Webinar: Winning with Corporate Startup Alliances

So winning with corporate startup partnerships. You might have seen the title change a couple of times. We were trying to see, okay, what actually makes most sense and what represents the content in the best way possible. So what are we going to talk about? Well, we're going to talk about open innovation and a particular part of open innovation. And I'll present myself. So I'm Arne, or Arnie if you want. I'm partner of, of doi and yeah, I've been working in, I would say in, in many different industries when it, when it comes to corporate startup partnerships. Yeah, Emily.

Yeah, hi everyone. Yeah, hi. I'm an associate director at BOI and I'm actually leading most of our work with corporate startup collaborations, especially in very complex environments. So happy to tell some war stories from me and sharing some of the learnings we got along the way today.

Yes. Yeah, before we start would actually be great. We have a poll here ready that will open up. When we saw all of you subscribing for the webinar. Yeah, we saw a lot of open innovation titles, a lot of other type of titles as well. So it would be great if we, if we could get an idea on how much experience you have when it comes to open innovation.

So just go ahead and vote. I think it's on the poll, within the chat there's a poll tab. So yeah, I see votes coming in. So there's some experience in organizing innovation challenges and idea contests, startup scouting and have collaborated with startups in a joint venture or partnerships. So more across the board.

So that's nice. Maybe on what we're going to present you about. So there's getting a lot of calls here. Like what we're going to talk to you about is very much a focus on like how I would say two corporates, big or small, work together rather than. Yeah, sorry, yeah, rather than. Oh yeah, rather than focusing on. Yeah, for example, in particular like crowdsourcing of ideas such as like legal, for example. Sorry, Emily, can you take over for a second because I keep on getting calls.

No worries. So what we're saying is there's many, many different ways in which open innovation actually happens. And as you know, these different words that you see here are different ways that open innovation happens. And I think that's important to know that there are. The one we're talking about today is really when a corporate works with a startups or scale up or external entity in a work format and not therefore going into hackathons ecosystem. Although that's very interesting to talk about. Just expectation Setting that's not what we'll talk about today as really we want to work on as we've realized also for the years for us it's really impactful when we get into piloting and actually going to market with such an external entity and therefore also what we want to talk about today. All right, maybe Arne, if you can move ahead. Or. Can you move ahead Arne? Thanks.

Yes, yes, I'll take over again here. Like we need to chat. So yeah maybe for you there is a lot of. There is a lot of different reasons why you would engage with external solution providers. Yeah. Can you maybe share like one of the top reasons for you from your point of view so there's no wrong answers here which is like would like to see like what the type of reasons are that either you work with external solution providers if you're a corporate or if you're not a corporate like why others work together with you.

So not enough resources in house for example as mentioned maybe other perspectives reducing costs, access to new technology. No expertise as expertise might be outsourced outside specialization. Yeah certain skill sets. I think it's going back about the resources specific skill sets and specialization. So definitely, definitely on point there time to market.

Yeah I would say there is a lot of here indeed on skills and resources. So you're definitely right there. I would say let's first zoom out and in the end everything is always about either increasing revenue or reducing costs. And obviously having the right technology to do something either way better would increase your revenue or like replace I would say a legacy technology which is costly. Same with time to markets. With time to market you you would expect I would say being quicker in the market than than other providers. Another corpus which so which would probably increase your revenue. Also development time would be shorter so that decreases costs. Again the whole point is here it's super important to know why you're doing it because if you know very well why you're doing it and you have a KPI linked to it then you also know for example like we can ask for budget internally it seems extremely basic but it's often attention to start with and then you have to go a bit more in depth and probably like adding some reasons that we heard from our clients as well over the past couple of years. There is for example an element like a bit more opportunistic but as a corporate you have certain assets that most startups don't have and I'm not talking about I would say the scale ups that have like huge VC funding Like from all over the world. I'm talking about, I would say yeah the spinoff coming from a university, having a good solution in markets but don't necessarily having the skills or the network to be globally active yet. Well as a corporate, if you become their first clients. Yeah. You might very well also help them to scale globally helping you as a corporate and at the same time also helping them to scale. So there, there is big win win in that collaboration and building on that. For example, like the open innovation programs that we had with or that we're running with with companies like Rush and Novartis there it even plays more overall because of the deep local regulatory expertise that is needed in order to be able to be active in the market. And this is true for both, I would say externally facing revenue generating services as well as what you can do in order to develop these services and then other things. Yeah, I've probably been mentioned like it's all, it's often about resource, it's about expertise, it's about let's say the yeah. Facing the reality of the best or the most niche talent that you need. Not wanting to work for a corporate like yours. And in this way you can, you actually can. So that's like these are relevant things to do. So then also I would say there is very small ways to collaborate with external companies and there is a lot between I would say having an agreement with a mature company and let's say on the other side acquiring a startup. And for us the nuance between all these is super important. What we will be talking to you about probably mostly today. So we'd say the smaller like the vendor supplier agreements or like licensing and distribution agreements that you would have with an existing, I would say technology provider or company. And here I would say the sweet spot is those companies that are mature enough to be on markets but in the same time not so big yet to be super experienced with. I would say the working with the big corporates that have huge procurement processes and I would say the companies, I would probably say no to the risk of working with such a small company. Yet there is also some context on why external collaborations and open innovation is becoming more and more relevant. And as you guess, technology is definitely related to it and mostly the speeds of acceleration when it comes to technology. So. So this is not only AI, this is just in general, if you look at Moore's Law there is like technology is increasing, it's more networked and so forth. And same time also obviously the last two, three years. And if you have been following us. We are all in on generative AI and autonomous systems and so forth. Yeah, well this is not only boi, this is also really there is so many AI related alliances and partnerships that are being built and that is just super, super interesting. And if you would think that it's only the small ones, think again. You all saw the news, this is not new for you. For example, it's super interesting to see, for example Apple being still the biggest company in the world. I would say playing, playing number one and two with Nvidia still saying we're not going to create our own LLM because others are better doing it. It's an interesting strategic choice, a choice that for example, Bloomberg burned itself on and this is more than a year ago now, but only a year ago where they said like, yeah, okay, we have a lot of data, great, we're going to build our own Bloomberg GPT with GPT 3.5 back in the day again, 18 months ago, sounds like ages ago. And then GPT 4 came out and it was better than what they spent more than 10 million on, which is painful, I would say. And then also in the same time, Notco, it's the article in the back there, look it up. For me it's one of the most inspiring, I would say startups of the last of the last couple years. So Notco is founded in 2017 so it's not too old as a company. And what they do is they as a mission, they say like we're going to replace all animal based molecules products by plant based ones and we're going to do this by using AI to formulate new recipes and they're doing this so well in the beginning with not milk and not burger and now they're having a joint venture together with Kraft Heinz to create for them not cheese for example and not macaroni and all these kind of things. And it's a nice joint venture and we work with Kraft Heinz and with the Nestle of this world, we know there are and the budgets and it's. Yeah, it's more than a billion of them. It's interesting to see that these companies choose to work with external providers to do this rather than doing it themselves just because it's quicker and often better as well. And by the time they get it to the level that it needs to be, then their competitors might have partnered with these startups and so forth. So let's say, let's get back to yeah, to our, I would say closer to home kind of examples because obviously we're not notco. We're not craft. You might be graph Heinz, but still, but like most of us are not, are not the CEO of Apple or crafty. So then say looking at like when do you work together with an external, with an external entity? First of all, if you have a super relevant challenge to solve internally, but where you just can't like building the solution yourself just doesn't make sense because you need more skill in order for the viability to kick in. So that's one big trigger. Then if you need to be faster than you can be, makes sense then if I would say like why build something yourself? If there is a lot of providers that are already offering this, often it's cheaper to get a license. I recently even got a European, I won't be too specific like a European retailer and 10 years ago they chose to build all their IT themselves, including their mail. And now it hurts a lot because they're having their own internal mail system rather than Outlook or Gmail or something like that. Sometimes you just have to use the standards and recognize what the standard will be. Same is true for LLMs right now. And then as you all mentioned, if you don't have the resources to do it, but you feel you should do it and you for example have limited headcounts but you do have budget to do things, then just go and look externally. I feel like I'm like preaching to the choir looking at all your roles. But still it's good to have this on one slide. So this is when you do collaborate externally sometimes and we would like to emphasize it as well, sometimes it does make sense to build things internally and yeah, just like two minutes on this before I hand over to Emily and like to hear about stories about open innovation so forth. But there is like a bit of the opposite. But like if there is no existing solutions is if there is either regulatory data restriction or if you have proprietary data where others don't have certain data and you feel like okay, we're not, we're not going to give this data away because for regulatory data instructions like for example if there is privacy concerns and so forth and this is big in the AI worlds and yeah, and then, and this is end by the way. So it needs to be have like no existing solutions, end something with data and it needs to create competitive advantage for you. If it's not going to create a competitive advantage for you. Yeah then just, just make sure you, you limit the cost and you have the same as your competition. One example because we, we get this question. Quite often linked to AI, so as many of you might know, is that we're building a lot of autonomous engines these days. Autonomous engines where you with AI, create, I would say, take build processes all over again, where you have an operating model where you use experts and AI agents, that by combining it and having access to LLMs and data sources and being always on that, you have a cheaper process and a better process and so forth. Very often it's actually better to wait than to build this yourselves. But sometimes it is. And these are the three. And just giving you one example of. Let's say, in our opinion, one of the more inspiring projects that we've done the last year started from like a big, big fashion retailer saying, like, yeah, okay, we, we've seen Shein in the markets. There is like a lot of controversy around Shein, but there is something that they're doing very nicely and it's inspiring to us and that is having an always on social listening engine. And I'm talking about Shein always on social listening engineering that looks at, okay, what are the designs, what are the things that people talk about, what are people wearing on their pictures and so forth. So what is trending? And then like automatically designing different apparels, then putting them on a webshop. The things that sell are actually produced and sent to consumers. The retailer that came to us was like, okay, we like this with a conversation with them. We managed to actually get the entire sustainability elements in there as well. Because apparel is. There is such a big. There is so much waste that what we together really wanted to do was making sure that you. Yeah. That you have less waste and more relevant clothing and that you only produce what needs to be produced. And in this case, you might wonder, like, why. Why is this an example of build it yourself? Well, the big why was because, first of all, it's a competitive advantage. If you have this, because this was from nine months to a week development time. And then second, they had all their retail data, all their sales data. They always kept it for themselves. So nobody had it. They didn't like, sell it to Nielsen and so forth for a moment like this. And this is why this is relevant for them. And I realize that I'm talking a lot about not open innovation, so I'll just give the words to Emily and dive into open innovation again.

Maybe just stop sharing your screen so I can share mine. Yes, thank you.

Very good.

So I think what we would. Thanks, Arna, first for setting the scene. I think it's important to go back and understand why do we do certain things and why do we not do certain things. Oftentimes we see a lot of companies just being very excited about startups and scale ups and we should be out there but with no real purpose. And I think yeah, you might get a lucky shot but oftentimes it can lead to failure and a lot of time being wasted. And so that's also what I want to talk about now. It's some of the tensions that we've observed in companies that want to collaborate with external startups and scale ups. What works, what doesn't and really our point of view on some of those tensions. So going from really all the way, very early, like how do you even collaborate? What's the length that you should be working with? So much more tactically speaking, what do you balance between scouting versus piloting preparation, all the way to measuring success. And I'll share some stories of what worked and what we've seen really drive value in such collaborations. Maybe going to the first one. So the first tension we've observed is saying well some companies, they say okay, well should we do like a quick pilot, maybe less tailor made, less co creative, but we just testing out the solution versus going for longer strategic alliances. There's nothing wrong with either and there's definitely nothing wrong with longer strategic alliances. I think what we want to say here is that we realized that before going into higher risk alliances, for example, it's always going to be very valuable to get a quick learning on whether or not this is a match between the corporate and the startup and how we did this. We developed different open innovation frameworks where within a four to six month time frame you can go from having that challenge and all the way to piloting in market. And I think this is ultra very valuable for a lot of people because you get to results very fast and you have that sense of well is it going to be a match with that company, yes or no. And that's also where we're really proud with our work that we did with Port of Rotterdam and the open innovation framework and pilots that we've launched there. As we see that we really were able to achieve a lot in a minimal amount of time. And this considering, I don't know if you know, but the maritime industry is really one of the most, let's say less fast industries just because of the complexities of it, because of there's so many different actors and stakeholders participating. So we were really proud to see, okay, we can drive change, we can do it very fast. And we've developed a methodology even with a very special setup because they are of course, facilitating change as support authorities and not owning the challenges. So also they're very, very good to see what we were able to achieve with a relatively little amount of time. A second tension that I wanted to underline was that there is a tension of, well, what is a good challenge? And let's spend some time on here because we've seen a lot of things and we've seen things work and not work, and it's quite important to understand, well, we have to have a good challenge definition because otherwise you're probably going to have failure in the sense that you won't be able to get to a pilot or even further along in the step. So I think they're just telling some stories that we've seen as well. Usually it's not a good idea to start working on such a collaboration if there's already an existing internal legacy project. Imagine you've been working on a certain technical challenges for five years internally, and you're saying, maybe now it's time to look externally, yes, that can work. But usually also it will mean that the challenge will be so sharpened on the challenges that you have identified that you'll kind of be blinded to any other possible avenues that you hadn't thought of. Second, I think it goes in hand in hand with the first one, of course, internal politics, let's face it, in corporates, they exist. And of course, creating an open innovation challenge or working with startups when you already have an internal project that is doing the same thing, not a good idea, it will probably be killed. So rather avoid that. And the third, and I think this is a super important one, as we've seen and we've written a lot of challenges in the past, be open to different approaches and really that allows you to get the best out of open innovation. Because if you go in there saying, well, I want to have a startup that does this exact platform in this exact way, you might be missing out on a different approach. And I think here as well, when we were working with Port of Rotterdam, we had the challenge of document errors. And really there we were able to test two different approaches completely either going from a centralized data depository platform or an AI powered error detection. And I think this is really, you can solve a challenge in way different ways. And that's really important when launching open innovation challenges. And moreover, I think they go hand in hand and that's why I show them together. We do understand that a lot of and you said it in the chat at the very beginning. It's about lack of resources and not enough time. And yes, that's true, but I think it's careful to say it's really important that it remains a strategic priority for the company. It's not about saying, oh, we have this like this deprioritized project. Let's just see if there's, if there's a startup that's willing to collaborate on this because that's never going to work and it's not going to continue. Also scaling internally and that also means maybe a good way to spot a good challenge is if it is a strategic priority, but it has a higher amount of uncertainty. Let's say it's something that has a longer time horizon, requires maybe some changes in your internal processes and therefore you keep on deprioritizing it because it's too complex or it's too much of a heavy lift. That might be a good idea to see if there is existing solutions externally. And yeah, obviously we've talked about this before, obviously if it is a challenge where you have to be super fast in the market, it's a great way to do that. Good. Tension number three. This is an important tension and I. Well, hopefully you'll also get to share some of your thoughts on the chat. We've seen a lot of companies spend a lot of time on scouting. There are even entire departments just working on scouting and being very opportunistic on finding opportunities and just being ad hoc looking at the market. And that's important. And there's definitely ways to do that which we've seen work really holistically, like looking very well at all the different angles from looking at your own research with partners and landing open innovation challenges, etc. What we do think is almost more important is nailing that pilot preparation. And that being said, and I'll share a bit of a grand number after we see a lot of those pilots stop before they even launch anything. And the reason why you see those highlighted in dark blue are for three precise reasons that we've come to see. The first one, I think in the end, starting an open innovation challenge, it's about people. Also it's about getting those stakeholders and the users, or in this case pilot leads on on board. Those that will actually be executing the pilot, have them onboarded. Oftentimes we see that these open innovation challenges are discussed at a higher level between leaders, but they're not onboarding the users or the pilot leads or the ones running the pilot. And that can cause an extraordinary number of mistakes or even just lack of interest and therefore the pilot just not going well. Second, is having a lot of the dependencies identified and mitigated. Sometimes we have the feeling, oh yes, it's going to be very easy, we know what the process is going to be, but we haven't done a deep understanding of the process, which is that other one of how exactly is the process going to change and who are we depending on. So that's very important because sometimes you'll get started into the pilot preparation and realize we haven't onboarded five different people or departments or external entities that need to be onboarded. So also super important when going into pilot preparation. So all in all I think a big watch out and maybe you'll disagree with me, but this is definitely our point of view is that it doesn't really drive value to just run continuous opportunistic scouting as you're just going into forever years of looking at something without a concrete landing of that opportunity. So really work on an, let's say always a strategic challenge with a clear testing and piloting phase and scope because otherwise it's never going to land in your company and the efforts that you're spending on scouting just on a generalistic phase will not bring its fruit. Good, I'll get us the fourth one and then also looking forward at the end to have your comments on this. What we've seen is in when looking at startups and communicating potential pilots and partnerships, oftentimes corporates tend to promise the world and really make it very rosy on yes, it's fantastic, etc. And instead sometimes the transparent truth is less rosy and it might be less clear. I think it's super important in our view to be very, very, very transparent and the reason why. And you see this number here. We feel very strongly about this. A lot of the pilots we see up to 80% never even start. And again because there's an unclear handover, for example, there's unclear incentives. So the incentives haven't been either communicated with that external entity for partnering. I think a big one. We never like to talk about budgets, but being very clear on how much money the startup is going to need and going to receive is extremely important. Sometimes we see that there's a vague range and then in the midst of the preparation we realize that actually the startup isn't satisfied with the returns it's getting for that collaboration. Yeah, high risk of failure. And fourth, a lot of time we expect a lot from the startups and we have to put ourselves in their shoes as well or in the scale ups. If you start planning pilots that are more than six months it's an extremely big and intensive resource for them. And so this might be also just one way where they realize and they drop off in the collaboration very early on. And finally I think let's think long term as I said it before, it's about landing the collaboration in the longer term whether or not it's about becoming a vendor or it's about actually landing a merger or even a joint venture. Longer term I think getting that scaling path extremely clear at the get go at scouting almost is going to make or break what happens down the line. So maybe I'll open up the floor on the chat. I'd be really curious to hear from you what other reasons you had for failing or failure in getting a startup or piloting just to see if there's other experiences there. All right. I don't know if you see any interesting messages I right now can't. Oh, there we go.

Yeah, I would say it's mostly, mostly the yeah, your point on scouting versus everything that is around it from stakeholder management to pilot preparation and so forth. Yeah, they're like yeah I be and.

Maybe just can I I'll just stop on the IP question. I, I we could spend a whole webinar just talking about that. We decided not to do that because of course it's a whole thing. Just going back to what Arne shared at the very beginning when we said we suggest to start with very easy collaborations that are more even just vendor tests or very early on tests with the solution as is just to see whether or not there's a match because of course as soon as you start going into co development there is that IP question but sometimes it's even just about testing out as is as much as possible without getting into those IP questions because oftentimes it's not even worth getting into IP discussions if there is no potential collaboration.

Yeah and I would say the bigger and this is also on like atit what you're mentioning the paperwork or like your legal procurement infosec and risk. These are all I would say they're often seen as showstoppers of collaborating with like less mature companies and when I say less mature it's like the type of companies that don't have a department to focus on like two months procurement kind of process and like due diligence and everything like that. Yeah, I would say it makes a lot of sense to create some kind of like Maximal or to seek maximal liability. And I would say try to try to keep it small and also work within it your legal and your risk to around that concept. So. So that you don't say like yeah, are we allowed to work with them? Because then they say no because they seek they can maximize the risk and then they say no, but that you say okay, how can we minimize the risk if we work with them? And then you have them as a sparring partner and then maybe they help you to work in some kind of sandbox environment which is then probably a link to the second thing. A sandbox environment is always nice to have where there is very, very clear boundaries from it or risk or legal where you actually have more freedom to operate. It's a bit of contradiction but like clear boundaries. Agreed with the yeah. And this again like legal and risk and so forth and procurement which might be like maximal amount of amount. Be quiet. And there's more slides.

Emily, now I'd just like to take a the last tension and I know we're over time. Thanks for staying. I think the last point and it goes more towards the end of our looking back at when you collaborate with startups, oftentimes we're trying to test a new way of doing things or we're trying to test a new approach to certain things. And that's great. And a lot of the time we try to maximize our improvement or our vision or our objective in that collaboration. But oftentimes we forget something basic which is comparability. What are we improving against? And so that's also what we've been really very diligently doing in some of our work is to make sure that we have very clear KPI definitions in what exactly we want to be improving with with that collaboration or what is it that we want to test for. And that also means something very important is just to measure today what are your metrics as today. And this on the right is an example of how we really went through the analysis of different documents so that we knew exactly what was our benchmark. And I'm saying this just because I've seen companies having a great and really successful pilot but then not being able to get the strategic buy in because there is no comparison as to what are we actually improving. And that really reduces also internal adoption. So that's something to keep in mind when also starting the pilots. And I think I'll close us off more with a forward looking statement. I think as you know we are very much pioneers at incorporating artificial intelligence and generative AI. In the way that we do our work. And that has also been the case within open innovation and we see, we have seen some snippets of how this can work and where we can really improve the way that we do open innovation programs and collaborations with startups, but we see a lot of room for growth and we'll also be really curious to hear what you've done on the bottom. You see, from the different steps that we have, I think in open innovation we've really been able to leverage a lot of core strategy in the way that we define our challenges and our rfp. So that's been leveraged very easily. There's and then using scouting platforms like Traction that are using artificial intelligence to scout for different solutions on the market is definitely a big help, although a lot of it will nonetheless be manual work. And finally, I think what's been very fun and interesting is to really speed up the whole evaluation step by doing automated analysis and integrations from one proposal to a one pager, for example, has been tremendously interesting. But as you see, we haven't cracked a nut on how generative AI is actually going to automate the whole open innovation or even scouting phase. So I'd like to open up a question to close us off on what else you've seen or tested that could improve piloting with startups and scale ups. Whether it is with artificial intelligence or without. We're very curious to see what has been tested in, in, in your experiences. So I don't see any more comments for the moment. I guess work in progress will keep you posted on what we find on this, but it's definitely been very interesting to see how, how we've, we've been able to speed up some of those piloting moments and, and how we're starting to use artificial intelligence where we can along the way. So unless there's any more comments. Arne, should we call it a wrap? 10 minutes late, but a wrap nonetheless.