secret is always the body.

in looking at organizations. And the body is represented by people who connect well enough with their colleagues upstream and downstream. And if you do not have a large enough body, the system is becoming fragile and you basically lose, how should I say, the long-term approach. Efficiency is good for the short term.

But effectiveness is much more important because it helps you go through periods of crisis. So when I'm picturing this brontosaurus now, the only part I think we haven't talked about is the head, who I assume is, you know, the most popular social butterfly person. How important is that upper, what is it, five, ten percentile of people in the network who are very connected? When we designed the tool, we decided that the head is the number of people that make

20% of the collaborations within the organization. Okay. So based on Pareto's principle, we took 20%. How many people do the 20%? And usually they are 20%.

5, 4, 5, 6, 7% of the people. They are not necessarily the butterflies because sometimes they are those that are overburdened because they have to connect with too many people. Sometimes you have people there that are, you know, they get there because they have a certain something that allows them to be able to deal with so many interactions.

And cope. So there are some, probably some. Sure. Yeah. There is a unique talent. There is a unique talent, yeah. They end up getting there on their own, not necessarily by, you know, this is how you promote people. You see they can deal with a lot of, you know, interactions and this is how they end up being promoted. For example, I wouldn't want a CEO or I wouldn't recommend a CEO to be in the head.

And when a CEO or the head of the organization is among those in the head, that is a signal of alarm because this is, I know that he's in, let's say, potentially he's one of the bottlenecks in the organization. So usually the CEOs are somewhere in the body. The reality is that you have a lot of people in the head and the top body, right?

that do not have official functions or do not have official roles within the organization. Still they have, through the sheer number of connections, they have a tremendous influence on the organizational culture. So you better know who they are. When you connect with intentionality,

these groups, the organization improves or the collaboration within the organization improves, meaning that the distribution is less abrupt. I mean, the concavity is less abrupt. And this is a way of measuring how the organization is becoming more effective rather than more efficient. I'm not saying that I'm not trying to say that efficiency is bad. OK, so it's not good or bad. But there is a

I would say a continuum between efficiency and effectiveness that you should always be aware of and also monitor over time because if it's getting too much towards efficiency, you are going to be surprised by all sorts of black swans or surprises that are going to show within your path as an organization.

When you look at an organizational network in a vacuum, you haven't really learned about the company yet or heard the complaints of the executives. Can just the network alone tell you something about whether or not it's healthy? I'm not so sure you can tell.

tell whether it's healthy or not. Because every organization is like, it's like in a continuum, it's somewhere between very healthy and very unhealthy. You need to define what are the expectations of the leaders of the organization from the term healthy. Because for some organizations, healthy means that the leader wants to know who's doing what and that they want to use even tools to better control.

I mean, for them, healthy means control. Otherwise, I mean, it's, of course, it's a soul killer to work in such organizations. They are not going to survive very long unless they simply change people like socks, which is not necessarily something that you can afford to do. Otherwise, healthy may mean also that you have a very good climate within the organization.

which might not be healthy from a financial perspective because everybody can collaborate very well. They can, you know, have a very nice environment and work, you know, productively. But sometimes when you look at the end, the,

The financial benefits for the organizations are not always there. So it depends on how do you define healthy, what you want to create. An organization that makes money and people feel engaged, probably, yeah, this is what I would define healthy. But you cannot from vacuum to tell what means healthy or not. It's like when you go to the doctor, tell me, doctor, is it good or bad?

And of course, we are also consultants. The right answer is always, it depends. So it depends on if it's good or bad. It depends on what you want to achieve. But there are some signs that, for example, the organization is fragile. And this you can see, you can measure it. But if you want to have...

Efficiency, squeeze the lemon. You know, how much efficient can you squeeze a lemon? There are some limits. The same is with organization. How much can you squeeze an organization to make it

efficient or the people within an organization to make them collaborate in an efficient manner. Well, obviously we've been talking a fair amount about OrgXO, your software tool, but I get the sense some of your work is consultative as well. Could you describe what a typical engagement is like for you? The most difficult part is to sell OrgXO and convince, let's say, the client that they need it. We do not have many competitors.

First of all, this is something new for at least for management levels or leadership levels. It's not something that is a proven methodology, although it has a lot of years. So among the data scientists, it's already old. But for the decision makers, this is not something that they learn in school usually. The difficult part is that they kind of look at it and say that, okay, but if I don't do anything...

Not doing anything, this is our biggest enemy. Because not doing anything, it's easy. It costs you nothing, almost, at least not at the beginning. So they say, well, leaders can handle the organization without necessarily having a network understanding of the whole thing. It will work. Good or bad, you don't even know whether you lost some potential.

But then after we manage to do this diagnosis, we just make it a picture with, of course, various pictures from various angles, but it will be in a timeframe that is, I don't know, one week, two weeks long, depending on the size of the organization. But then you need to do some interventions because much like you do in the medical

medical profession. You go do a diagnosis, you do an MRI, and then you have to go to, of course, someone who reads the whole MRI and understand what the problem is. And then if there are some real issues, you need to address them and

And this approach of organization network analysis allows you to make precision interventions within clear spots within the system without addressing the whole organization. But those interventions sometimes

they need some specialized professionals. This is where our partners, like we have partners in Brazil, in South Africa, in Kenya recently, and Morocco, and in Romania, of course. So with our partners, we can extend the interventions within the organization after we do the diagnosis. And then we do it again after usually one year,

or continuously if the client agrees to let us connect with APIs to their systems and try to get some data continuously, which allows us to see whether there are some changes in the meantime. Frankly speaking,

Things do not change that much in organizations unless you have a project of transformation that is very focused or limited in time.

Have you seen any cases where leaders decide to make pretty radical changes in response to your analysis? Oh, yeah. Probably another very interesting example is the case for an investment fund, a private equity fund, that hired us to do an analysis on a company that they were buying. So they buy the transportation company from Romania. They wanted to invest 15 million into this company.

They had some sense that they wanted it, but they were afraid that this may be a failure because they were afraid that this might collapse based on their analysis. So they did the normal due diligence, financial, legal, and so on. Almost all instances in due diligence and M&A's

When they look at the people, they simply rely on some tables with people, a chart eventually with the organizational chart, and that's all.

Usually M&A's or investments of this kind fail and the rate of failure is very high due to the human factor, not necessarily due to the financial analysis that they did before. So they asked us to do this evaluation of the organization using organization network analysis.

how well and with whom is the CEO connected because they wanted to get rid of the CEO and they were very interested in understanding who are the key players in the organization that this guy is connected with. After they discovered that the guy was not what he was telling or selling to them, they simply decided to remove him quickly.

Just by removing that person in two weeks rather than in one year or two years like they were planning to do so, it makes their investment in our analysis worth. There was another question. Okay, this is transportation business. We need to have these operations functioning. We cannot stop the contracts that continue. The organization has to function.

And if the CEO that we fire is going to live with 10% of the drivers, we are in big trouble. So their problem was who can we nominate or who should we nominate? Of course, they had two options, someone from the internal or someone from the outside. But they decided they want to focus and find the right person from the inside to take over. So they did some, of course, research.

They had some interviews, they had some discussions, but then they turned to us and asked us, do you see anything in the data in Orgexo that would be a red flag for this person? Okay, so they thought they found the right person in their discussions. And I was looking at the data and said, yeah, he's an influencer. Yeah, makes sense. But then I hit the button about bottlenecks and we discovered that he's the main bottleneck.

And then when you know that he's the main battle, you start digging into the feedback that colleagues provided when they answered about him. And they discovered that if they would nominate that person, probably a lot of people would leave the organization because they nominated that person. So it was not evident only from the discussions with the people around because you cannot discuss with everybody. While our approach means that you collect data from everybody and big data is

when you want to aggregate them, provides a lot of data that you can rely on to make decisions. So in the end, we recommended someone and their first reaction, I do remember like,

Like it was yesterday. And the first reaction was, oh, no, no, that guy, he's too weak. He's supported by all the drivers and everybody's very happy with him. And I said, what's wrong with him? Because you want to nominate someone at the interim until you bring someone professional. The idea is that the interim would have to keep the organization floating and functional.

The fact that you have someone, you know, has support from almost everybody around in the organization, that's ideal. But they said, but he's not a CEO because he doesn't know a lot of things that the CEOs should know, like in finance and a lot of other stuff. And I said, but

Things like that you can learn. After one year and a half, the guy was still the CEO. They never replaced him. Now, after two and a half years, he's still the CEO and probably they want to replace him because, you know, the CEO job you can learn. But the human relations job, it's much more difficult to learn if you don't have it. Well, Gabriel, where's the best place for people to go online to learn more about the tool? First of all, it's the website, orgxo.com.

It's a solution that is on Amazon Web Services. We are GDPR compliant, or they can look me on LinkedIn. Gabriel Petrescu is my link. And yeah, so we are on all the platforms. They can book a demo with us on the website anytime. Very cool. I have links in the show notes for listeners to follow up. Thank you. Yes, thank you so much for taking the time to come on and share your work. Thank you, Kyle. It was really a pleasure. Thank you.