with Ulrich Irnich & Markus Kuckertz
Shownotes
In this episode, Markus and Uli talk with Ahi Gvirtsman, Co-Founder and Chief Knowledge Officer at Spyre Group, about why the traditional Proof of Concept no longer works in today’s innovation landscape. They explore how companies can move from slow, over-engineered PoCs to fast, low-cost experiments that create real learning and impact. Ahi explains why AI only delivers value when it truly changes how you work and why a great demo means nothing unless it drives real decisions, transformation, and measurable business outcomes.
Ahi’s new book: https://www.amazon.de/Spark-practical-handbook-managers-innovation-ebook/dp/B0FVW61NFP/
Connect on LinkedIn:
- Ahi Gvirtsman – linkedin.com/in/ahigvirtsman/
- Ulrich Irnich – linkedin.com/in/ulrich-irnich
- Markus Kuckertz – linkedin.com/in/markuskuckertz
Alle podcast episodes: digitalpacemaker.de
We would love to hear your feedback. Let’s talk!
Transcript
Ahi Gvirtsman:[0:00] So, you know, I think that when we talk about AI, it’s a bit misleading to think that AI is somehow fundamentally different than other disruptive technologies
Ahi Gvirtsman:[0:10] that we experienced in the past. I understand that its effect could be much more revolutionary than anything else we experienced in our lifetime. But at the end of the day, it is still a technology that has to be adopted inside an established organization. So it really doesn’t matter what type of technology we’re talking about. the principles remain the same.
Markus Kuckertz:[0:45] Welcome to the Digital Pacemaker Podcast. Hey, Uli.
Ulrich Irnich:[0:49] Hey, Markus. How is it going?
Markus Kuckertz:[0:51] Thanks, Uli. Our guest today is Ahi Gwurzman, co-founder and chief knowledge officer at Spire Group. With over 30 years of experience, Ahi has built a reputation for turning ideas into impact. He holds a Bachelor of Science in Computer Science from the Technion and an MBA from Tel Aviv University and also several books about innovation management. At Spire, he and his team help organizations to turn employee ideas, external technologies and new business models into scalable ventures, led by trained innovation leaders from within the company. It’s already his second time at the Digital Pacemaker Podcast. Welcome, Ahi Gwitzmann.
Ahi Gvirtsman:[1:29] Thank you. Thank you for having me, Markus. Great to see you again. Hi, Uli. Great to see you too.
Markus Kuckertz:[1:36] So he will raise the question, is the proof of concept dead? And we will speak about the following topic. The traditional PUC is dead, is expensive, slow, over-engineered and too often ends in the infamous valley of death. Innovation today requires faster, cheaper and smarter experiments that focus on learning, not just feasibility. Secondly, we talk about AI only delivers value when it changes how you work. Companies need to stop thinking of AI as a shiny tool and start rethinking their business and operating models around its potential. Without transformation, AI is just theater. And lastly, a great AI demo means nothing until it connects to real decisions. If your AI stays in the lab, you are not innovating. The challenge is not just building it, but embedding it into workflows, metrics and leadership priorities. Uli, how many PUCs have you seen in your career that look promising, only to quietly disappear before anyone made a decision?
Ulrich Irnich:[2:35] Well, I saw a lot of them, right? And I stopped counting because too many, right? And, you know, the proof of concept gives you some kind of certainty because you check the system overall and does it make sense or not, right?
Ulrich Irnich:[2:50] But interesting piece which is not answered is how you scale it and how you come really to value, right? Because at the moment, if you see in the industry, there are two metrics which are key for success right first is really the time from let’s say sandbox or whatever into production that’s number one and number two is from production into value is the second key metrics what all i would say all c levels are very interested in right and the number of pocs because most of them and ahi will talk more about that most of them are very isolated and just looking at a few things, but it doesn’t mean it flows and scales over time, right? And that’s, I would say, the most thing. And I got really tired over time for the next POC and the next POC and the next POCs. It’s like, it remembered me a bit like a management consultant where you have a lot of PowerPoints, right? You look at it, you have a good view on it, and then you take it under the desk, right? And then it’s there, right? And it’s a pleasure to have you again here in the Digital Pacemaker podcast. And great to see you.
Ulrich Irnich:[4:06] Before we start with the first question, now you triggered my interest with your new book, right? Because I had already my question in front of me, but I couldn’t ask it because I want to know a bit about your book today.
Ahi Gvirtsman:[4:22] Sure. So again, thank you for having me. It’s great to be back. So the book is called The Spark, a practical guide to anyone who was tasked with leading innovation, but no instructions on how to do it. And I like to say it’s a book that I wish I had back in 2012 when I started running innovation, HP’s global software division. And I wanted to make innovation management more accessible. So it’s more like a story. It’s a story about a mid-level manager called Maya who gets the task to run innovation, do something with innovation and she gets help from an experienced mentor. And there are also technical parts in the book that explain things more at the technical level. But I really wanted the book to be easy to read and accessible and have a lot of information and I really packed it with everything that I could think of that is relevant and is based also on many years now that I’ve been working with innovation managers in various industries, various countries. And I hope it turned out okay. I guess I will get the feedback later on.
Ulrich Irnich:[5:27] I’m pretty sure, I’m pretty sure. I know that innovation is really to the core of your heart, right? And you have run that for decades, right? And have seen, I would say, the good, the bad, and the ugly on that, right?
Ulrich Irnich:[5:42] Now, the question for me is a bit, what does really innovation mean to you? And what generates real outcome from your perspective?
Ahi Gvirtsman:[5:51] So, I think it’s a really interesting question. I mean, I wouldn’t say what it means to me. I tried to define innovation in a way that would make sense to everyone. and not just to me.
Ahi Gvirtsman:[6:01] And actually, not very long ago, I came up with, my team and I, we came up with a very, I believe, clear definition of what innovation is in organizations. Because when you work with organizations, especially the ones that are advanced technologically, or you work with a department like IT that is working with the latest technology, you have advanced R&D in your organization, it’s very difficult to say, you know, if you’re an innovation manager, this is where I end and you begin, right? This is innovation, and it means that this is not innovation. So it generates some pushback in the organization. You get less cooperation. People like you a little less because you imply that what they’re doing is not innovation. So what we chose to do is to say, okay, look, we’re not trying to take ownership of the term innovation, but we’re calling it entrepreneurial innovation inside an organization. What does it mean to be an entrepreneurial project? It means that you are, if I could give the intuitive explanation, this is a project that is outside the organization’s comfort zone. How do you break down the word comfort zone into things that are objective? There are five axes. One is the technology. Is this a technology that the organization has experience with or not? The other one, the next one is the business uncertainty. Is this something where you know that if it works technically, how sure are you? that the business aspect of it is going to be successful.
Ahi Gvirtsman:[7:29] Then there is how fitting is it for the organization? How many people are going to have to change their behavior in order for this to actually work, right? Is this something that you could just turn on and people will start using? And another one is, are there internal functions that might object to the project outright? So you have safeguard functions in the organization, right? So if you take all of this and look at it together, then you can then assess how far out of the organization’s comfort zone is this project, or is it just advanced R&D? With all due respect, right? There is advanced R&D, and it’s great, and it’s innovation. But you can apply standard project management techniques to a project like that. And if it is entrepreneurial innovation, it means that the appropriate tools and techniques are the ones that come from the startup world and were adapted to a corporate environment. So it was kind of a long answer, but I think that if you look at it through this lens, it takes this really amorphous term innovation and really brings it down to earth in a way that you and I can have an objective conversation.
Ulrich Irnich:[8:50] And I like that, Ahi, to be honest, because if I condense now what you explained a bit, first of all, it’s an inclusive approach, right? So that there is no just an innovation department and the rest of the company, it’s all over the place, right? And the other thing, which is more important, it’s a kind of mindset and an
Ulrich Irnich:[9:10] attitude, how you drive things, right? And how you move yourself out of your comfort zone.
Markus Kuckertz:[9:14] Maybe, Uli, before I start with Ahi, I have another question to you. What does PUC thinking at a company mean? What is meant by that?
Ulrich Irnich:[9:22] Yeah, first of all, if you look at companies, right, you have different profiles of companies, right? You have a very bold company who is driving very much innovation, right? And this is, especially for startups, essential to survive in that kind of environment. And then you have, I would say, corporates, which are already very mature. And some of the corporates are already in a kind of, I would say, risk-averse mode, right? And a POC proves that the thinking behind the concept is proven and even double-proved, right? And then you can make a decision. And you see already in that kind of logic that there is a failure in between, right? Because you are debating instead of experimenting and prove it, right? And, you know, if you are just doing that in a sandbox, you are not getting smarter, right? Because you are still working in your environment with your knowledge. But as soon as you are… Proving that kind of thing with your customers, ideally, right? And getting experience into that or other business partners or client partners, right? And get feedback to that. Then really innovation starts to work, right? Because you adopt, you not just take it for granted. It’s a flow.
Markus Kuckertz:[10:48] Ah, and based on your observations and experiences, why do so many POCs fail in the ballet of death? And what’s, from your perspective, a better alternative?
Ahi Gvirtsman:[10:58] Yeah. So you asked in the beginning whether the POC is dead. I think that in the context of innovation, it’s not dead. It’s just misunderstood. Yeah. If I could really try to point out the main weakness of the traditional POC in the context of innovation, it is that the traditional POC focuses on, is this going to work? It’s a technical approach. So I have an MBA with a specialty in IT management. So when you study IT management, POC is one of the things that you study. And POC comes from the point of view of, we have this technology that we want to apply to the organization. let’s run a POC to test that the technology works.
Ahi Gvirtsman:[11:40] Now today in most organizations unless you’re really in SpaceX maybe technologies that you’re going to deploy in your organization or implement yourself you’re probably going to do it. Your IT team will be able to create an AI model right so you know that this is technically feasible 99 times out of 100 so it’s not the main question whether you can do it. The main question is, can you do it here in this organization and make it deliver value? And you can test that at an order of magnitude less investment in much, much shorter timeframe with a much higher level of learning at the end that will allow your executives, the investors that Uli mentioned that are ignored in the POCs, to make much higher quality decisions about which POCs to invest in going forward. And another thing that we need to make a mind shift is that in the traditional sense, if you start a process in an established organization and something comes in through one end of the process and doesn’t come out of the other side, it’s a bad thing.
Ahi Gvirtsman:[12:56] So if you start a POC and the POC failed, that’s not a positive experience, especially if you invested a relative high amount of budget and resources on it. But if you reduce the cost by an order of magnitude, if you make it run over a duration of not more than three to four calendar months, then even if you fail at the end, it’s a learning experience and you’re running a funnel. And if you understand that it’s a funnel, you understand that during the stages of the funnel, you’re going to weed out the weakest options. And if you run, we like to use the acronym SMART. So let’s say we call it a SMART POC. The term I like to use is an experiment.
Ahi Gvirtsman:[13:37] But some organizations, they have pilots. So we say, okay, keep calling it a pilot. We’ll call it a SMART pilot. Okay. You have a POC, we call it a SMART POC. The SMART is an acronym. I invite you to read my book to see the actual meaning of the acronym. But the idea is a lot cheaper, a lot faster, with clear KPIs that executives care about. They approve what success looks like before you run the smart POC. And if you adhere to these principles, then something really magical happens because executives like this format. It gives them control. They understand that they’re managing a portfolio and And they understand that to keep trying again and again in various ways is a good thing for the organization. So I think this is like the main thing that you can do to change the traditional POC approach. And then if you think about it, if I’m an executive and we started by, I was presented with a few options to run experiments. I approved the experiments and what success looks like for each one. I followed up. I see the outcomes. Let’s say that they hit my thresholds. Now I’m a lot more motivated and engaged to make that next step of investment, which leads us to, you know, overcoming the valley of death, as you called it, Marcus.
Markus Kuckertz:[14:58] So what does a successful or smart early stage experiment look like in your framework related to budget, timeline, outcome, and so on?
Ahi Gvirtsman:[15:07] So as I said, you know, what I like to do is I take the approach of asking the people in the organization, usually my contact, the innovation manager, what is the acceptable amount that people ask for? So usually when somebody comes in and asks to run a POC, how much budget do they ask for? What is the range that you usually ask for? So let’s say I worked with a hardware company, and they said, we like to ask for about $200,000, $250,000. That is the traditional request because we have labs and we have equipment and so on. And I said, okay, so 10%, an order of magnitude less. You cannot ask more than $20,000 to $25,000 to run your smart POC, to run your experiment. And it takes time for people to adjust because they’re not used to thinking this way but when i explain to them that the purpose is not to test whether it is going to work technically.
Ahi Gvirtsman:[16:05] But to say let’s assume that it works technically but let’s test whether it’s actually going to deliver value where let’s test whether it’s going to work in this organization then the mindset set shifts and they come up with clever plans that adhere to these standards. And another thing that is really important is that before we even start an experiment, we start earlier on by, first of all, working on problems that are aligned with what the executive team in that organization in that period of time, what they care about, right? So we, from the beginning, we bring options that are aligned with the executive priorities.
Ahi Gvirtsman:[16:54] And then when experiments are proposed, an executive has to sponsor it. That’s the S out of the SMART acronym. So we need an executive sponsor. If you don’t have an executive sponsor, you’re not going to run an experiment. I don’t care how much I like the idea as an innovation manager. The executives are the internal customers of innovation. As an innovation manager, I work for them. So if I don’t have an internal customer, there is no point in proceeding. Right so in that way i start generating momentum and engagement from the executive team and that allows me to make progress after successful pocs so.
Markus Kuckertz:[17:35] Willie what would you say is there still any case where a classic proof of concept makes sense or should we retire the term
Ahi Gvirtsman:[17:41] Completely, It’s like I said, right?
Ulrich Irnich:[17:45] If you make it in that new definition to make it actionable, and of course, with a fast feedback loop, right, it still makes sense. But I agree with you. I would say, except rocket science or something else, most of the technology will not fail anymore, right? So it’s not a proof that the technology will work, right? It’s more how able is the organization to adopt to that new style and make them adoptable to that kind of things. Maybe you have read the recent new MIT study on generative AI POCs, which was really, or they call it projects, not POCs, right? 95% fail, right? And now you ask your question, why does that happen? And it’s not why the technology is not capable to fulfill that. It’s the way how companies are able to adopt to that new way of working and also to use that kind of technology. And it’s, of course, also that the organization needs to learn how to scope with that kind of new technology. But at the end of the day, I would say technology is only a very small portion of the success formula. it’s more the adaptivity of the organization.
Ahi Gvirtsman:[19:10] So I could give you a simple example, and I think it’s very applicable to a lot of AI-based projects nowadays. So let’s say that you want to create a conversational AI that will help a certain team with a certain task. And the traditional approach would be to say, let’s take all of the data, all the information, let’s create an initial model. And then once we create the initial model, let’s have a few people play around with it and let’s see if they like it, if it creates value and so on. What I would say is create like a chat environment and have a person use predefined scripts for a limited period of time, let’s say for a week, okay? They will dedicate a week of their work time to address questions and issues from the team, simulating the AI, being the AI.
Ahi Gvirtsman:[20:06] What we will what we will learn from this is are people using it if you build it will people use it if you give them these types of answers will they be happy with it we can measure maybe it’s something where we can measure the actual effect because one of the things that people miss oftentimes is they say let’s use ai to do x okay they don’t take the time to really dig into this x and say, why is this a must problem for us? Why is this not just a nice-to-have problem? If it’s just nice to have, it’s going to be difficult to go beyond the value of death. If it’s a must problem, it means that there is an important KPI attached to it. So one of the things I can measure during this short and cheap experiment is to see whether if I provide that support, simulated by a human, can I actually measure a positive effect on that KPI? So imagine the difference between saying, we created the model, it’s partially correct, people like it, as opposed to, we simulated the model, and we were able to reduce the response time of the team by 10%. Completely different conversation. And now you have an engaged executive that understands exactly what is in it for them if you actually put in the work and the effort over time to create that model. And you also know, we call this operational fit.
Ahi Gvirtsman:[21:33] That’s the A out of the SMART acronym so i’m giving everything away so you also know already the third exactly right exactly i’m giving you the entire chapter 10 all of chapter 10 i’m giving it away for free, So that’s another thing that you learn. Will people actually use it? Which is something that you miss when you focus on technology, because we forget that oftentimes when you’re trying to do an innovative venture, people inside the organization, there are functions that are going to have to change their behavior for this venture to be successful.
Ahi Gvirtsman:[22:13] So if you can incorporate testing the operational fit as part of the experiment, it becomes much more valuable. and in order to do that you need to stop focusing
Ahi Gvirtsman:[22:23] on the technology and start focusing on the value so.
Markus Kuckertz:[22:27] Talking about value let’s switch to ai from your observations and companies where do most companies go wrong in trying to do something with ai
Ahi Gvirtsman:[22:35] Yeah so you know i think that when we talk about ai it’s it’s a bit misleading to think that ai is somehow fundamentally different than other disruptive technologies that we experienced in the past. I understand that its effect could be much more revolutionary than anything else we experienced in our lifetime. But at the end of the day, it is still a technology that has to be adopted inside an established organization. So it really doesn’t matter what type of technology we’re talking about. The principles remain the same, right? Even the example that I just gave you is based on a conversational AI solution. So at the end of the day, it’s a disruptive technology. Organizations need to adopt it. And unless they decide to invest in some top-down mega play, like a change management, AI transformation, like we did, you know, in older days with IT transformation.
Ahi Gvirtsman:[23:41] They’re going to have to take the entrepreneurial approach. I mean, that is what at least we recommend, which means that instead of dictating AI top-down across the board, which probably we’ll get to that someday in some organizations, what you need to do is to identify where today using AI can give you the biggest impact in the shortest period of time. So if you look at AI, it’s a technology. Where can I apply this technology? In my finances, in my operations, in my customer support, in my R&D, where and to what ends can I apply AI today in this organization so that it will give me the biggest effect? Why is this so important? Because as you said before, there is a valley of death. And we already know that 70, 80% of AI-based projects fail, right? So it’s not like it’s a magical thing that allows you to be successful at a high ratio.
Ahi Gvirtsman:[24:43] It fails at a very high ratio, like technologies did in the past. So it means that we still need to apply the same principles that have proven themselves to be effective for previous technologies for this technology as well. What would you say?
Markus Kuckertz:[24:58] What does it make to move from an AI pilot to meaningful business transformation?
Ahi Gvirtsman:[25:03] So it really is all about, you know, being very focused on doing just enough, just enough technology, because any iota that you add to the technology capabilities add complexity, adds budget, adds time, adds risk. So this is why it’s so important. So think about startups, okay? Startups can’t afford to spend a dime more than necessary they have a limited runway they have a limited budget and so that’s why the successful startups are so clever in how they structure their value proposition they’re they’re laser focused on what is the value that they deliver they’re laser focused on their target audience and if they find that they miss the mark they immediately pivot and they do something different. So these projects, these AI-based projects must follow the same principles. You have to be laser-focused about what is the problem that you’re solving, who you’re solving it for in the organization, and keep testing yourself. And if you are not accurate, pivot. Don’t try to do more. Again, the corporate thinking.
Ahi Gvirtsman:[26:13] Is the more I say I’m going to do, the more interesting I’ll become, the more budgets I will get, right? The more executive attention I will get. But let me dispel this notion. The more budget you get, the less patience executives are going to have with you. The more you say that you’re going to do, the higher the expectations and the lower the threshold where they’ll say, okay, let’s stop. So actually being small is an advantage. If your spending velocity is lower, it means that you have more runway, you have more time, you have more chances to make mistakes. And executives will look at you as a much more, let’s say, you’re much more
Ahi Gvirtsman:[27:00] mindful about their budget and their resources. And so I think that they will appreciate you more. Uli, as a former executive, I would love to get your feedback on that statement.
Ulrich Irnich:[27:11] It is, especially because when you prove that you’re realizing things, because talk is cheap, we know that, right? And budget is also a thing which gives you room to maneuver, but it doesn’t mean you’re adding value to the company, right? And therefore, execution is king, and successful execution is king. And I like your view to say small is beautiful because it gives you always the power to move yourself out of your comfort zone, right? Because this kind of appetite, this kind of curiosity, and what you said, especially on startups, right? It is their surviving strategy to be laser-sharp focused to deliver something which customers are buying, right? And that’s something… You need to continue with this kind of small, right? And therefore, this kind of lean budget and all kinds of things which saying, okay, you get less, but if you are successful, you get the next chunk. If you are successful, you get the next chunk.
Ahi Gvirtsman:[28:17] Exactly.
Ulrich Irnich:[28:17] Gives a different, I would say, speed and attitude as well, right? Because you are focusing on the value of the company. Yeah.
Ahi Gvirtsman:[28:25] And another thing, I think a mistake that people keep doing is they get so enamored with the technology that they forget to justify the value. So I meet so many teams that say, we’re going to use AI to do this. And I say, why? What is the problem? Well, this is manual today. It’s not efficient. It’s not a good use of our time. And I tell them, listen, this might be an issue for you. It’s not an issue for your management. Your management pays you for your time. There’s always something that you can make more efficient in a large organization. Don’t talk to me about nice-to-haves. Everybody knows that we can make this more efficient. It’s a nice-to-have. Tell me what is a must-have. Don’t talk to me about efficiency. Talk to me about the fact that today, because of the lack of AI, the team makes mistakes. And when they make a mistake, it costs the organization a lot of money. Talk to me about…
Ahi Gvirtsman:[29:20] You know, something where you could win business opportunities at a higher rate, close more deals, right? Talk to me about generating revenue or saving the organization tons of money. Something that when you talk about it to executives, they will shift in discomfort in their seats. That’s how I teach people to talk to management. I tell them, I want you to make them feel embarrassed that this is happening in their organization today, if you can find it, right? Find a problem that is so painful that they will be compelled to act. And if you can demonstrate it, you can remedy this problem, then they will support you over the valley of death and to eternity. But you have to be laser focused. It should be something extremely painful. And then you should prove, you have to prove that you are the one who is going to solve it for them.
Markus Kuckertz:[30:13] So, yeah, seeing AI and the value is one thing, but what would you say are the critical blockers that prevent AI from being operationalized? And how would you deal with them? What do you recommend?
Ahi Gvirtsman:[30:23] So when you talk about blockers, I think it’s a technical word. At the end of the day, I like to call them, they’re like safeguard functions in the organization. Every organization has functions whose job it is to protect the organization from litigation, from breaking regulatory constraints, from making cybersecurity errors and so on. You have the legal department, regulations, data privacy, data security, and so on and so forth. And usually when we talk about innovation, these functions are regarded as kind of, you know, they’re the wet blanket of innovation, right? They’re always, they’re the ones who say no to everything and they don’t cooperate and so on. And I look at these functions from a different perspective and I say, you know, they’re doing their job. Our job is to demonstrate to them that we are responsible corporate citizens, and that what we are about to do falls within the constraints of their requirements. So I know that sometimes this is nearly impossible to do, but first of all, if you look at how we structure an experiment.
Ahi Gvirtsman:[31:34] You structure an experiment also by simplifying it. You tried to delay any collisions with those blockers to a later stage. Why? Because if I can, for example, as by the way, we did in one of the successful pilots at Vodafone at the time, we take historical data, we send it to a startup, we know what the analysis that we already did on this historical data, what results it gave us. We ask the startup to analyze it offline. We anonymize the data so there’s no private data. And then we check whether the startup was able to generate results from the historical data. So that is an approach that allows you to do an experiment fast and cheap, not collide with data privacy, not collide with data security. You don’t have to install or allow access to the startup to any live data, which, by the way, if you think about it, that’s the traditional POC approach. Let’s bring the startup in. Let’s kick the tires, take it for a spin, connect it to the data, and let’s see what happens. But what we’re saying is let’s avoid any delays, any roadblocks early in the process. Why? Because if we complete the experiment.
Ahi Gvirtsman:[32:40] And it is successful, now we can go to management, to the executive sponsor, and if there is any conflict that we cannot resolve, or if there’s a delay because we need resources from IT and we can’t get them, now we have an executive in our corner who is the customer of this venture, who is enthusiastic and excited because we demonstrated to the executive what the outcome is going to be, what value this is going to generate if we take the further steps and we actually deploy and start deploying it or do like the traditional POC, right? So the blockers are just doing their job. We need to try to delay any conflicts to as late in the process as we can and work in cooperation with them. So for example, when we train, you know, innovation champions, innovation leaders, innovation coaches, you know, every organization has its own terminology, we make sure to have representatives from these teams as innovation coaches. Why? Because then we can, I like to call it to indoctrinate them in the innovation ways, in the innovator’s way, so that now we have somebody who understands both worlds. They understand, for example, the legal world, but they also understand the innovation approach. So they can actually help us sometimes to mitigate. They’re very valuable resources, very valuable participants once they become part of the innovation network.
Ulrich Irnich:[34:05] And therefore, i.e. early involvement of this kind of protection functions in the process, which makes them part of the journey of the experimentation, helps, right? To get acceptance, that’s number one. And number two, it helps you also to might already build in functions which are necessary later on in the stage when you scale, right? Because they will give you all your information. Information but imagine yourself in the shoes as a protector right and somebody worked already half a year in such kind of pilot right and then comes to you and say here now we are surprise surprise right and and that’s that’s also what i always recommend right that people if you know if you know you’re working with data with sensible data why not involving friendly friendly guardrails right into that kind of process which helps to make the process faster yeah
Ahi Gvirtsman:[35:04] Yeah absolutely i mean the way we guide internal entrepreneurs when they come up with a value proposition we ask them okay who could object let’s think in advance what safeguard functions might object to this project and if we can think about these functions let’s go talk to them let’s ask them what needs to happen for you to support this? What are the boundaries within which I can operate in order for you to support this at least until I complete the experiment? And then we can talk, right? But then I have, if I’m successful in the experiment, when we talk, I also have a supporting executive. And oftentimes, what the legal department simply wants is the executive to say, listen, it’s on me. I take responsibility. Give him a waiver. Let them proceed, right? But the executive will only do that if they get tangible understanding of what the potential is if we take that next step following the smart POC.
Markus Kuckertz:[36:06] So you mentioned data and insights. Can you share an example where a small AI experiment led to a real operational shift?
Ahi Gvirtsman:[36:13] Yeah. So I work with a maritime port. And, you know, what we do is not just internal development, it’s also the deployment of external technologies, usually startup technologies. The traditional approach that organizations take when they need a technology, an external technology, is they write a document. In a lot of organizations, it’s called a brief. and you define exactly what are the requirements from an external technology in order to fulfill a certain task. In their case, what they wanted was to have video cameras pointed at cranes unloading merchandise from a cargo ship and tally the number of elements that are being unloaded automatically using AI that is analyzing the video. They created a brief. If the brief had every possible type of merchandise that you can think of, every shape, every kind, in every situation, lighting, etc., and so on and so on. So they created something that could not really be fulfilled, right? The technology was not ready.
Ahi Gvirtsman:[37:18] So they said, okay, we looked for startups, but we couldn’t find any. When we came in, we apply a technique that we like to call focus, right? We apply Pareto’s law. We asked them, is there a certain type of merchandise that if you could only count that type right now, it would still generate a lot of value for you? And there was. And it was a very expensive type of merchandise. It has a distinct shape. And so when we brought in the startup technology, which was evolving, by the way, was learning out to identify more stuff, just by applying that to a smart POC and being able to count that particular type of element was very valuable for the organization. So again, an entrepreneurial approach. Instead of saying, we’re not going to do this until the technology can do everything, we assume that the technology is evolving. We focus on where it’s going to bring us the biggest impact, which is already feasible. and we start from there and we start expanding afterwards.
Markus Kuckertz:[38:16] So if you look like 5, 10 or 20 years on, what would you say from a company perspective? What could be your vision or innovation? What must be the key change companies should follow or what should they trigger already today? What’s your view?
Ahi Gvirtsman:[38:36] So I like the term ambidextrous organization, right? There are things that you need to do in the existing business in order to thrive in the short and medium term. There are certain things that organizations need to do. Their operations, their products, their services, whatever. They need to keep optimizing and doing it. At the same time, they have to constantly assess what opportunities lie in front of them and what threats are relevant for them. And the way to do that is by running an innovation funnel that assesses both internal opportunities for development and also external technologies that you can use for various purposes in the organization. And the thing is that it is much less expensive and complicated than people imagine. You can run a very cost-effective funnel, at the end of which you can decide, you can really cherry-pick exactly which investments you want to make. And still, the investment in the funnel will have brought a lot of learning to the organization.
Ahi Gvirtsman:[39:44] And the vision that they have for several years now is that this should be the customary thing in every organization. I mean, just like you have marketing and you have sales, you have operations, you need to have innovation in the sense that we talked about it, right? In the sense that you’re kind of running an internal VC for opportunities to improve your product services, increase your revenues, increase your profitability, become more efficient and so on. But constantly doing that and constantly investing in scaling the best opportunities that you have as a VC, gradually, in stages. And it’s a skill. It’s a skill that organizations need to develop. And I think that innovation management, first of all, I think it’s an amazing role to have if you do it properly. That’s why I wrote my latest book, The Spark. And I think it really gives a very nice overview of all of the topics that you need to master and to understand in order to be an effective innovation manager.
Markus Kuckertz:[40:53] So thank you very much for the exciting conversation. Let’s do a joint wrap-up. Uli, what are your key takeaways?
Ulrich Irnich:[41:01] Well, there are a lot of key takeaways, right? But I would say let’s summarize it to three, right? There’s a new definition of smart and you can read it in Ahi’s book, right? We get already a glimpse that S is for sponsor, M is for must-have or must-die situation and A is the kind of adoption, right? But R&T, you need to figure out if you read the book. But nevertheless I would say the second summary is POCs are not that you need to adjust them into smart POCs which means you get going you get fast results you make it smaller and get experience because in the world we are living in you need to learn right because when you when you see one thing what is happening at the moment with all the AI development and everything what’s happening You think you know a product, right? But after three months, it’s already so much enhanced that your knowledge is already outdated, right? So that means that’s a fact, right? And that means you need to keep up and you need to run such kind of experimentation to keep yourself up to date.
Ulrich Irnich:[42:16] And number three is it’s not about technology. It’s all about adoption of the organization.
Ulrich Irnich:[42:23] We have grown up in organization verticals, right? But in the world we are living in now is more a horizontal world. So that means departments need to work cross-functional together. And your knowledge you are gaining with this kind of cross-functional improves the company. And that’s very essential because we see otherwise a big shift between technology speed and also the adoption of the organizations. And, you know, to be successful also in the future, you need to work cross-functional. And that means not IT alone. IT together with business and all critical functions which are involved into that. That makes you more successful. And you need to buy Ayes book.
Ahi Gvirtsman:[43:10] Thank you, Uli.
Markus Kuckertz:[43:13] Any further thoughts from your side?
Ahi Gvirtsman:[43:15] I really liked Uli’s comment about the horizontal cooperation. I think it’s very astute observation and I like to say that innovation is an extreme team sport. You really need many different functions to pitch in. You can’t do it by yourself simply because it’s not just the technology. You need a lot of functions to cooperate in order for this to be successful. Some by allowing certain blocks, you know, and, and, and, uh.
Ahi Gvirtsman:[43:43] And delays to be removed some by contributing their subject matter expertise some by generating connections to the right people in the organization and of course the leaders to make the right decisions but this has to work as an this has to be a network it cannot be hierarchical you cannot you cannot command innovation into existence there there is an organic element to it there There has to be something that happens horizontally, as Uli said.
Markus Kuckertz:[44:11] So, and if anybody is interested to read your book, where can you buy it or where can you access it?
Ahi Gvirtsman:[44:18] So you can get it on Amazon. You can look for the book, The Spark, and my name, Achi Gewerzmann. Hopefully you’ll have a link to it in your website as well. For the Hebrew readers, there is going to be a version in Hebrew in the Ivrit website, which is coming very, very soon as well. So if somebody is looking for a reference, you can also ping me on LinkedIn and I’ll be happy to send you a reference to the book.
Markus Kuckertz:[44:44] Great. And we will definitely include in our show notes. So don’t worry about that.
Ahi Gvirtsman:[44:48] Wonderful.
Markus Kuckertz:[44:49] That will be a really great link to Amazon. Ahi, thank you very much for being our guest and thanks for the time and insights.
Ahi Gvirtsman:[44:55] Absolutely. It’s always a pleasure to talk to you guys. And again, thank you for inviting me.
Markus Kuckertz:[45:00] That was the Digital Pacemaker podcast with Ahi Gewürzmann. and follow us now on your favorite podcast platform and never miss an episode. Thanks for listening and see you soon. Yours, Uli and Markus.
