IGNITE 2017 CONFERENCE

Cracking the Innovation Genome

Jim Hornthal, Launchpad


Jim Hornthal
Co-Founder & Executive Chairman
View Slides
Jim is the co-founder and Executive Chairman of Launchpad. He is a serial entrepreneur, angel investor, and lean innovation educator at number of leading business schools. His private investment fund, Hornthal Investment Partners, has made investments in over 50 early stage companies, including Lending Club, Lyft, Change.org, Hightower Advisors, Krave Jerky, KIND Snacks and Hipmunk.

Video Transcript:


Amy:
I'd like to introduce, today, as part of that, Jim Hornthal, who's the co-founder and executive chairman of Launchpad Central. Launchpad is a partner of ours, and also Jim is a professor at Stanford and Berkeley. He teaches the next generation of business leaders about innovation and entrepreneurship. He's a serial entrepreneur, an angel investor. If you've got any good ideas, let us know. And his private investment fund, Hornthal Investment Partners, has made investments in over 50 early-stage companies, including Lyft. Ideas, Jim? We're open. Crave jerky and kind snacks. So if we could all give Jim a warm welcome. Thank you.

Jim:
Good morning, everybody. I realize I'm standing between you and lunch, so hopefully this will be brief and to the point.

I liked the way Scott started today, taking us to the past to talk about the future. And what better day to talk about the future than today? You all know that this truly is Star Wars Day, the Fourth of May, and it's true. It's May the Fourth be with you. That's not a joke. This is Star Wars Day. And Scott took us back to the Gold Rush. I want to take you back to 19 years ago, to an important event in my life. This would be season two, episode 17, of South Park. And for those of you who are intimately familiar with South Park, you may know that the issues of the day are often reflected in our common media, be they cartoons, comic books. But there's an important issue in this particular episode, and this is an episode where something strange is going on in South Park. Everyone's underpants are disappearing. And the boys of South Park set out to find out what happened to the missing underpants. So if we could run that video, we're going take you to season two, episode 17 of South Park.

<Video>

Jim:
We were actually going name the company Phase Two, but no one would get it. So what we've got is an interesting situation today, where being strong at execution is critical to the success of every business. Everyone who's here is here in testimony to the fact that the businesses that you're involved with can execute, and execute well. But the question about how to go from being ambidextrous and being able to both execute and discover is a really difficult challenge.

So phase one: idea generation. And Spigot is the world leader in helping companies become very efficient and effective in that. Phase three is something you're already good at: execution. But phase two, in this case, is the abyss. It's commercialization. And to go from phase one to phase two ... I had a great teacher who taught me "Never confuse a clear view with a short distance or a straight line." You'll be on the 30th floor of a building and think you can just reach out across the street. I don't advise doing that.

So the question is, how do you navigate this abyss? How do you take a company and its strengths and execution and help them not just search, because search has an implied answer, but help them discover? There's a great article in the recent issue of the Stanford Graduate School of Business Magazine that great companies are often accidentally discovered, that it's the process of going from an idea to something great. So the question is really understanding the process involved, and let me rewind the tape and tell you some of the things we've learned about innovation.

Our company began five years ago. Steve Blank, who is the godfather of the Lean Startup Movement, and I were doing some work with the National Science Foundation. They wanted to help the 11,000 men and women from all around the world who were doing academic research in this country do better at going from innovation to commercialization. The reason that we spent $7 billion a year for the last 40 years as a bipartisan support in this country is we firmly believe that innovation leads to job creation and industry formation.

Well, there's a metric of innovation. There were 550,000 patents last year granted by the US Patent and Trademark Office. You can check the box. But everyone wants to be the folks who gave Sergey Brin and Larry Page their NSF grant to Stanford and Qualcomm, really great companies, and were kind of bad at that. Well the problem of being good is here's what we've learned with 15,000 teams, is that no successful business winds up where it starts. Google was going to be a box inside the enterprise to help you search for proprietary data. Amazon was only going to sell books, maybe CDs, because they look like books. And the idea is how do you go from one to the other and accelerate the time to truth and minimize the pain? We know that you make lots of mistakes and changes along the way, and embracing failure and the culture of failure is actually embracing the opportunity to be successful.

The thing is, also, there's no structured way, historically, to keep track of all the learnings. I am a data hoarder. I confess. 12-step program: first admit your problem. I am a data hoarder. Every time a white board is erased, I cry. Every time a Post-It note is thrown away, I wonder who could use that. So I have a garage full of Post-It notes and a bunch of un-erased white boards. The question is, how do you make this accessible? How do you democratize access to all the failures we've made along the way to being successful?

The other thing I like about Scott is not only does he use history as a parley, we both have that same famous Edison quote, but you'll notice that mine isn't the picture of Thomas Edison. Mine is a page from his notebook in 1876. Edison took copious, copious notes of every experiment. He wasn't sure if it would succeed. He wasn't sure if it would fail. But unless he could capture at the artifacts of the experiment, he'd never know.

Anyone ever use Pandora? Pandora is a beautiful example of a recommendation engine. If you do one thumb down in Pandora, you know what it knows? Absolutely nothing. But a series of thumbs down, it looks for commonality and then began to customize and evolve the music genome. So if we can take music, and Pandora's adjusted over a million songs, carved into 400 different elements, each rated and scored by experts, you can find patterns in the music.

It turns out that in the laboratory of today, they don't need copious notes. The electronic lab notebook of today has sensors built into everything. If you go into Lawrence Livermore Labs, they are measuring things that seem to be irrelevant for the experiment: temperature, humidity, oxygen levels, because you're just kinda never sure what might correlate with a future question. So imagine if you're ingesting all of this data in your electronic lab notebook inside the lab, giving you more data than you ever thought you could imagine. Why aren't we doing it with the lab notebook outside the building?

Anyone ever take a course in biology or chemistry or physics? You don't have to tell me how you did, but you took the course. Scientific method, I have a hypothesis. I think there's lead in that water. I'm going to run an experiment. I have a lead test kit. I'm going dip it in. Oh my god, it turned blue. Hold on, let's do a quantitative analysis. How much lead is in the water? One part per trillion, drink the water. So the idea is I have a hypothesis, I run experiments, I gather evidence, and it's a recursive process. If we do that in the building, in the lab, with so many great sensors and data acquisition, why are we failing to do that outside the building where customers, products, and markets replace bunsen burners, beakers, and test tubes? Kind of makes sense, right?

If we can repurpose and animate all of this intellectual exhaust and make it available, transparent, and collaborative as a resource for all of our innovators within our company, we have an opportunity to leverage institutional knowledge in a way that's never been done before. There are externalities beyond our company, beyond the walls. Regulatory environment, competitive environment, technological advancement. You certainly don't want to build a solution that's not going to be reasonably durable. We did some work with one of the three-letter acronym agencies in Washington, and it's really hard for them to get new technology approved, so their tagline is "Yesterday's technology tomorrow." I think they're still running Doss. At least, I think that's true.

But the question, and Watson's kind of led the way in cognitive theory, is it's not necessarily predictive, but it's suggestive. We've seen this movie before in this way. If that's okay with you, enjoy the movie. But the idea is to develop and underscore some new metrics. If you believe, and I'll speak personally, a lot of the better decisions I've made in my investments have something to do with pattern recognition. What comes before pattern recognition is pattern acquisition. You have to collect the dots before you can connect the dots.

So the question would be, if you believe it that way, there's some systematic thinking that goes beyond the normal frame of data analysis. A fancy word, ontology. Let's organize the data in some rational way. If we were in the map making business, we might decide that we're going to have countries, states, cities. As reasonable as any. Then we have to come up with taxonomy. We're going to name this data. This will be forever called San Francisco, California. If you call it Paris, France, good for you; no one will find it. So you start having common phrases.

The next thing you want to do is come up with some level of algorithmic intelligence. Some people like to call it AI. That sounds scary; we'll just call them algorithms. In order to synthesize and analyze the data, you're looking for trends. You're looking for acceleration. You're not just looking for velocity. If I have three cars on the freeway doing 60 miles an hour, that's interesting. This one went from zero to 60, this one went from 100 to 60, this one's doing consistently 60. If I ask you which car's going to be ahead in 20 yards, you'll probably go the zero to 60. So acceleration is a much more important measurement than just simple velocity.

Visualization. If I want to synthesize and analyze the data, I damn well sure better be able to visualize it, and Spigit's done some fabulous visualizations when you look inside of your idea portfolio. As a person responsible for finding, filtering, and funding the most promising projects, those visualization tools become critical. So I'm not going to teach you anything about the wisdom of the crowds; that's something that Spigot has brilliantly tapped into. You want the diverse opinions, independent, e-centralized navigation. What we've seen in successful teams, though, is the more diversity, resiliency, and tenacity are what characterize great teams. If we all come from the same background, the same degree, the same school, the same knowledge, we're going to see the same problem the same way. But you want to have age diversity, gender diversity, technical diversity, poets and engineers together, because they bring something unique to the problem. Their pattern-acquisition skills have a whole different portfolio.

If you've ever played the game Boggle, where you shake up 16 letters in a cube and you want to find words, if you rotate the cube 45 degrees, you see new words. They were there, but your perspective has changed.

So the question for innovation and to discover its scale is a lot of times we'll start, importantly, inside the building. We want to tap into the wisdom of our employee base, their richness, their talents, their skills. Using market data and reports, where Watson was showing us early this morning, is really valuable. Outside the building, I frankly don't care what you want to build. I care what your customers want to buy.

We often get stuck with this problem where you take a microscope on what's feasible? So we build it. What we often lack is the telescope of what's desirable. Just because it's feasible and desirable, you want to make sure it's got adaptable. What's coming around the bend? What's the next thing? And by the way, if you're lucky enough to find things that are both feasible, desirable, and adaptable, you may well have something that's valuable.

For those who are familiar with business model canvases and other visualization tools for business models, the left side of the canvas, where you have key resources, key activities, partners, that's feasibility. 90% of projects start because we can do it, not that we should do it. Right side of the canvas, value propositions, customer segments, partners, channels in customer relationships, that's what they want out there. Wouldn't you rather build what they want?

The work we've done with the National Science Foundation and the Nationals of Health is amazing because the engineers and scientists that touch this approach go back to the lab and start making new things because they want their molecule to be used. It's one thing to have a patent; it's another thing to change lives. So the way you make a dent in the universe is you keep your eyes outside all the time. Within the companies you work in, it's amazing. You are blessed with amazing amounts of what we call innovation capital. Innovation capital is the talent pool you've been fortunate to recruit and retain. It's the ideas that they individually and collectively can come up with, enabled by powerful technologies to make it easier to get those ideas into a pipe line or a funnel of some sort.

There needs to be a culture that is not only tolerant, but embraces failure. The average team, 15,000 teams later, 500,000 interviews, 51,000 documented pivots, we're wrong 7.6 times for every time we're right. 7.6 times. The average team has about 40 plus hypotheses they're testing, and what they'll find out with is they will make at least 28 changes from the time that great idea becomes maybe a really great business.

So if you are of the faint of heart and don't like to be told no, A) Don't go into sales. And B) Think about your innovation career path, because it's going be fraught with bruises, scrapes, and changes. But those changes create the data. And if you capture the data, again, data hoarders, we're going to meet after lunch, you've got this intellectual exhaust that feeds future learning. It feeds understanding. It is the source of insight that can lead to breakthrough.

Everyone likes to say, "We're going do a moon shot." Joe Biden's doing a moon shot on cancer. Let me take you back to 1960, the very first moon shot.

Imagine if you are an astronaut. It is 1960, and John Kennedy says, "We're going to the moon by the end of the decade." How many of you are really happy? You made the right career choice. All that training, all those low-carb diets, working out in the gym, dammit, I might be the one to the moon. Now the bad news: 1960, half of the orbital launches blew up. So have a big life insurance policy and make sure that you're comfortable with those odds.

The newly-formed National Aeronautics and Space Administration was not happy with those odds. So what they did is they sat down with the Department of Defense and they came up with a new metric. Now, you can imagine, two federal agencies collaborating on a document. It took forever and the document must be like Trump's tax returns. One page. A one-page document. They defined technology readiness level. A pencil sketch is a TRL-1; breadboard prototype's a three, field test is a five. TRL-9, you know what? Not only are you going to the moon; you're coming back. Bring some moon rocks and we'll have a party on your return. The technology is ready.

So why can't we do that with investment readiness of innovation? Why can't we say, "Which of our projects has the lowest risk, the strongest business model, and in fact, it starts taking the opacity and the politics behind decisions and makes them very transparent?" I love your project. Unfortunately, that's an IRL of four. Yours is an IRL of seven. That is amazing. We've never seen that. IRL of seven. I can't wait to see what your project becomes. And what you're doing, and IRLs are quantified differently by divisions of my companies, but it's an expression of relative risk inherent in anything new, and that tool is a formidable tool. You say, "How can you do that? There's so many different innovations. How can you normalize them?"

Anyone been to the supermarket? Ever? Everything you buy has a nutritional fact label. If you are diabetic, I suggest avoiding the sugar. If you're on a low-carb diet, if you're looking at fat. We know the building blocks of nutrition. We can plan our diets accordingly. Well if you know the building blocks of nutrition and we know the building blocks of innovation, it kind of seems like we should share that learning.

So the question really is, aggregating all those dots, collecting new dots, and allowing you to connect them to discern the patterns that are apparent only once you've done the work. Only if you put in place the infrastructure and the architecture to do that. So if you want to synthesize, analyze, and visualize, let me give you a case study.

Within our portfolios, within all the teams that have played in our software, I mentioned that they're wrong about 7.6 times every time they're right. And it's really hard to get an honest, authentic signal evaluation. And I don't mean confirmation bias; don't you all love blue blazers? Of course you do. Don't you think striped shirts are fabulous? Of course you do. But really you think it's the ugliest outfit you've ever seen; you're just too kind to tell me the truth.

So we had one client, very diverse, lots of technology in their business, and they took a cohort of about eight teams down this eight-week experiment. Let's get them out of the building and talk to customers and see what's real. We were very nervous when they came back with their first cohort. We'd never seen a group more pathetic. We figured that this was a brain-dead experiment, it doesn't work, and I'm going to short their stock, except they're a private company, so I can't. They thought that hospitals was a segment. How many people think hospitals is a market segment? Anybody? No. Urban hospitals that are 300 beds or more with a cardiothoracic surgeon unit, that sounds like a segment. How many people think millennials is a segment? Really? So an 18-year-old fraternity boy at Purdue who's drunk and a 27-year-old married woman in New York with a family of two, they're both millennials, tell me why they would be in the same bucket.

Once you understand what a segment is, what a value proposition is, you can start getting better. And what this company did was remarkable. They had a member in the first cohort mentor the second cohort. And they had a member and a cohort member on the third cohort. They've now established a gold standard. We've never seen this either. We'd never seen companies be so efficient, and part of the efficiency of accelerating the time to truth is being able to articulate the experiments. Jeff Bezos was saying that Amazon's success equals experiments. The more experiments they run every day, every week, every month, the more successful Amazon will be. They are experiment machines. And our belief is that if you are an experiment machine, you will have better hypotheses, a faster path to validation, and understanding that you can't do that unless you have the talent and the culture and the data to allow you to synthesize, analyze, and visualize what is possible to make the best decisions as you find, filter, and fund the next generation of innovation in your businesses, in your industries, and Lord knows in our world, because we sure need it.

So that is the idea of cracking the innovation genome, and we can take some questions if someone wants to throw a yellow cube at you. We'd be happy to do that. Just make sure you duck when the cube is thrown.

Amy:
Questions for Jim?

Jim:
I know they want to go to lunch, so the best question is "What's for lunch?"

Amy:
There's still one more thing before lunch.

Lara:
Sorry, I'm apologizing for everyone, but I did find this really compelling. I'm Lara. I run an innovation program in Spigit, and I think everyone who uses Spigit here can understand there's both the technology, and then there's the program, and it sounds like your technology depends upon the internal group creating of program around the structure of using it. Can you talk a little bit about the key practices you're seeing successful? You've hit on a few, but I think that's sort of the linchpin of a lot of this succeeding.

Jim:
If I heard the question right, it's "So what do I do in a practical sense to go from this idea ... how do I have an investment race-level concept actually done? What do I have to do?" There's a little table back there, so my colleagues will be happy to explain this in a little more detail. It really is about believing that, ultimately, the decisions are not going to be made by your engineers or your programmers or your product marketing people, but by the customers. The worst thing we've seen is, "By the way, we're sales people, so you know you can't talk to them or my customer." Well, that's part of the problem, is we need to talk to them. So as we look at defining the hypothesis crisply and articulating what is a rational experiment? What do you need? I want to talk not to the receptionist at the hospital, not to the surgeon at the hospital, but the person in charge of cardiothoracic supplies and purchasing. And what are you going ask them? And what are you going to learn when they tell you? And how are you going to take that signal that you get and share it so that other people can come and can learn on it?

People using Slack is an interesting artifact, but it's kind of a flat screen. You want to be able to have the collaboration based on the experiments and the evidence. The one thing that is common across lean, agile, open-design thinking are hypothesis experiments and evidence. They have different parts of the world they're looking at. Some use different canvases, different visualization tools, but if you believe that the future belongs to the companies and the individuals that can decode that quickly and fast, then you want all the assets in charge.

So first, what are our best and most promising ideas? Cog Weise and others kind of give us a fighting chance of, man, these seven, eight, 12, 15, they really look good. Trust me, none of them are going to go to the market and be successful without at least some minor or major changes, 28 changes on average. I want to do those 28 changes really fast and really cheap, and I want to know that if they can't really be done, let's just move on.

One of the first teams in that first failed cohort, they pulled us aside and said, "We've got eight teams, but team number seven, the guy who started our medical division, it's his idea. It's a brilliant idea." So they're in here so the people can learn from them, but it's already been green-lighted. In five weeks they shut it down, because there's some fundamental assumptions that everyone believe were true because of the person who said them, so they must be true. When you really got into the marketplace, never going to happen. No one lost their job. They got to do other cool, fun stuff and the company saved over $10 million that was prepared to be spent on something that they would have found out two years later. Probably shouldn't have spent.

So a fast lose, a fast no, is a huge win. Again, giving them the tools to share, to collaborate, to be visible, to be transparent, and aggregate them in away that give them metrics that matter to them was what was critical for them to go from an idea factory to actually driving the bottom line. "Innovation" is code word for "growth." If we were all growing at 80% a year and our bottom line was just exploding, we might pay less attention to innovation, but for them, innovation was survival.

Amy:
We've got time for one more for Jim.

Jim:
Wow.

Amy:
Thank you.

Jim:
Thank you.

Amy:
Great job, Jim. Thank you.