Latest / Growth After Headcount / The 5 Tiers Of AI Vulnerability — Bram Kaashoek (Main Capital) #008

Transcript

The 5 Tiers Of AI Vulnerability — Bram Kaashoek (Main Capital) #008 · Growth After Headcount

• Bram: Bram, thank you very much for inviting me to the main capital of the C &D Hard to record this podcast. Good to have you, Harris. My name is Bram Kaashoek, Haris Odobasic: Thank you very much for inviting me to the main capital of the sea in the Hague to record this podcast. Good to have you, Harris. My name is Bram Gaslow. Bram: one of the parents here at the firm, and I'm also CEO and lead the operational team here at Maine. Which is perfectly relevant for what you're going to talk about. Haris Odobasic: So CO and Elite, the operational team. Which is perfectly relevant for what we're going to talk about AI in businesses. And you just told me before that this building and especially also this room has a very special history. Can you tell us again? Yeah, of course. it's a very, very special building. Bram: AI businesses and you just thought before that this building and especially also this room has a very special history. Can you tell us about Yeah, of course. So it's a very, very special building. So this was the workplace of the prime minister until somewhere in the seventies. And this place in this room, there was like a meeting of all the ministers like every month and the picture in the back, you see them sitting there all together. So this is a place where history was Haris Odobasic: somewhere in the 70s and this place in this room there was like a meeting of all the ministers like every month. The picture in the back you see them sitting there all together so this is a place where history Bram: was written. That's very fascinating and you like replicated the picture right? Exactly, so we have on the right hand side the picture with the current partner team of Main Capital and that will grow of course over time. Haris Odobasic: was written. It's very fascinating and you like replicated the picture right? Exactly so we have on the right hand side the picture with the current partner team of Main Capital and that's where we grow of course over time. Okay beautiful I like the story. Bram: So I will start actually from a little bit difficult question or challenging question. I will read this one out because you are familiar with YC, they plan a startup incubator in the world and every year they're publishing a guidance to which companies they are going to invite. And this year's summer guidance said... Haris Odobasic: So I will start actually with a little bit difficult question or a challenging question. I will read this one out because you are familiar with YC, the largest startup incubator in the world. And every year they're publishing a guidance to which companies they're going to invite. And this year's summer guidance said, in the words of Jared Friedman, Bram: In the words of Jared Riedman, Haris Odobasic: AI has collapsed the cost of producing software by a factor of 10 to 100x. That's the change he's seeing. And his simple recommendation is, clone an existing product and sell it for the one tenth of a price. You guys are investing a lot into SARS and many incubants. They're doing this for years, Bram: AI has collapsed the cost of producing software by a factor of 10 to 100x. That's a change we've seen. And his simple recommendation is clone an existing product and sell it for the 1-10th of a price. You guys are investing a lot into SARS and many incubants. They're doing this for years, How does this statement make a difference? Haris Odobasic: How does this statement make you feel? Thanks. Bram: Thanks. Yeah, well, I do believe indeed that AI will change the economics of software development. So it's easier to write code. It's cheaper to document, to ship new versions to the market. So then that indeed will impact the way how software is created and developed. Haris Odobasic: I do believe indeed that AI will change the economics of software development. So it's easier to write code, it's easier to document, to ship inversions to the market. So then that indeed will impact the way how software is created and developed. Bram: But I do believe that the opportunity is bigger, not only for AI native companies, but also for current. Haris Odobasic: that the opportunity is bigger not only for AI native companies but also for current Bram: software companies. What I mean with that is that in the past or historically software was always seen as a tool, like rendering software for designers or bookkeeping software for accountants or contract review software for lawyers and so on. Haris Odobasic: software companies. What I mean with that is that in the past or historically software was always seen as a tool like rendering software for designers or bookkeeping software for accountants or contract review software for lawyers and so on. But now with agentic AI, see a trend from software as a tool for professionals to drive their productivity to software that acts like a platform to execute Bram: But now with agentic AI, see a trend from software as a tool for professionals to drive their productivity to software that acts like the platform to execute Haris Odobasic: the job to be done. And that means that software companies can tap into way larger budgets, only IT budgets of their customers. HR budget and that sort of Exactly, exactly. So we definitely will change the Bram: the job to be done. And that means that software companies can tap into way larger budgets, not only IT budgets of their customers. exactly. So we definitely will change the game Haris Odobasic: and rules of the game. But this is not only for AI native companies but also for existing software businesses. Bram: and rules of the game. But this is not only for AI native companies, but also for existing software businesses. Haris Odobasic: That makes a lot of sense. And what advantages do you think existing software businesses have against those new challenges? Yeah, couple of things. So I do believe that it's not just about creating the product, it's about building a business. Bram: That makes a lot of sense. what advantages do you think existing software businesses have against those new challenges? Yeah, a couple of things. So I do believe that it's not just about creating the product, it's about building a business. And we invest in companies that deliver solutions to enterprise customers. And then it's also about the trust. It's also about the relationship. It's also about Haris Odobasic: companies that deliver solutions to enterprise customers and then it's also about the trust, it's also about the relationship, it's also about deep domain expertise, it's also about the delivery capability. So are you able as a company to deliver the solutions and critical work processes of these enterprise customers? And that's exactly I think where the boat sits of the current software vendors and they know Bram: deep domain expertise. It's also about the delivery capability. So are you able as a company to deliver the solutions in critical work processes of these enterprise customers? And that's exactly, think, where the mode sits of the current software vendors. They know Haris Odobasic: the market, they know the work processes, Bram: the markets. They know the work processes. Haris Odobasic: they know how to deliver value in these processes and they can increase their added value from there. So if you embed AI in your solutions and if you increase the value that you deliver, we also have the opportunity to monetize on that. So I do believe that indeed there will be Bram: They know how to deliver value in these processes and they can increase their added value from there. So if you then embed AI in your solutions and if you increase the value that you deliver, you also have the opportunity to monetize on that. So I do believe that indeed that it will be rising stars in the AI native field. But I also believe that current software companies are well positioned to benefit from the opportunities that AI will bring. Haris Odobasic: stars in the AI native field but I also believe that current software companies are well positioned to benefit from the opportunities that AI will bring. Bram: Yeah, definitely. And I think there are a few... Haris Odobasic: Yeah, definitely and I think there are a few use cases that are easier to challenge and you guys have factored that in with your five level investment category if you can walk us through that. Yeah, of course. So, late last year we developed our own Bram: use cases that are easier to challenge and you guys have in fact that in your five level investment category? Yeah, of course. So, late last year, we developed our own AI assessment framework and it's a way to look at the market to kind of categorize the market of enterprise software. And we have defined five levels in tier one. We have companies that Haris Odobasic: assessment framework and it's a way to look at the market to categorize the market of enterprise software. And we have to find five levels. In tier one, we have companies that Bram: deliver Haris Odobasic: deliver system on record. So critical solutions and without these solutions, enterprise organizations cannot operate. If the system is down, the customer is in serious problems, right? Bram: system or records. So critical solutions and without these solutions enterprise organizations cannot operate. If the system is down, the customer is in serious problems, right? Then in tier two, you have solutions that support critical processes. In tier three, you have solutions that are delivered to specific Haris Odobasic: Then in tier 2 you have solutions that support critical processes. In tier 3 you have solutions that are delivered to specific organizational departments of an enterprise organization. In tier 4 you have productivity tools. And in tier 5 you only have solutions that automate one specific task. Bram: organizational departments of an enterprise organization. In tier four, you have productivity tools. And in tier five, you only have solutions that automate one specific task. And we typically invest in category one through three. the real enterprise grade solutions. Haris Odobasic: We typically invest in category 1 through 3. the real enterprise great solutions. And we also see that the AI vulnerability is different for these tiers. So in tier five and in tier four, it's easier to come up with an AI native tool that replicates or substitutes that process. And in one or two, it's way different and it's way harder. You cannot simply replace your ERP, for example, if you are like a big manufacturing company. Bram: And we also see that the AI vulnerability is different for these tiers. So in tier five and in tier four, it's easier to come up with an AI native tool that replicates or substitutes that process. And in one or two, it's way different and it's way harder, right? You cannot simply replace your ERP, for example, if you are like a big manufacturing company. Yeah. And also if you would like to replace... Haris Odobasic: No. And also if you would like to replace, Bram: maybe there is this competitor who has exactly the same use cases and is maybe not 30-40 % more affordable, what would still protect the incubator that they don't switch? Do you see some factors? Yeah, it's again a couple of things. So I already mentioned the deep domain expertise. Haris Odobasic: maybe there is this competitor who has exactly the same use cases and it's maybe not 30-40 % more affordable, what would still protect the incubant that they don't switch? Do you see some factors? Yeah, it's again a couple of things. So I already mentioned the deep domain expertise. Bram: but also the capability of product management. So really understanding how you help your customers and what's the value that you deliver. And from there, continuously make the product better. Haris Odobasic: the capability of product management, so really understanding how you help your customers and what's the value that you deliver and from there continuously make the product better. What we read now a lot in the newspapers is that AI native companies they come up and they have a very modern and innovative solution and often it's positioned a bit black and white like the current software companies don't innovate anymore. Bram: What we read now a lot in the newspapers is that AI native companies, they come up and they have a very modern and innovative solution. And often it's positioned a bit black and white, like the current software companies don't innovate anymore. But that's not the truth, right? So we invest in more than 50 companies and I see innovation all over the place. So these companies are agile as well. They also have R &D budgets. They also are eager to do something with new technology like AI. So it's not that the current vendors in software stand still. Haris Odobasic: But that's not the truth, right? So we invest in more than 50 companies and I see innovation all over the place. So these companies are agile as well. They also have R &D budgets. They also are eager to do something with new technology like AI. So it's not that the current vendors in software stand still. They also develop their solution continuously. Bram: they also develop their solution continuously and with AI that's going faster and faster, right? So in your introduction, you mentioned that it's way easier to create software and that's exactly what software companies are doing. So they have their current offering and then on top or layered in, they come up with new features, new product, new functionalities and they monetize on the value that they deliver and... Haris Odobasic: with AI that's going faster and faster. So in your introduction, you mentioned that it's way easier to create software and that's exactly what software companies are doing. So they have their current offering and then on top or layered in, they come up with new features, new product, new functionalities and they monetize on the value that they deliver. We see a couple of examples in our portfolio. For example, Fortes, it's a company based here in Holland. They are specialized in PSA, professional service automation for account firms. They integrate with CRM, for example. They notice that especially account firms... Bram: We see a couple of examples in our portfolio. So for example, Fortes, it's a company based here in Holland. They are specialized in PSA, professional service automation for accountancy firms. And they integrate with CRM, for example. And they notice that, especially accountancy firms, they like specific functionalities in a CRM. So they decided to create a CRM solution themselves. Haris Odobasic: they like specific functionalities in a CRM. So they decided to create a CRM solution themselves. And they Bram: they did with the help of AI and in a period of I think three, four months time, Haris Odobasic: did it with the help of AI and in a period of three, four months time. Bram: they had a product ready to ship it to the market and they now sell it to their current customer base. So what you described in terms of the product velocity and develop a productivity in AI native companies, that's what I also see in the current landscape basically. Yeah, I think what I'm also observing is Haris Odobasic: They had a product ready to ship it to the market and they now sell it to the current customer base. So what you describe in terms of the product velocity and develop a productivity in AI native companies, that's what I also see in the current landscape basically. Yeah, I think what I'm also observing is if delivering a product is easier, like the Bram: If delivering a product is easier, like the... development that's democratized. It's way easier to build. What becomes the premium is domain expertise. Because if you have done this for 10, 15 years, you may even learn something on the way. Exactly. And this one is way harder to replicate, right? That's one part. And the second part I would like to add is your delivery capability. Haris Odobasic: development got democratized. It's way easier to build. What becomes a premium is this domain expertise, because if you have done this for 10, 15 years, you maybe learn something on the way. this one is way harder to replicate, right? That's one part. And the second part I would like to add is your delivery capability. So you create a product, Bram: So you create a product. But then bringing that product to the market and making sure that is well embedded in enterprise organizations And that customers can trust it that they can reach out to you if they have questions That if the system is down, they know who to call Haris Odobasic: but then bringing that product to the market and making sure that it's well embedded in enterprise organizations and that customers can trust it, that they can reach out to you if they have questions, that if the system is down they know who to call. Bram: Right that part of organizational maturity and and delivering capability is also very very relevant in the area of enterprise software So so yes, it's on the product side Haris Odobasic: That's part of organizational maturity and delivering capability is also very, very relevant in the area of enterprise software. So yes, it's on the product side related Bram: related to deep domain expertise. But next to that also, it's Haris Odobasic: to domain expertise, but next to that also, yeah. super important that you have this delivery capability at the highest level required for selling solutions to enterprise organizations. I think in the delivery there are like in a way also two pieces. The one piece is once they are customer onboarding them continuously delivery of the service and reliable. Bram: super important that you have this delivery capability at the highest level required for selling solutions to enterprise organizations. I think in the delivery there, in a way, there's two pieces. The one piece is once they are a customer, onboarding them, continuously delivery of the service and reliable. On the other side, it's also Haris Odobasic: On the other side it's also Bram: how do you actually sell the deal? Like the go-to-market piece. And that piece is also big because if everyone can build, well, everyone also needs to sell the product. And I think there... Haris Odobasic: how do you actually sell the deal? Like the go-to-market piece element. And that piece is also big because if everyone can build, well, everyone also needs to sell the product. And I think there... Bram: I think there it's like it can go both ways. can be innovative, creative in getting your product out. The AR and ATVs also have a chance. But the thing in the enterprise domain, it's also a little bit more special of how you sell it, Correct. Yeah. So in go-to-market, indeed, we talk about products, go-to-market and delivery, basically. And if you then zoom in on Haris Odobasic: I think there it's like it can go both ways. Like you can be innovative, creative in getting your product out. AI native VMs also have a chance. But the thing in the enterprise domain, like it's also a little bit more special of how you're selling, right? Correct, correct. Yeah. So in go to market, we talk about products, go to market and delivery basically. And if you then zoom in on Bram: the go-to-market part, especially in enterprise environment, Haris Odobasic: The go-to-market part, especially in enterprise environments, Bram: Of course, you can do a lot with your marketing campaigns and the way how you do the sales enablement and prospecting and that type of stuff. But then once you have a foot in the door and you have the room to give a demo, for example. Haris Odobasic: of course you can do a lot with your marketing campaigns and the way how you do the sales enablement and prospecting and that type of stuff. But then once you have food in the door and you have the room to give a demo, for example. Bram: then it comes down to really understanding the pains and the gains of the customers and showing a demo that resonates, right, that clicks. And that still is, in my perspective, human interaction and understanding the real... Haris Odobasic: then it comes down to really understanding the pains and the gains of the customers and showing a demo that resonates, that clicks. And that still is, in my perspective, human interaction and understanding the real... Bram: use cases of customers and making sure that you have a strong pitch there. And of course you can automate and you can create like segmented demos or personalized demos. But in the end, in enterprise environments, it's also a matter of relationship building and making sure that you build that trust. Yeah, I think that's very relatable. Also, when I think of my own... Haris Odobasic: use cases of customers and making sure that you have a strong pitch there. And of course you can automate and you can create like segmented demos or personalized demos. But in the end, in enterprise environments, it's also a matter of relationship building and making sure that you build that trust. Yeah, I think that's very relatable. Like also when I think of my own consulting services in DevOps is like at some point, Bram: consulting services in Revops is like at some point Revops Haris Odobasic: it's like you are talking to people and you know with people it's not like as one person is going to be 10x better than the other Bram: is like you are talking to people and you know with people it's not like as one person is going to be 10 weeks better than the other. Haris Odobasic: like it's still human knowledge which is limited to some extent so a distinguishing factor of course is also like do I trust that person as well and that is very universal of B2B software also in order level you also have to work with that person together right and if you don't trust the team there's no Bram: it's still human knowledge which is limited to some extent. So a distinguishing factor of course is also like do I trust that person as well. And that is very universal of B2B software. Also in order to work with the person together, Exactly. If you don't trust the team, there's no connection there and having these capabilities and if you have a very young team who also does not maybe invest Haris Odobasic: connection there and having these capabilities. And if you have a very young team who also does not maybe invest into more senior people who can do this relationship building and can get political as well, everything, you need many things you need to consider and that's still intuition and AI is not good at intuition. Bram: into more senior people who can do this relationship building, can get political, everything, you many things you need to consider and that's still intuition. AI is not good at intuition. Exactly, and I mean, I can imagine that in your job in Rev-Ops had the same situation as in selling and bringing enterprise solutions to the market in the sense that yes, of course, you can rely on the data, you can do something with systems and automate process. Haris Odobasic: in your job in Redbox had the same situation as in selling and bringing enterprise solutions to the market in the sense that yes of course you can rely on the data, you can do something with systems and automate processes Bram: But in the end it's about collaboration, it's about people, it's about trust, right? It's Haris Odobasic: but in the end it's about collaboration, it's about people, trust. Yeah, definitely. I think what AI native companies do very well Bram: super important. Definitely. I think what agronomical companies do very well... Haris Odobasic: is really efficient in the internal operations. How can they go to go to market? How can they build a product? And I think the classical model in SaaS is like, if you look at the 2010s, like Bram: is really efficient in the internal operation. How can they go to go to market? How can they build the product? And I think the classical model in SaaS is like if you look at the 2010s, like I think the flag was like how many employees does my company have? Because I have more employees, it means like we can build and ship and deliver more. That script switched. You can do more, but you still of course Haris Odobasic: I think to flex was like how many employees does my company have because I have more employees. It means like we can build and ship and deliver more. I think that script switched because you can do more, but you still of course need the people for your high value strategic topics for stakeholder management, delivery, et cetera. Bram: people strategic topics for stakeholder management delivery etc. And what I am observing still is to making this transition for companies to become leaner that that one is a difficult one. Yeah it is. Haris Odobasic: What I am observing still is to making this transition for companies to become leaner that that one is a difficult one. Yeah, it is. Bram: I do believe, by the way, that this is not something that you... Haris Odobasic: something that you need to do or have to do overnight. You can do that in a more gradual way. Indeed, we also look for efficiency in our organizations. By the way, we only invest in profitable growth companies, so there was always a thesis that we supported. Bram: need to do or have to do overnight, right? So you can do that in a more gradual way. But indeed, we also look for efficiency in our organizations. And by the way, we only invest in profitable growth companies. So that was always a thesis that we supported, right? So yes, of course, we want companies to grow, but we also want companies to do that in a very profitable way. So looking at the organization Haris Odobasic: Yes, of course, we want companies to grow, but we also want companies to do that in a very profitable way. So looking at the organization and Bram: and how lean certain teams are is always part of our playbook. And with AI, I think that's kind of in the line Haris Odobasic: how lean certain teams are is always part of our playbook. And with AI, think that's kind of in the line. Bram: more and more. And we know that in certain departments, you can be more efficient. For example, in your support organization, you already mentioned your development organization, your go-to-market, especially enablements capabilities can be way more efficient. So indeed, Haris Odobasic: certain departments you can be more efficient. example in your support organization, you already mentioned your development organization, you go to markets, especially enablements capabilities can be way more efficient. So indeed and making Bram: that transition is for certain organizations I think a bit harder. Haris Odobasic: the transition is for certain organizations I think a bit harder. Bram: What I do observe is that enterprise software organizations are still pretty agile. They like tech, they like Haris Odobasic: What I do observe is that enterprise software organizations are still pretty agile. They like tech. They like... and to improve continuously, that's think way different than Bram: to improve continuously. That's I think way different than real enterprise organization, like very, very big multinational, for example, right? So these companies are still quite used to adapt, learn, adjust processes, automate things. Haris Odobasic: real enterprise organization, like very big multinational for example. So these companies are still quite used to adapt, learn, adjust processes, automate things. Bram: So yeah, although this is a shift that they have to make, they are first of all willing to do so. And what I observe is that they are making the steps. I think from my experience, these steps, they're going to take time and a lot. Haris Odobasic: So yeah, although this is a shift that they have to make, they are first of all willing to do so. And what I observe is that they are making the steps. No, I think from my experience, these steps, they're going to take time and a lot of effort and also a lot of money to make the steps. What I'm often observing is Bram: effort and also a lot of money to make the steps. What I'm often observing is Haris Odobasic: and a conversation with Jakob von der Koi from Winningbird Design. He loves the notion of building a revenue system, the whole concept around revenue architecture, which means basically that all your data is connected. Now we're talking about the commercial or Bram: in a conversation with Jakub von der Koi, he loves the notion of building a revenue system and the concept of revenue architecture, which means basically that all your data is connected. Now we're talking about the commercial. Haris Odobasic: marketing, sales, customer success, support, product data, feeding into a data layer that we can get insights. Bram: marketing, sales, customer success, support, product data feeding in into a data layer that we can get insights and base decision on it ideally in real time. This foundational data layer is really essential to rolling out an AI in motomarket. Haris Odobasic: and based decision on it, ideally in real time. This foundational data layer is really essential to rolling out an AI and go to market. The reality is how far many companies are from that is still striking and making this transformation because this transformation is required to grow. Bram: The reality is how far many companies are from that is still striking and making this transformation because this transformation is required to grow in a more efficient linear way. this joint is not only for the AI. Haris Odobasic: in a more efficient, leaner way. And this disjoint is not only for the AI Bram: Problem because historically it was an issue for people right marketing and sales is unaware of the campaign once they find out it's rolled out and doesn't make sense right yeah, this is a typical story or like something was sold by sales and they wanted to rise a customer success to onboard like we don't even support this use case it's in the roadmap for three months yeah, but these these joints Haris Odobasic: problem, because historically it was an issue for people, right? Marketing and sales is unaware of the campaign. Once they find out it's rolled out and it doesn't make sense, So these are the typical stories. Or like something was sold by a sales and then wanted to rise of customer success to onboard. Like we don't even support this use case. It's in the roadmap for three months. But these disjoints, Bram: and They are I think people struggle before that before AI Haris Odobasic: And they are, I think people struggled before before AI. Bram: I think AI now shows the urgency of this is really a blocking point. Yeah. And the question is, companies taking this really serious? do you mean? That's a fair point. And indeed having that data foundation in place is super important to base your decisions on. And what Haris Odobasic: think AI now shows the urgency. Okay, this is really a blocking point. And the question is, are companies taking this really serious? What do you really notice? That's a fair point. indeed having that data foundation in place is super important to base your decisions on. Bram: we see in our portfolio and also broadly in the market that Haris Odobasic: We see in our portfolio and also broadly in the market that the good thing for current vendors, like for the existing software companies is that they have the data, they have rich data because they have the product for many years, they have a customer base. Bram: The good thing for current vendors, like for the existing software companies, is that they have the data. They have rich data. Because they have the product for many years, they have a customer base, Haris Odobasic: They have prospects, know the market, they know the competition, so the data is there. Bram: they have prospects, they know the market, they know the competition. So the data is there. Haris Odobasic: The challenge is how do you get the data out of certain systems and how do you bring data sets together? So that are the two main questions that these companies have on the plate. If you then look at AI native companies or new businesses coming to the market, they have potentially the design for a strong data foundation. Bram: The challenge is how do you get the data out of certain systems and how do you bring data sets together? So that are the two main questions that these companies have on the plate. If you then look at AI native companies or new businesses coming to the market, they have potentially the design for a strong data foundation. Haris Odobasic: But as long as you don't have customers, you don't have the real data. So the benefit from the existing companies is that they have the data. It's a matter of bringing the data sets together. But I see as the biggest struggle in reality because Bram: But as long as you don't have customers, you don't have the real data. So the benefit from the existing companies is that they have the data. It's a matter of bringing the data sets together. And what I see as the biggest struggle in reality because Haris Odobasic: the orchestration is not the hardest part anymore. And you have tools and platforms for that. Bram: The orchestration is not the hardest part anymore. You have tools and platforms for that and providers like Snowflake and Databricks bringing these capabilities and data sets together and finding patterns in your data and do like the mapping and the matching and that type of stuff. The hardest part is get the data out of your product to really understand how customers use your product. Haris Odobasic: snowflake and data breaks, bringing these capabilities and data sets together and finding patterns in your data and like the mapping and matching and that type of stuff. The hardest part is get the data out of your product to really understand how customers use your product. Bram: And especially if you have Haris Odobasic: Especially if you Bram: like legacy solutions and especially if you still have on-premise solutions, it's really hard to get the data out. Haris Odobasic: solutions and especially if you still have on-premise solutions, it's really hard to get that data out. It's Bram: is super important. So we push for that like significantly, not only now, but for years. Because we believe that this data on users patterns learns you a lot about the added value of your software. Haris Odobasic: super important so we push for that like significantly, not only now but for years, because we believe that this data on users' patents learns you a lot about the added value of your software. Bram: Really understanding what are the use cases, what's the difference between a heavy user and a regular user, are the specific parts of your solution not used at all? How frequent are certain functionalities you Haris Odobasic: really understanding what are the use cases, what's the difference between a heavy user and a regular user, are the specific parts of your solution not used at all? How frequent are certain functionalities used? Bram: what are flows in your system. That's super relevant data and that provides lots of insights, not only for product development and product management, so making your product stronger, but also for Haris Odobasic: What are the flows in your system? That's super relevant data and that provides lots of insights, not only for product development and product management, so making your product stronger, but also for... Bram: the commercial teams like... Haris Odobasic: teams like churn reduction. So if I see a customer that has decreased usage in the platform, can that be a red flag that the customer dislikes the product and potentially churns? Bram: churn reduction. So if I see a customer that has like decreased usage in the platform, can that be a red flag that the customer kind of dislikes my product and potentially churns? So having these type of insights is super important for running a software company well for the product side and the go-to-market sides. But this is indeed for many incumbents the biggest challenge in my perspective. And I think some traditional software fix the code to get to there that sometimes I think tricky. Some people have logging data, some even don't check if they activated the license or not. So there are big differences and for some time Haris Odobasic: So having these type of insights is super important for running a software company well for the product side and the go to market sides. But this is indeed for many income, the biggest challenge in my perspective. Yeah. And I think some traditional software to fix the code to get to there that sometimes I think tricky, some people have logging data. Some even don't check if they activated the license or not. Like, so there are big differences. And for sometimes Bram: you have Haris Odobasic: you have You can have data of every module who's activated, how often it is used, who uses it, and that's what you want to have. And then you upsell, if there's something to upsell, many companies have something to upsell, it's like way, way, way easier. On the same time, I would also not underestimate the data orchestration. Bram: you can have data of every module who's activated how often it is used who uses this and that's you will have and then you sell if there's something to upset many companies have something to upset it's like way way way easier on the same time I would also not underestimate the data orchestration it's one side getting snowflake configuring everything still at this moment AI can do a lot Haris Odobasic: It's one side getting snowflake, configuring everything. Still at this moment AI can do a lot, but not everything in gathering the insight out. It's also where we are seeing a lot of companies struggling to really investing in a few data engineers that are supporting only the go-to-market work, not the other rest of the business. That's what I mean with significant investments that you need to hire data engineers. You probably going to need to... Bram: Everything in gathering the inside out. It's also where we are seeing a lot of companies struggling to really investing in a few data engineers that are supporting only the go-to-market or the other rest of the business. That's what I mean with significant investments there. You probably wouldn't need to... Haris Odobasic: revenue operations teams, if you don't have one, multiple people, to build the orchestration that those systems are connected, the processes are connected, and everyone is enabled, the enablement piece we talked before. And that's what I'm seeing now. There's more demand for this, to build this out. Yeah, I agree. And this is not just a matter of bringing the data together technically. It's also about Bram: revenue operations teams if you don't have one, multiple people to build the orchestration that those systems are connected, the processes are connected and everyone is enabled to enable the piece we talked before. And that's what I'm seeing now. There's more demand for this, to build this. Yeah, I agree. And this is not just a matter of bringing the data together technically. It's also about understanding what are the patterns you are looking for. Haris Odobasic: understanding what are the patterns you are looking for. So Bram: What kind of user patterns are we analyzing? And on the basis of these insights, what do we decide on the product side and on the go-to-market side? So having this defined clearly is indeed a capability that we should bring on board in enterprise software companies. Haris Odobasic: what kind of users patterns are we analyzing and on the basis of these insights what do we decide on the product side and on the go-to-market side. So having this defined clearly is indeed a capability that we should bring on board in enterprise software companies. Bram: So it's not just about bringing data together technically, you're right. Haris Odobasic: So it's not just about bringing data together technically. Yeah. And yeah, Bram: What do you think can incubants learn from aerated companies? I think if existing companies are in the market for many years, they are run like an institution. They Haris Odobasic: And what do you think can incubants learn from AI native companies? I think if existing companies are in the market for many years, they are run like an institution, right? Bram: know the market, they know the customer base, they know their partners and so on. Haris Odobasic: they know the customer base, they know their partners and so on. Bram: What I think is refreshing to learn from AI native companies and startups in general is their focus on finding a solution for a specific need. And this is a discussion that we have frequently in our portfolio as well. Haris Odobasic: What I think is refreshing to learn from AI native companies and startups in general is their focus on finding a solution for a specific need. And this is a discussion that we have frequently in our portfolio as well. Bram: So what's now the added value that we deliver? Can we write that on paper sharply? What's the single use case that we want to make the difference for? Haris Odobasic: So what's now the added value that we deliver? Can we write that on paper, sharply? What's the single use case that we want to make the difference for? Bram: So that's one thing, having a very clear focus on use cases, added value, the impact basically that you make for your customer. And the second thing is, like you mentioned, the Haris Odobasic: So that's one thing, having a very clear focus on use cases, added value, the impact basically that you make for your customer. And the second thing is, I'd like to mention the... Bram: kind of day one mentality. Haris Odobasic: kind of day one mentality. And what I mean with that is that you as an organization behave like it's day one, you are excited, you are eager, you wanna deliver, you wanna make impact. So that kind of entrepreneurial spirit, day one mentality is what I think incumbents can learn from startups or AI native companies in general. Bram: what I mean with that is that you as an organization behave like it's day one, you are excited, you are eager, you want to deliver, you want to make impact. So that kind of entrepreneurial spirit, day one mentality is what I think incumbents can learn from startups or AI native companies in general. Haris Odobasic: Yeah, you can maybe guess the next question. What can AI native companies from Bram: The other way around. What can SaaS, AI-native companies learn from SaaS companies? Yeah, from existing companies. So I think a lot in the sense that Haris Odobasic: SaaS companies? Yeah, from existing companies. So I think a lot in the sense that Bram: these companies you were talking about, the SaaS companies or the incumbents or the existing software vendors, Haris Odobasic: these companies you are talking about, the SaaS companies or the incumbents or the existing software vendors. Bram: they started as a startup once, right? And they went through a journey, through a scale-up phase, and they became like institutional organizations with a very strong delivery capability with the brands, the products, the customer base, delivery capabilities, and so on. So I think what startups can learn from these type of companies is how do you navigate through Haris Odobasic: they started as a startup once, right? And they went through a journey, through a scale-up phase, and they became like institutional organizations with a very strong delivery capability with the brands, the products, the customer base, delivery capabilities, and so on. So I think what startups can learn from these type of companies is how do you navigate through difficult times and how do you build consistency in your delivery? Bram: difficult times and how do you build consistency in your delivery so that customers can really trust on the products or the solutions that you bring to the markets. So what kind of organizational capabilities do you build to have that trust and to build that relationship? So professional services, your CSM, Haris Odobasic: so that customers can really trust on the products or the solutions that you bring to the market. So what kind of organizational capabilities do you build to have that trust and to build that relationship. So professional services, see them really understanding the problems of customers from product perspective. Bram: really understanding the problems of customers from a product perspective. These type of capabilities, I think, are typically well developed in the current landscape for current software vendors. Haris Odobasic: These type of capabilities I think are typically well developed in the current landscape for current software vendors. you recently heard in the news that ClickUp is laying a off like 20 % and most of them middle management because they are betting all okay we need our expert individual contributors so we're gonna make them more Bram: recently he's heard in the news that ClickUp is slaying a off like 20%. And most of them middle management. Yeah. Because they are betting all, okay, we need our expert individual contributors. So we're going to make them more... Haris Odobasic: powerful. and they should act quicker, faster. And now they're removing the middle layer. And the podcast book is also called, Growth After Head Count. So how do you see this topic of restructuring? Yeah, think in general, in enterprise software, organizations are already pretty Bram: and you should act quicker, faster. And now they're removing the middle layer. And the podcast book is also called, Growth After Headcount. So how do you see this topic of restructuring? Yeah, think in general in enterprise software organizations are already pretty... Agile. Haris Odobasic: agile. So in the market and also in our portfolio, we already see companies that have like 30-35 % EBDA. And that's not just because the product is so scalable, it's also because you're organized in a more lean or manner. So the different layers that you typically see in like enterprise organizations and corporates, that's what I don't see so often in Bram: So in the market and also in our portfolio, we already see companies that have like a 30, 35 % EBDA. And that's not just because the product is so scalable. It's also because they are organized in a more lean manner. So the different layers that you typically see in like enterprise organizations or corporates, that's what I don't see so often in Haris Odobasic: a lot Bram: like mid-market software companies. Haris Odobasic: mid-market software companies. Bram: What I do believe is that if these companies make a change and if they adopt AI in their internal processes, of course you can do way more with the same team or do the same with a smaller team. And then you can, based Haris Odobasic: What I do believe is that if these companies make a change and if they adopt AI in their internal processes, of course you can do way more with the same team or do the same with a smaller team. And then you can, based on... Bram: on your business case and your growth trajectory and your plans, you can make a decision. Do we do more with the same team or do we let go certain people in specific areas and do Haris Odobasic: business case and your growth trajectory and your plans, you can make a decision. Do we do more with the same team or do we let go certain people in specific areas and do Bram: more and do the same with a smaller team. And what we see in our product development teams, in our engineering teams, is that Haris Odobasic: more and do the same with a smaller team? And what we see in our product development teams and our engineering teams is that Bram: the thesis is mainly let's do more with the same team, but in supporting areas. Haris Odobasic: the thesis is mainly let's do more with the same team. But in supporting areas like customer support, for example, typically you can do the same with a smaller team. And then you have to restructure indeed. Yeah, that's like a process you have to go through then as an organization. Yeah, I think it's also like Bram: customer support for example, typically you can do the same with a smaller team. And then you have to restructure indeed. And yeah, that's like a process you have to go through then as an organization. Haris Odobasic: How much of an opportunity does AI unlock for a specific business and industry? It might even create completely new use cases, right? But if, for example, we talk about the sufficiencies that YC mentioned 10 to 100 times, like, okay, maybe that's true, but the market is definitely not growing at the rate of 10 to 200X, right? So there are... Bram: How much of an opportunity does AI unlock for specific business and industry? It might even create completely new use cases, right? True. But if, for example, we talk about the sufficiencies that Weiss mentioned 10 to 100 times, okay, maybe that's true, but the market is definitely not growing at the rate of 10 to 200 eggs, right? So there are Haris Odobasic: probably some who will probably be a bit on the losing side and on the winning just like in a way basic economics and numbers. Exactly, Yeah, I'm with you. But again, and this was my answer to your first question. I don't believe this is so much about the efficiency game. This is not so much about productivity. This is about finding new pathways for growth for existing companies. Bram: probably some who will probably be a bit on the losing side and on the winning just like in a way basic economics exactly yeah yeah i'm with you but again and this was my my answer to your first question i don't believe this is so much about the efficiency game this is not so much about productivity this is about finding new pathways for growth for existing companies And we see that in different dimensions. So I already mentioned the Fortress example where they did product expansion, right? So they created the CRM product next to the PSA offering. We also see portfolio companies that can unlock a new category that arises now with AI. So we have, for example, a company in the Nordics called Point Chop. Haris Odobasic: And we see that in different dimensions. So I already mentioned the Fortress example where they did product expansion. So they created the CRM product next to the PSA offering. We also see portfolio companies that can unlock a new category that arises now with AI. So we have, for example, a company in the Nordics called Foinshop. Bram: and they deliver identity and access management software. So managing identities of users in enterprise organizations. And now they Haris Odobasic: and they deliver identity and access management software. So managing identities of users in enterprise organizations. And now they... Bram: target managing the identities of non-human agents in these environments too, right? Because agents are running around in the IT landscape and do certain tasks. Haris Odobasic: target managing the identities of non-human agents in these environments too, right? Because agents are running around in the IT landscape and do certain tasks and Bram: And that's also one part of your ecosystem that you have to manage. What are the rights that you provide to these type of users in your system? So they grow into that kind of new category that didn't exist like three, four years ago. Haris Odobasic: that's also one part of your ecosystem that you have to manage. What are the rights that you provide to these type of users in your system? So they grow into that kind of new category that didn't exist like three, four years ago. Bram: And then we have companies that tap into like new markets. Haris Odobasic: And then we have companies that tap into new markets. Bram: that like I mentioned shift from software as a tool to software to execute certain tasks to be done. And then we have one example in the portfolio called Woodwing. And Woodwing is specialized in software for advertising companies, for media organizations, and publishing houses. And they created a tool to do the editing part of your front page of newsletters. Haris Odobasic: that like I mentioned shift from software as a tool to software to execute certain tasks. And there we have an example in the portfolio called Woodwing. And Woodwing is specialized in software for advertising companies, for media organizations, and publishing houses. And they created a tool to do the editing part of your front page of newsletters. And instead of providing the platform just to do that as an editor, they created a platform that automatically comes up with a layout. So takes over part of the work that was done by the editor in the past, right? And I think the beauty of this example is that it not only creates a productivity gain for the customer, Bram: And instead of providing the platform just to do that as an editor, they created a platform that automatically comes up with a layout. So takes over part of the work that was done by the editor in the past, right? And I think the beauty of this example is that it not only creates a productivity gain for the customer, It also creates an extra window for the newspaper to publish the latest news because typically in the past, the editorial team started their job and that takes like three, four, five hours. Now with tool, we have extra hours to keep the window open for news. Haris Odobasic: It also creates an extra window for the newspaper to publish the latest news because typically in the past the editorial team started their job and it takes like three or five hours. Now with Toolbook we have extra hours to keep the window open for news. Bram: Right? So the added failure that you bring as an organization is more and you can monetize on that. So. Haris Odobasic: Right? So the added value that you bring as an organization is more and you can monetize on that. So Bram: What I see is that yes indeed this is a productivity game changer, but it also unlocks opportunities for value creation Yeah, okay. Let's try to summarize what we just talked about and maybe give winning recipes for incubates. think one of the first points is being in the right Haris Odobasic: what I see is that, indeed, this is a productivity game changer, but it also unlocks opportunities for value creation. Okay. Let's try to summarize what we just talked about and maybe give winning recipes for incubates. think one of the first point is being in the right Bram: industry in a way, like would you describe level one to three? Yes. It can probably be described in the term stickiness. Haris Odobasic: industry in a way like what you described level one to three like which can be probably described in the term stickiness Bram: We have a product that is sticky. Yeah, mission critical. Exactly. We have the big domain expertise, like the years of experience, having this knowledge, which probably is not, you cannot just JGPD it and find it. Exactly. We have really unique data and experience. that is- And the Haris Odobasic: we have a product that is sticky yeah then exactly we have the big domain expertise like the years of experience being having this knowledge which probably is not you cannot just jpg it and find it so you have really unique data and experience so that this one Bram: network. And the network, which kind of is the next thing, trust and brand, you are this respected authority. Haris Odobasic: and the network, which kind of is the next thing, trust and brand, you are this respected authority in that space, right? Why would we start with a nobody if we have this trustworthy party? I think also time is also an advantage because we didn't talk that much about it, but for critical software, it's not something you are changing on a monthly, yearly basis. You keep those tools for years, if not... Bram: Yes. Nobody if we have this trustworthy party. I think also time is also an advantage because we didn't talk that much about critical software. It's not something you are changing on a monthly basis. You keep those tools for years if not decades eventually. Like we see hospitals running on software from three decades ago. Maybe they need an update but that's a different topic. Haris Odobasic: decades eventually, like we see hospitals running on software from three decades ago. Maybe they need an update, but that's a different topic. Bram: Yeah, right. And we have now also the final, like the fifth one, there are many more, that AI gives also the opportunities to those companies and they have the spirit to develop new use cases. Exactly. That they can apply in the existing product or... Haris Odobasic: And we have now also the final, like the fifth one, maybe there many more, that AI gives also the opportunities to those companies and they have the spirit to develop new use cases that they can apply in the existing product or. unlock a new vertical because if they have to finance that, add the CRM to it because let's look at Salesforce HubSpot for many businesses, it's overkill. Like they need two, three modules and they don't need the rest. Right? So in this, Bram: they have to finance that right now, add the CRM to it because let's look at Salesforce or HubSpot for more than any businesses it's over care like two, three modules they don't need the rest. Exactly. So this If you follow these five points, you're in a good position. Would you add anything else? Yeah, so I would like Haris Odobasic: if you follow, guess, these five points, you're in a good position. Would you add anything else? Bram: to add one. And you mentioned that in your fifth point, it's also the agility. Right? So I believe that also the incumbents have the agility and the innovative kind of position. Haris Odobasic: fifth point is also the agility. I believe that also the incumbents have the agility and innovative position to create something new and to innovate and to bring these new solutions to the market. Bram: to create something new and to innovate and to bring these types of new solutions to the market. Haris Odobasic: Yeah, so it's like really the right team. Like if you had a good team the last 20 years, you've made it so far, chances are like, okay, this team can also innovate in action like this entrepreneurial day one spirit you mentioned. Great. This is an optimistic end. love it. Exactly. Bram, thank you for finding your time. Thanks Bram: So it's like really the right team. Like if you had a good team the last 20 years, you've made it so far. So that's it. like, okay, this team can also innovate and actually like this entrepreneurial day one. Exactly. Yeah. The day one mentality. Right. This is an optimistic end. love it. Thank you for finding the time. Thanks Haris Odobasic: for a... Bram: for Haris Odobasic: And this is for the listeners, this was the second page. We did already one and failed due to some technical issues, but I think this time we are good. Let's hope so. Thank you, Errols. Thank you.