Gambling on Innovation: How To Be Productively Wrong With Luis Perez-Breva Of MIT

Read Full Transcript

This is Mark Bidwell. With me today is Luis Perez-Breva from MIT’s School of Engineering where he is a lecturer and a research scientist. Luis is also the author of a great new book called Innovating: A Doer’s Manifesto. So, welcome to the show, Luis.

Thank you. Thank you for having me.

So, when we spoke earlier on you were telling me a bit about your background and you said you didn’t intend to be an academic. Now, what happened?

Well, I’m not even sure I’m an academic today but let me tell you what happened actually. So, really early on when I left Barcelona as a chemical engineer, I did a startup on cell phone location using artificial intelligence which already showed to me that whatever I learned before did not really match what I wanted to do, and that has been a constant. So, after I did that successfully, to look at cell phones worldwide in case of emergency, I decided I needed a PhD because there was a way to use this fantastic new thing, artificial intelligence, to solve real-world problems and the problems I care about are fundamentally industrial, so I did a PhD. That did not turn out to give me exactly what I wanted. What I wanted was a means to solve real-world problems and so, I thought I was going out to do another startup that would mix artificial intelligence with industrial and infrastructure companies, and in some odd way I was drawn back to MIT to try to teach students how to do precisely that, how to take new technologies and bring them to impact in a really meaningful industrial way, not just apps. So that somehow created the opportunity for me to learn what I had not learned throughout my PhD which was, how do you define a real problem that can be solved by throwing technology at it? And I did that, I continued to do AI, and what happened was they spent ten years trying to figure out how to teach and explain this in layman’s terms, so the result is the book, and now I guess I’m an academic until I decide to no longer be one and go back to the corporate world, so it’s a fluid condition, I guess.

 And in that decade, I think you worked with hundreds and hundreds of startups, enterprises, and I think you said you had a 30% success rate which is typically in the venture capital industry is more like 10%, so what are others doing wrong that you’re doing right?

We don’t try. I know it sounds odd but most programs in university and in academia, academia in general and then also in incubators and accelerators, obsess about funding and getting a company idea and then running with it, and we are a bit more modest. What we have is – we started with MIT technologies – and so what we have is a fantastic result from research that’s done, that someone has figured out that something is possible. Now, at best that’s a great paper, at better it’s evidence that we have a new superpower but frankly we have nothing to go on so it would be pretentious for us to assume there is a company there, so instead we try to figure out, how can this thing be to turned into a superpower, and then we try to shoot it down, effectively, until we are persuaded there is a real opportunity ahead that would work even if we’re wrong about how this can become a superpower. The result is that people get a plan that’s actually executable. In a reasonably short term, actually, most of the companies that emerge or licenses that emerge from the program at MIT take about a year and a half of time inside MIT to fully de-risk many, many outcomes and so when they leave they tend to succeed, mostly because they’ve tried every single possible way they could figure out it would fail before they leave MIT, and it turns out that it only takes a year, which, all of these things, by the way, were shocking to me. Perhaps the most important thought out of that experience was that really early on when we saw that this was actually working inside MIT, I got curious. Is this working because we have MIT technologies and MIT passion and talent or could this work elsewhere? And so I’m even more proud of the things we’ve built outside because at MIT, to be honest, even if they did this poorly, we have such a great pool of talent and passionate people that self-select to come here, that pretty much anything with all of them works. So, I’m more proud of everything we did and made work elsewhere because it proves to me that we are up to something.

Yeah, and I think at the heart of the book, let me just read the subtitle it’s A Doer’s Manifesto for Starting from a Hunch, Prototyping Problems, Scaling Up, and Learning to be Productively Wrong, and in that last piece about being productively wrong, I guess you wrote this book almost as a reaction to the recipes for success that tend to be out there in the innovation and the entrepreneurship marketplace around, you know, ‘Do these ten things or these twenty things and you’re going to be successful’, but as you say in the book, most recipes are exactly the opposite of learning, so learning is really at the heart of your approach, is that fair to say?

Definitely fair to say and I’ll give you an anecdote. When I started to teach this subject at MIT, it took me two weeks to dispel all the recipes and to get us really working on critical thinking and scaling up because that’s what almost every recipe forgets, and nowadays, it takes me a month, and that was a very big motivation, and the reason why it takes me a month is not that the students are any better or worse, it’s just like they have been fed so many recipes that would be awesome if they worked. So, think about it, most recipes tell you something like, ‘Find whoever wants to buy these and then sell it to them’. Sure, that sounds like a circular reference to me, but because I don’t know what it is, I don’t know who they are, and I don’t know what they want to buy, so I need to make lots of things up. So, my students come with that preconception early on and those recipes get in the way actually of us really engaging in the meaningful kind of learning we need to engage in to figure out how is it that something new, a new idea, a new hunch, whatever, is actually going to make it into a world where no one has actually seen it, and I don’t care how vague it is, it just means that no one has seen it, there are no users, there is nothing, so those recipes do not even apply. Most recipes are actually extensions of product design and we don’t have that, so I wrote it a lot in large part in response saying, ‘There’s no recipe that can get you anywhere unless you pretend you already have a product and then that’s a high-risk gambling exercise.’

So, the focus is very much on how do you actually produce the innovation in the first place, right? So, it’s the early stage before you get to the pitch, before you get to the raising the money, that early stage that, as you said, you kick the tires, it takes a long time to look at it from multiple angles and tear it apart before you’re ready to take it live. Is that where you put most of your energy?

So, at first, I thought that was the case, and then as I am writing the book, as you probably know, it informs you quite a bit about what you actually mean to say, so I realized that what I’m talking about is a set of skills that actually you carry along all the way. So, most people think that it’s that you come up with an idea, you pitch it, and then you run, right? It’s a neat process because it’s easy to understand and grasp but it just doesn’t happen that way. Funding is just a step along the way. What you need to do regardless of whether you’ve gotten funding or after you get funding is exactly the same thing, which is, at the scale at which you are, you need to make sure you anticipate what would be failure points at the next scale so that then they’re wrong, you’re basically just wrong, and then as you move on, if you happen to fail, it will not have been in a way that’s actually predictable. So, what resulted was even though what I say at the very beginning, if you start the book in chapter one it looks like I’m talking about the messy process before the idea and before the pitch. If you start the book from chapter six and then move on, what you realize is that it actually applies and it’s the exact same skillset I’m applying at different scales all along, and the words change a bit but the mechanics are the same. So, if you’re starting from scratch what you build up is a resilience that comes with you all along, and the same principles and mechanisms you used at that messy stage actually apply the same at the next stage.

Yeah, now obviously at MIT a lot of the folks coming through are entrepreneurs and they’re looking to build their startup, but you also work with, I mean, your message resonates a lot with corporates and you work with corporates, our paths didn’t cross but I know you worked with my previous company, Syngenta, in the past. How relevant is this way of thinking about learning and innovating in a non-linear way, leaving behind all of this language which gets, I think you use the word ‘contaminates’ our thinking, how relevant is that in large long-product-life-cycle industrials?

I think it’s even more relevant than for startups. We live in a moment where being an entrepreneur is fashionable and that will pass as most fashions, but what’s critical is to understand what it takes to actually produce innovation and it’s different from what everybody actually thinks. At the very beginning there’s nothing new, and if you stop at that thought and you look at what a corporation has achieved, a corporation is an incredibly efficient beast that manages to put together very disparate areas for the purpose of serving either a product, a service, an industrially produced product and so forth. Now, take this company a hundred years into the future just for the exercise; you are very unlikely to rebuild the entire company from scratch, rather you’re probably going to have figured out a way in which part of what you did in the past helps you going forward in the future. So, innovating the skillset is about that, it’s about how you figure out or how do you learn how to slowly or incrementally repurpose what you already do today to support your existing product so that in the future that same activity with small changes supports an entirely different product. So, at the beginning, nothing is new, but as a corporation what you have is already way more efficient than what a startup has, so, if at all, it’s even more relevant to a corporation than it is to an entrepreneur. That may be perceived as having freedom but that freedom comes with an insane amount of inefficiency. Startups are inefficient by their very definition. They may be fast, agile, but it’s mostly because they need to be because they don’t have anything that works, whereas in a corporation you have a lot of things that work, and you have a small resistance to understand that whatever you may need your pieces or your parts or your business units to do in the future, is just only slightly different from what it does today. If you take that approach, then learning is possible. It’s about-

Sorry, keep going.

No, go ahead.

I was going to say, another area which I just wanted to touch on here, is this idea of the importance of being multidisciplinary and I guess this is about exposing yourself to multiple perspectives versus innovation just occurring in one function, in one discipline, in research and development, for example. Can you tell a little bit more about why multidisciplinary thinking is so important?

Sure. So, first of all, these disciplines that we have should not be seen as constraints, right? They are, fundamentally, assets, and that’s true for knowledge in general, so whatever you acquired whenever you went to school, whatever you were trained on, it is typically taught as if it was a constraint but really it should be seen as an asset. Now, if it’s an asset and you actually accept that going forward you don’t know what’s ahead because it’s fundamentally an innovation, then it would be preposterous to assume that you have all the knowledge you need right now, so multidisciplinary thinking, there’s a lot of praise for it but my statement for it is fairly plain. It’s very unlikely that the discipline you have right now or the one silo you live in today is going to be the same silo ten to twenty years ahead, and so the default should be multidisciplinary communication because for that thing that’s new, we need to develop a new language and the multiple vantage points force you to develop that new language, meaning if you’re talking to a lawyer, a conversation between a lawyer and an engineer, which by the way happens all the time in my classes, requires a different language and so the problem you’re solving gives you that language and you guys each bring your own perspective and backgrounds to it as an asset, so multidisciplinary is just basically, I don’t how you could possibly do it otherwise, but you can bring this up earlier if you focus on the problems. So, multidisciplinary is praised mostly because we don’t see it often, but it should be the basic default, otherwise you’re saying the opposite is that you can have a company built out of only engineers or built out of only lawyers, it makes no sense, and then you have to solve for that in a different way bringing very complicated words into the picture like culture, and team dynamics and so on that only makes the problem more complicated. Well, just start out by realizing that you only have half of the language you need to solve that new problem and let’s make it multidisciplinary from the get-go and the idea of culture becomes actually easier to palate because you’ve had to wrestle with it from the get-go.

And so, coming back to the large company with a history, with products, with customers, maybe under attack from an invasive species, a new entrant into its market, what are you seeing these companies that are tackling these issues really well from an innovative point of view, what are they actually doing differently from the vast majority, if you like?

So, let me tell you what I believe the vast majority is doing because it will be easier to see what the exceptions are. So, the vast majority these days is trying to identify great ideas to pursue. So, they will have their own form of idea box, they will have their own form of internal competition, demo day, patent award, whatever you want to call it, and that’s great but then the hope is that somehow an idea will reveal itself as being awesome to pursue. That is gambling effectively. I mean, we dress it up, we add in statistics, but all we have at the end of the day is a long list of ideas that are barely thought through and the statistics about which ideas seem to please most people and whatever, so we’re effectively thinking about new ideas as if they were recognizable and then a matter of how many likes you get, and everybody will have a different process to perceive this. This makes no sense to me, it literally makes no sense. Actually, math is against the idea because all you’re going to get is to the average idea, you have to believe it. Now the companies that I believe do these incredibly well are companies that take a step back and say, ‘It’s not about coming up with the next product idea, because we’re not great at predicting products or the products we predict are tend to be rather incremental and really hard to build,’ but rather they take a step back and say, ‘You know what? We have a new problem to solve and we have parts’. Now those parts may be the brand, maybe the legal system, maybe the contracting office, maybe the engineering folks, maybe the research folks, management, and those are the parts you have to work with. Now, what they do is they conceive of a way in which they can demonstrate to themselves that a given problem is worth solving further so that they change the decision making towards, instead of being about choosing great ideas, it’s about figuring out whether the demonstration that the money would buy is worth obtaining. Now, that’s a worthwhile inquiry because now every idea is looked by its own merit as opposed to how it looks when compared to a bunch of other ideas, and so companies that do that end up starting pretty much the same way which is they produce something that looks like an embryo of an organization. Take a very common example that everybody has heard of is how IBM started their PC business. I’m using that not because I don’t have others but because that’s one I’m sure everybody has read or heard about, it’s been in the literature for thirty years. So, if you look at what they did, they had a small unit that was tasked with exploring PC’s and all the might of every single other business unit within IBM, they could tap into, the lawyers, that programming, that had been developed for something else. In that case, it was the mainframes. Now, instead of thinking, ‘Oh, we’re competing with each other’, and so on and so forth, they gave it a try, a try at scale. It’s not lean experimentation because it was costly. It’s mostly like, ‘Let’s give it a serious try, bringing in the weight of what we know how to do, and allowing freedom of operation to some degree to allow for us to explore these meaningfully.’ Now, it could have backfired, right? Maybe IBM could not have actually pulled out a PC but that would have been a small investment on the part of the company and then they would know that that doesn’t work but they would also know, how could their existing units have served another unit? They would have learned a bit along the process. It did turn out to work and so the next up was bringing that to scale, but you see we’re talking about bringing an organization to scale, it’s not a single product, and that’s the key difference from most recipes and most companies. What most companies are doing today, they’re focusing on a product and mostly doing product management on a made-up product instead of actually trying to figure out how to build a new organization with what they have there. That’s the fundamental difference, and of course, you cannot plan for this ahead of time so that’s why I emphasise that the PC business could have not worked out, but the only way for you to know is to understand the decisions that are to be made as you bring this to scale and to realize whether this is a good use of your resources, and that’s, by the way, that’s what every management manager is trying to do.

Yeah, capital allocation.


That’s an example as you say from a number of years ago. Are there any companies that you’re seeing you using this approach in contemporary times that gets you positive about the future of corporate America, for instance? You know, the rump of the industry which is obviously under attack from these new entrants across many industries?

Several. Several are following – first of all, I don’t claim that these people are following the approach I created because these people obviously did it before I published the book, so let’s be clear about it, I just see trends, patterns, and principles but I don’t claim that they followed my recipe, which is a common mistake for most recipe books.


So, there are many companies that are doing this continuously, large scale and small scale. For example, IBM has done it again with Watson and using that infrastructure for business and if you look at the mechanics it looks a lot like what they did for PCs. Not only that, along the path they actually sold their notebooks to Lenovo which means that they make this a matter of process to build up an organization that then eventually are no longer part of their core. 3M has been doing this forever. In the case of 3M, they do this more through acquisition, so it’s not so much they actually have their ideas in-house, they also consider requisition as a means to actually introduce these ideas and reuse parts they already have for the purpose of advancing. The mother companies, to give a hundred examples, like Google and Facebook, they claim to do something similar to this, but their process typically starts with the idea of gelling the product really quickly. So, they may start with a problem then they gel the product they want to do and then they hope that product will be a mass-market product. So, you actually see more of that – you know, it conceals a bit of gambling because you need to be right about the product from the get-go. It doesn’t build any robustness so it’s kind of high-risk.

And the distinction there is versus what you’re talking about just to be clear is, you start with this hunch, you start with not even an idea, if you like, but a gut feeling you have something and then you iterate and iterate and reiterate until you feel confident that you’ve got a problem that you can go after. Is that the distinction?

It’s close. So, on one part, yes, you started with a hunch, you describe the problem you believe that started with a hunch and in the book, I explain how to do that very quickly, and then your job is to realize that if this is a real problem then there is a role to be played by an organization at the other end that actually solves this sustainably, continuously, and so your job is to bring this problem upscale one step at a time. So, it’s not only iteration, it’s actually every iteration needs to bring you closer to a demonstration at the next scale, so the key word for me, the most important one is scale. Iteration is just what you do whenever the problem doesn’t have a closed formula, right?

And then as distinct from, you were saying, how Google and Facebook do it, or Amazon do it, which is a different process, it’s more accelerated around the product, if you like?

So, actually, Amazon does it differently. Amazon is awesome.

So, it’s Facebook and Google that do it the way you just described. What does Amazon do then?

So, Amazon seems to follow, and I am glad you brought them up because I had forgotten about them. So, I think they are just admirable in everything they do. They’re just unlike any other company, unlike any of the companies, any of the companies that survived the in that they build stuff for themselves. So, think about the Elastic Cloud of Amazon. Amazon needed that for themselves because they needed to serve computation, they needed to serve fast search, they needed to serve a lot of different products and it needed to be seamless for you. Now, one day they realized that that, and that’s as far as I know the way it happened, someone at Amazon said that this could also be a product, so they had developed this enormous expertise that could conceivably be repackaged and made as a product. So, then they had a question as to how to do it but that’s actually a very easy starting point. They have something that works that’s a good prototype of what they have, they suspect that a lot of the world is going to need cloud computing, it’s not clear what kind of cloud computing is going to be needed but it’s going to need big resources and the question is, can this be brought to be a separate organization within the company? So, they did it and as far as I can tell that’s what they continuously do. Every time they build something for their own needs, they pause, or maybe someone inside brings up the idea, ‘How can this become a powerful means for us to proceed forward?’ So, it’s not the Kindle which would be the product, it’s more like the idea that they can build an infrastructure that can be supported or needs to be supported by an organization, and that’s very different from the idea of saying Google when they came out with Google Glasses. I’m not saying it’s not a good idea-

It’s very different process.

It’s a very different process because they decided on the glasses really early and even though they may have thought of a problem early on, they ditched the problem and went for the product.

Interesting. So, again, a lot of our audience are sitting in large organizations, can we just talk a little bit about measuring innovation because this is one of the things – actually, I was at a meeting last week with a bunch of procurement professionals and they were asking me how to measure the innovation that they lead in their procurement functions. You have KPI’s, you have ROI’s, but I think you’ve got a slightly different angle on this around uncertainty reduction, I think – can you say a little bit more about how do you think about measuring innovation?

Yes. So, the first thing to realize is that if you are using a KPI, KPI’s only work for a linear problem. Actually, that’s something we know in the run from physics and engineering that if you have a linear problem then you attach to it an indicator, if your problem is not linear, your indicator will not work. So, the first step in coming up with an indicator is linearizing a problem, so that’s why those recipes are so great because they enable KPIs, even though the KPI isn’t conceived for what you want to achieve but at least you can have a linear indicator. So, the first thing you need to accept is that because you try to produce something new and it’s a highly non-linear process that requires a lot of exploration, KPIs is not the right way, or KPIs on the outcome or the desired outcome are not the right way. So, instead what I-

But KPIs on the process, perhaps, or not?

Yeah, to some degree on the learning, actually, that’s where I would put it, and you’ll see why, so,  think about it the following way; you have a problem, you have a group of people working on that particular problem and trying to figure out what needs to be brought together to demonstrate this problem, right? So, the first point for those people is to actually figure out what kind of demonstration they are seeking to achieve with an extra set of resources. Now, that should not be a demonstration about a single product but rather about how robust the space of opportunity ahead of them is, so in so many ways, the more ways they can show to you they could possibly solve the problem, the better you are, provided they are sufficiently distinct and that you believe the demonstration that comes afterwards. So, one good indicator is actually, ‘Am I getting enough ways to solve the problem from these people? Are these people going to survive what happens to us every time evidence hits us in the face that we were wrong mostly?’ So, if they are, then you have a good measure these properties may be worth exploring. The other thing is just traditional analysis which is cost-benefit analysis, ‘Is the demonstration going to serve me well in terms of, if this turns out to be true, if it turns out to be an opportunity, is this something I care to know about and most importantly, is the money asked for too much or too little for the demonstration? Will they want to know something else when you bring it up into a learning component?’ And the last piece of it is to realize, are you going to learn something about your particular company? For instance, say, at the end of the day, all you buy is knowledge that the opportunity isn’t there, which is good knowledge to have, so will you have learned at the very least how you could possibly see every single one of your units in a different light so that when the next innovation product comes in, you know more about how flexible your organization is because that’s really important. Now, if you want learning principles or indicators and learning that you can do as a corporation, most people will be hesitant to accept this because they don’t tell you whether they don’t have had success, but success is survival and growth of your company. It’s not success about a specific product idea which is where most KPIs get it wrong. The more flexible you understand your company to be, the easier it is to incorporate that learning or those lessons into any other part of the company, provided you do this continually, and the more you learn about how flexible your company is, the easier it is to bring in new product ideas or new organization ideas and test them at scale, because you know so much more about how you can actually serve any new organization that emerges from within. So, it’s KPIs and the skills and the learning that the company goes through and on building a sustained innovating capacity more so than a specific product which is where most type KPIs are actually geared for.

Yeah. So, I guess you live in a space where you’ve got one foot in the startup world, you see a huge amount of startups coming through MIT but also, you work with large organizations. How bullish are you on corporate America’s ability, not just America but the corporate world, the largest average company’s ability to actually manage themselves through these big VUCA environments that we’re facing today?

I am of two heads there. On the one hand, many of the processes I’ve seen implemented by companies, to me, sound like an enormous waste of resources. All of these methods bank it all on having a good idea and demo days, and these things to me sound wasteful because they’re not really thinking about building any organizations. So, someone will have to think about that later on when so many things have been decided, and then that turns out to be inefficient. On the other hand, I’m actually quite hopeful because another group of corporations has tried to actually buy these methods and realized that they don’t really work for them and they wrestle with them because they seem to be too much oriented towards consumer products, they seem to be very narrow in scope, and so I’m very hopeful that I hear more and more corporations telling me, ‘You know, these launch pads’ and whatever names they receive ‘don’t really work for us because we have important infrastructure there, we have long cycles, and this fast failure mode and celebration doesn’t really work for us,’ and I’m very happy because it turns out it doesn’t work for startups either, and so let me just give you a couple of numbers about startups that few people realize. So, there is this culture of failing fast and that’s what most corporations think. Corporations feel bad about that idea because failing is not a nice word, and I agree it’s not a nice word, you should succeed. Most companies that succeed startups take a good four years to get to a point where they have something that an external person would agree adds value and let me tell you with an example. Google AdWords is the real first demonstration of sustainable value for Google. That wasn’t there in 1998 when they started, or 1996 when the first version or prototype of Google went live inside Stanford. That appeared, if my memory serves me well, around 2003, so that’s three to four years after funding. The same for Genentech. Genentech took four years from funding to create the first clinical trial which is the first moment in which you’re trying to prove the value of what you’re actually producing. So, it takes four years, so I don’t know what ‘fast’ means and, you know, if you’re going to invest four years into something you better ensure that failing is not an option or at least failing in ways that you could have anticipated is not an option, so the mentality of ‘fail fast’ has not resulted in great startups. The mentality of ‘don’t fail predictably’ seems to work out, so I’m hopeful because most corporations are seeing this, that this mentality doesn’t resonate well with what they do and although they try, it ends up being a separate entity inside their corporations rather than being something fully internalized in the corporation as a means to grow or change.

Yeah, yeah. Well, I guess, quite a lot of, as you say, the language that comes from the recipes are oriented towards a misunderstanding, I guess, of how this happens in the startup community which is completely different from, as you said earlier on, in the corporate world where you have all these resources, you have these models, you have these customers, you have this capability which can play out very, very differently if you manage them effectively.

Absolutely, but I’ll give you another data point. The Wall Street Journal published in April a report on corporate America and entrepreneurship and they were trying to answer the question given the emphasis on entrepreneurship over the last decade whether we had more startups or not now and whether they had been more successful, and to their surprise they discovered that for the first time in recorded history, more Americans work for large corporations, and there’s been twenty-five-hundred employees more than for small companies understood to be a hundred or less, which means that even though we have more entrepreneurs than ever, those people are not doing the things we wanted entrepreneurship to do, which is to create fantastic innovations and enormous amounts of work. In the report they go at length to show that the growth is really coming from either, acquihires, which is one form entrepreneurship has taken but then it’s mostly up to the large corporation to kind of bring into fruition, or from large corporations themselves. So, in so many ways we have evidence that seems to suggest that even though we have more entrepreneurs out there, the message given to entrepreneurs is not really taking over corporate America, rather it’s making it more prevalent, and by the way, in the United States up until 2004, it had always been the case that more people worked for small companies than for large companies, so it’s a reversal.

Interesting, yeah. But following the media, you’d be forgiven for thinking that it was exactly the opposite.

Yeah, exactly and so, that’s part of what I want to accomplish. I want to bring a bit of voice of reason to this. You know what? It’s great that we have entrepreneurs, it’s absolutely fantastic that it’s become an accepted professional career and I’m fully supportive of that, but innovation is not a prerogative of entrepreneurs, it that happens pretty much everywhere and that’s what we actually need. We need more innovation otherwise the economy doesn’t grow and that’s been known by economists for a long, long time. Economies grow because of innovation not because of more entrepreneurs.

And you make this distinction, just because you’re an entrepreneur doesn’t mean you’re an innovator or vice versa? They’re two different animals.

Two different animals that on occasion overlap and when they overlap, we love it, and we love it so much that we remember their names; Bezos, Musk, Jobs, Ford, and so on so forth. But every single one of these, they’re building a huge company. They did not build just a small company for an app. We don’t remember the app builders, we remember the people that managed to go from zero to a large corporation.

Yeah, yeah. Super. Fascinating stuff. Thoroughly recommend the book. Listen, just before we wrap up where can people get in touch with you?

So, the easiest way to get ahold of me is through my LinkedIn profile and send me a message. I think I have it set up so that people can actually send me messages from wherever they are.

OK, and we’ll put the link in the show notes. Twitter as well? Any other social media?

Yes, I do also some Twitter. I’m not particularly good at it but I’m getting better.


So, my Twitter handle is @lpbreva, and then they can get in touch with me via email. My email is public actually but they can get it through the LinkedIn profile as well.

OK, good. So, we’ll put those in. Now, before we end, I sent three questions across to you last week. First question – what have you changed your mind about recently, and I’m really interested in the answer to this given how important learning is in all your work?

So, what have I changed my life recently about? So, you started asking me whether I had become an academic and for a while I thought I had, but the kinds of ambitions I have are bigger than the academia we have, so I’m not sure the next step for me is inside academia, to continue further this work from an economic vantage point or to create a whole new other thing where people can actually engage in innovating meaningfully in the way I talk about in the book. Of course, I do that already in academia and I’ve done it worldwide with over three to four thousand people before I even wrote the book but now, these days I’m thinking carefully, ‘Am I still an academic or not?’ So, I’m still changing my mind about that pretty much every day.

Well, we’ll watch this space for breaking news. Next question – have you got a personal work or practice you can share with our listeners that’s helped make you more effective?

I take all my conversations walking, actually all of them, typically in the street, even if it’s snowing. So, every time I have a phone call, obviously not today, today we’re sitting over Skype but every time I have a phone call I’m actually walking in the street and it helps me think a lot, the accompanying exercise as well, but most important it actually keeps me in touch with the real world otherwise I’m just stuck in an office, so, whenever possible, every single phone call I’m taking in the street, taking a stroll not far away from the office, but just is taking a stroll, and I think it makes me more effective because I factor in what I’m talking and thinking and thinking is something that’s really hard to do when you are sitting in your office, going through the motions of your email, and so many other things that hit us every single second.

Fascinating. Great, great, and third question – what’s been your most significant failure or low, what have you learned from it and how have you applied that learning?

So, when you sent that question, I started to think about what’s been my most significant failure, and you know, I don’t know, but the reason why I don’t know is not because I haven’t gotten a gazillion noes. I think I get about ten noes every week about stuff, crazy stuff I want to do and then someone tells me no. It’s because I don’t understand the word failure because failure to me is fatal, it’s final, and so what I get is an insane number of noes, many, you wouldn’t imagine. For example, I was told no the first time I applied to MIT to become a student, but I didn’t take that ‘no’ as more than their opinion, so I figured out another way. I would say this doesn’t make me different from any other person, it’s just I’ve never understood the word failure per se because you’re not just going to stop, right? So, I’ve been told no about MIT, I’ve been told no about startups, I’ve been told no about new programs I’ve wanted to create then they’ve come back to me to actually create those programs so, every supposed now has taken me in another direction. That’s the way I see it. So, the more I think about significant failures, I can’t come up with a significant one that I couldn’t tell you something I learned enormously from it and at which point it mostly sounds to me like I was wrong about something, more than it being a failure. I never truly saw that massive moment of change.

So, it was a nice reframe?

But I don’t want to be cured, right, either, as soon as I work hard – and you know, I could claim many things in hindsight, but since the next day I was doing something else already, all it was was a no. Don’t get me wrong, each no hurt enormously and they still hurt. The last one I got two days ago hurt a lot but I’m just already thinking about how we’re going to do that differently. So, I don’t perceive any fantastic failure in any spectacular way which is not to say that bad things haven’t happened.

Yeah, but it’s how you look at it, right? It’s how you choose to respond versus the meaning you attach to the event or the step in the process?

Exactly, and you said very neatly, the meaning you attach to it only comes in hindsight, but to which extent it wasn’t even a failure at the time it actually happened if you were able to change the meaning of it in hindsight.

Yes, and that’s a theme we haven’t touched on in the book but you talk about if you look at innovation backwards you see one story, you look at innovation forwards you see a very, very different story and I think what you’ve done is this book is all about looking at innovation forwards, it’s all about going through the reality of it versus the rose-tinted view that we are spoon-fed in the media about innovation looking at it backwards?

Absolutely. I’ll give you a quick example so that you see what failures mean. Ten to eight years ago I started a startup on fashion with artificial intelligence with two colleagues. We started working on it, we got IP secured, we were securing funding, it was all working all the way there, and then one day I woke up and I said, ‘Oh my God, I care nothing about fashion. Why did I even engage in this? Why did I do this?’ and so I talked to my partners, I said, ‘You know what? I really don’t care. I like you guys, but I just don’t care’ So, I moved out and I never looked back, and I told them that everything we had done together was theirs, I didn’t want any claim on it because I really didn’t care so I wished them success and whatever I could do to help as a friend but I just didn’t want to do it. So, they continued on their own. Now, is that a failure? I don’t know, I think I was successful enough to figure out I didn’t care before I spent four years in it, right?


I was most likely wrong. At the time though I was out of a project though, I had to come up with another project to do.

Right, nice story. Luis, it’s been great to get you on the show, sorry it’s taken so long. I know it’s been hard work to schedule this but really appreciate your input, your work and I’m sure our listeners will as well, and thanks very much for your time today.

Thank you. Thank you for having me



In this episode, we are joined by Luis Perez-Breva, a lecturer and research scientist at MIT’s School of Engineering and the Director of MIT’s Innovation Teams Program.

Luis has extensive experience in both innovation practice – via his involvement in multiple startups – and innovation research – through his academic work.  

We are talking about his first book, Innovating: A Doer’s Manifesto for Starting from a Hunch, Prototyping Problems, Scaling Up, and Learning to Be Productively Wrong.

What Was Covered

  • Why Luis sees following “innovation recipes” is inherently wasteful and essentially high stakes gambling
  • How the best innovators both prepare for scale at each stage and excel at applying their “parts” to identified problems
  • How a corporation’s existing products and services give it an innovation advantage over startups

Key Takeaways and Learnings

  • Luis’s tried and tested method, anticipating failure at each ‘scale’, which can help innovators to prepare and solve as many foreseeable faults as possible – what he calls being “productively wrong” as a way to avoid “failing predictably”
  • How to use linear processes to improve the non-linear process of building innovation
  • Innovating the skillset; how companies learn and re-purpose what they do today to provide entirely different products in the future

Links and Resources Mentioned in This Podcast

Mark Bidwell

Mark Bidwell

Join Thousands of Members for our FREE Newsletter

Share this post

Share on linkedin
Share on facebook
Share on twitter
Share on email
Share on print


Recent Articles

Follow Us

Latest Episode

Stay Updated

Share with a Friend

Share on linkedin
Share on facebook
Share on twitter
Share on email
Share on whatsapp
Share on telegram