Here are the transcripts to my discussion with entrepreneur Steve Blank. I apologize for the general lack of punctuation.
______________________
Eric Lofgren: i’m speaking with steve blank who is a serial entrepreneur, a former air force officer, a founder of hacking for defense, a member of the defense business board, and an adjunct professor at Stanford. steve thanks for joining me on acquisition talk
Steve Blank: thanks for having me
Eric: great so i want to start off here with uh blank’s rule which states that in order to predict the future you basically have to have one third of you be crazy and this is incredibly important kind of concept in my mind and it strikes me as related to Michael Polanyi he was essentially a philosopher of science back in the 20th century and he had this concept of tacit knowledge that real innovation and real progress require these anticipations about what’s likely to be correct in the future and it can’t really be verified by third parties using explicit rules like math or known facts like natural laws now of course in the department of defense for a lot of acquisition programs this is exactly what you need you need these types of articulated information that can be consensually agreed upon by layers of bureaucracy and neutral experts. so craziness isn’t really allowed in the defense acquisition programs in my view. can you just talk a little bit about blank’s rule and then how that manifests in commercial and government activities
Steve: sure i mean it was just an observation about the distinction of how startups work and how entrepreneurs think and this overused word visionaries think versus how people come to work daily to do their jobs and what i’ve found is that most people whether they’re in the commercial world or in the dod tend to be great at predicting the past and while that’s fine in a static environment and when your competitors are also predicting the past you know no harm no foul but in fact when you have people operating with different mindsets you could rapidly fall behind and I think that i mean to diverge from your question but that’s obviously what’s happened in the last two decades and then dod versus china and even Russia as well we were focused on making sure we got you know dealing with non-nation states right thinking great power conflict is over because we won in the 20th century we woke up to realize that not only were our adversaries or new adversaries buying new systems they basically obsoleted all our old ones and yet we haven’t gotten that memo you know and it’s not just an acquisition problem it’s an imagination problem and it has all the other accoutrements of the problem of acquisition of a de-pockets democracy lobbyists you know congressional interests you know jobs etc that really makes those transitions from battleships to aircraft carriers you know massively hard or from horses and horse blankets to tanks and uh you know people written about this for the last 50 years and so i’m not the first to to talk about it but I and neither are you but i think our voices need to get much louder because um our adversaries have figured this out uh really well and we still haven’t gotten the memo but what was your question sorry
Eric: so i think you’re addressing the question but i guess some of the question is that in the department of defense in order to get an acquisition program approved it basically has to go through layers and layers of bureaucracy some of which are neutral experts some of which are you know laymen on the on the business side and they all basically have to agree to this thing and so it needs this articulated plan a cost estimate where are you gonna go over the next five ten what’s your production plan before you even start you know full scale development. so it’s really like this big waterfall process where you’re supposed to know exactly what the end state’s gonna be before you really start and that seems to kind of go against blank’s rule and as well as michael polanyi and a lot of these other guys that are saying hey you kind of have to have some imagination. i want to just read a quick quote here from polyani he says “looking forward before the event the act of discovery appears personal and indeterminate it starts with the solitary intimations of a problem and bits of pieces here and there which seem to offer clues of something hidden they seem like fragments of a yet unknown coherent whole,” but we can’t really start a program unless we think that we know exactly what that coherent whole is .
Steve: so i’d summarize what you said in a different way you know most visit most people who think they’re visionaries are actually hallucinating yeah but there are a few who actually do see the future and the problem is that finding that signal and noise really unnerves people whose jobs it is to in fact simply fund a better version of the current thing and so in in business we you know about 30 years ago mckinsey came up with a model that helps companies think about how to how to think about this called the three horizons of innovation and it’s a it’s a kind of a good shorthand that says look every good organization needs to incrementally add features or additional things to their current product or think of this as their current weapon system bigger whatever even and that’s called a horizon one innovation and of course everybody should be doing that and that’s classic requirements and acquisition is translated into dod and but then there’s you know like horizon two which says let’s take an existing system and maybe repurpose it uh like putting the new missiles existing missiles on new uh on navy ships uh what are we doing putting harpoons or their equivalents now on ships for longer reach and whatever the missiles exist we’re just kind of repurposing them etc and it’s a new program it’s true truly kind of innovative in in how it’s used but didn’t require a huge insight it requires someone to say let’s take this and put it over here to solve a new problem and by the way traditional requirements and acquisition can do our traditional okay job if you think it’s okay but the ones that really screw us up are the ones in commercial world we call disruptive innovations disruptive innovations there is no five year plan and this is back to the quotes you made is that you have to imagine what the future is and for the dod the problem is one is we tend to think that those are all technology driven. Let’s spend several billion dollars building shiny object x or y and sometimes they are and i’ll give you some great examples of where we need to meet to but the real trap is thinking they’re only technology driven. you know the biggest disruptions that’s happened to the dod in the last 75 years probably since pearl harbor were the islands in the south china sea what kind of technology did that require that was a huge disruptive idea and I hope there’s some 30-foot statue of some chinese colonel somewhere who said well sir we need bases in the middle of the ocean and before they shot him he said no no no we own the biggest dredging ships in the world well we just dumped sand on these reefs and all of a sudden they basically negated 75 years of our ability to project power in the western pacific along with the f-21s and everything else and they’ve made it navy’s life incredibly hard and so disruption could also be my point is disruption is not only tech um it’s also sometimes just innovative thinking in a way that just is non-consensual non-traditional but specifically for acquisitions of systems it’s pretty clear that we’ve invested for the last you know since and during the cold war and survivable systems where it’s pretty clear we need whether it’s navy or air force or anything else to go to attritable assets you know we invested in a large and best in breed platforms well it’s pretty clear that those are you know have little targets or big targets painted on them we now need distributed assets and a great example is you know we have battlestar galactica’s up in space one could imagine we talk about but spacex is about to launch 44 000 you know satellites for in starlink as a communication network imagine that kind of targeting problem for our adversaries versus forty four versus forty four thousand but that’s not where we’re going or at least publicly we were proud that we built best in class features our you know things have the best resolution or the best whatever but in fact if you decide that distributed assets are the way to go then you start thinking about aggregation of mass good enough features well that also changes the game about the entry of adversaries they don’t have to be as good as errors but they could have a thousand times more and that’s not just some hyperbola we’re used to features being fixed and updates to systems taking months or years if ever and a new contract and whatever i have to tell you in the startup world i mean people are aghast we update our systems in hours or days continually and no features are fixed and if you don’t believe that could be do done with complex systems teslas are updated that is car is updated literally within every couple weeks for the last eight years i’ve owned a tesla you know we built stealth which was great when you had one or few platforms but you know i think people are going to realize when you’re when you’re spending billions for a single platform you might want to decide that perhaps a mass form of things that are observable rather than a single thing that’s unobservable i could go on and on but my point is it’s not just you know shiny object tech it’s that the whole kind of class of systems we need to build are radically different than the things we built in the 20th century and more importantly and this is for me the big one it’s just not an acquisition problem these new systems will create a whole new set of operational concepts we won’t use robotics on the battlefield the same way we were using men and people we won’t use drone swarms exactly like we were using aircraft yes we’ll have you know unmanned wingmen thank you for being innovative but that’s not how future air battles are going to be fought they’re going to be fought by machines for dominance of airspace um you know and what that means is you know the existing primes and this goes back to acquisition who essentially and again i don’t mean this is a pejorative but as a way to think about it the old primes were essentially sheet metal vendors who had to figure out software and we’re seeing some of the problems of that when that’s not built into your dna i think we’re going to see new primes that actually were built as software companies that learn how to bend sheet metal and and if i had a bet on what the future is it’s not going to be the sheet metal vendors and and again i don’t mean that as a pejorative and they have a long history but it’s like discussing the long and valued history of the cavalry great whatever guys but that’s not where this is going to be in 20 years and if we think it is we’re going to be pretty unmatched and you know this waterfall engineering versus agile i think we hear this all the time and it’s just again to people in the commercial world no one builds fixed features specked up front not understanding that you know the problem’s going to change you truly don’t understand the requirements until you get out of the building you need to build products iteratively and interactively you know something as simple as like gee what do you mean we didn’t build minimum viable products of those elevators on the on a ford class carrier or tried some of these things out always too expensive well you know I think we see the even on things we should have known better on and what that means is today for example in headquarters staff innovation is you know kind of like some disconnected activity or gee you know we’ll worry about that later or gee my requirements people got it worse in fact in particular our adversaries continuous innovation is integral to the mission at headquarters level and more importantly civilian military fusion is like what they do and equally interesting for acquisition i see this the air force being the classic example you know we love to buy things and operate them in the military and the consequences for example we love to buy rockets you know okay let’s buy a rocket but now we’re going to launch it well that means that the contractor has to design things for with a tech manual for operators with a certain grade and experience level etc but the model really is going to go to for a good number of our systems we ought to buy outcomes not things and what i mean by that is you know why are we not just contracting for i need this much bandwidth and i need this much you know payload into space and i mean well because real men love you know to own rockets or real men who love to own satellites or real men i could go on and on some things obviously can’t be acquired like that but there are a lot of things that now we could be thinking about services and even as dod internal services let alone commercial services so sorry i’ve been babbling way past what you asked
Eric: no that was great that you touched on just an incredible number of things that i actually wanted to get to. actually i heard recently that the army was uh talking to spacex about contracting for starlink and even they were saying something about starship it wasn’t clear what what they wanted out of starship but yeah i think there’s there you know there’s a lot of interesting things there you know when we think about the cavalry you brought up the cavalry back after world war one they were actually they did a big study on that and they said okay what’s the cavalry you know gonna do in this age of tank warfare and the study came back said horses make people smarter on the battlefield we’ll just drive them up to the battlefield and then we’ll do the exact same thing that we used to do it’s just like so i think a lot of what you’re talking about kind of gets around ambidexterity so can you just describe you know what is ambidexterity and like how do large organizations need that?
Steve: in order to do things across the three horizons that you were talking about yeah so ambidexterity for those of your listeners is not familiar with it because it’s a like a weird word to use in an organization essentially means your ability to chew gum and walk at the same time and by that i mean look you have your current mission and you know in a commercial company you have your current customers and products and services in in the dod think of it as your current mission and you want to focus your energy and your requirements and new acquisitions on current mission and as i said earlier if the world is static that’s fine you know and in fact if your adversary is moving at the same pace with the same tech as you are great you could kind of keep up and you guys will leapfrog each other year to year but it’ll be kind of that’s fine but that’s not how the world is today the world is uncertain and people are and more importantly technology is being driven not just by dod labs but a good chunk of it everything but maybe hypersonics that’s about it maybe some advanced quantum stuff are not just dod tech they’re commercial tech being driven by much more money than the dod has in their budgets so now all of a sudden innovation is occurring at a rate that’s different than what mission requirements are for current systems so how do you separate out those separate lanes ambidexterous says look i need to obviously support my current mission but i need to be thinking about this is goes back to the crazy things that might occur in the future how will we build swarms of systems not just swarm drones that look like you know existing concepts for what we’re doing today or how do we not have oh it’s an optionally manned bomber oh what a surprise you know why don’t we think about what the concept is about securing airspace or delivering payload on target or delivering payloads from space or not just the issue is payloads in space but the issue is survivable communications and intelligence and other things we do up there those are different outcomes and to do that and here’s the big idea you need different people most people and i’ll that’s my experience of 40 years of doing this most people in large organizations want to come to work do a job that’s prescribed by a job description get promoted and you know like and then go home and hug their kids and wives and really have a good time but there’s a few people just a few not better you know not smarter always but very different and these people hear things that other people don’t they definitely see things that other people don’t and they think differently and in most organizations they’re called the mavericks or you know the innovators alliance and whatever and they’re beaten to the ground because there isn’t an organizational structure that embraces the ability to do execution and innovation in parallel and what’s worse is in the last five years good parts of the dod have adopted a good chunk of what startups do my lean startup methodology and if you know there’s an explosion of incubators and accelerators across the dod and and i think that’s wonderful in shaping culture but actually it’s created innovation theater and innovation theater is hey look you know we point to the general and the general could point to the civilians we have one of these two it looks like just like large company x and then you go uh sir uh how many programs of record have come out of this thing oh let me show you our coffee cups and posters and lanyards and you go no no no you know how many billion dollar acquisition programs have come out of all these things and the answer is there’s probably you know probably one acquisition program you know of a billion dollars exceeds anything that’s been funded outside of these accelerators and so it’s not that we should stop doing them is that we really don’t have what i and pete newell call an innovation doctrine we really don’t understand how to connect all these what what i’ll call point activities into a transition program of how do we use innovation and disruption to create a whole new series of both weapon systems and more importantly operational concepts that not only in some cases catches it up to our adversaries but actually get us ahead you know i think as a civilian now looking at it from the outside i think the biggest disservice the dod does to itself today is describing china as a near peer adversary i’m not sure that’s true in the western pacific you know you know i’ve seen enough to convince me that there might be a lot of stuff going to the bottom of the ocean and though hopefully we’ll never figure out that’s not a near period description and so we ought to be and obviously there are there are things we can’t talk about that we’re doing but i don’t think congress understands this i i don’t think the message has been delivered to the primes and there’s just some real i mean if you look at the total number of dollars the prime spend or or get every year from the dod i don’t know what is it the top ten you’re the expert is like a quarter trillion dollars or something or and then you say well how much are they spending on acquiring startups and you know my mother requires more startups than they do you know what’s lockheed’s venture capital budget 40 million dollars and you go that’s their pencil budget no no that’s their vc budget that would be like a in in silicon valley that wouldn’t even make the list to have get online for lunch somewhere and so there’s no incentive from the dod to the primes that said hey you know i’ll tell you what everything you acquire that ends up in a weapon system from startup will pay you 2x or everything innovative that you buy and deploy and want to that is there isn’t an end-to-end uh innovation process that has incentives and i don’t believe in sticks i believe in lots of carats to motivate our existing primes to kind of like welcome to the 21st century and more importantly to motivate the new set of primes that have stood up the palantirs and the andurils and the spacex’s you know spacex’s fight and palantirs fight with the dod should be illustrative about like you know all those horse blanket and horseshoe generals who fought them if it were me i would be taken away their robot to retirement that it’s like excuse me but you did a better job that china could have done and like delaying our ability to respond sorry for the solo queen no that’s great you know with this transition problem in the innovation theater that you’re talking about you know a lot of it for me it kind of gets back to you know how do you actually transition that program and it seems like well we want to adopt what you’re saying in lean methodology but then every like statutory or policy type thing that we have an apartment offense is waterfall so the air force the platform one which is their software factory they have a motto that says waterfall only you can prevent but you know we’ve been doing this for many years you know for decades we’ve been talking about incremental or evolutionary or iterative methods and they just seem to be stuck in this you know waterfall process where it takes you know one of the big parts is you have a multi-year budget process in order to get funding it takes me multiple years to even line up that funding and then transition it assuming i can get everybody on board so can you just talk a little bit about agile fall what is agile fault and you know why do large enterprises like the department of defense get stuck in this yeah you know it’s probably just worth if if you don’t mind going back and actually explaining for the two listeners who don’t know the difference between waterfall and agile and you know i grew up in the 20th century where waterfall was the methodology of how we built things wonderful is just a shorthand for a description of how you build products and services in in in the old days and the ultimate were in some parts of the dod still today and it basically said look why don’t we specify every possible requirement for this new thing we want right we’ll write it down it’s a and it’ll turn into a long feature list and we’ll get it approved and whatever and then we’ll acquire it and by the way we’ll set dates and times and you know for delivery and we’re going to have that entire feature list and we’re going to send it out to a contractor and they’re not allowed to make changes because they just signed up to that feature list and so what they do so you start with a set of requirements you go to a feature list and then you basically lock your engineers into a room and like they start developing either hardware software and you might have milestones for demos and whatever or proof of principles or whatever it is and then the product or services comes out the other end and typically at least in when i was building companies for years and then you discover oh i really didn’t understand those customer needs because there was no requirement to actually talk to them and or often oh yeah those kind of were the needs three years ago but like the world has kind of changed and if you would have asked we would have told you know 18 months ago those weren’t needed or technologies moved on and the thing you built is a fixed system is unupgradable unless you unless we give you another 100 million dollar contract to add free features and that’s the world of waterfall now if you kind of deconstruct it it says really on day one that you have to be able to predict the future it’s a big idea not only do you have to be able to predict the future when you write those requirements you have to assume that you as the requirements writer sitting in your dark office deep somewhere in a big building with no windows or maybe you have a window are able to kind of understand every possible customer need because you’re the smartest person in the world but in fact we know that’s not true there’s no way you could be smarter than the collective intelligence of potential customers but waterfall doesn’t allow us to kind of take that into an account lean in contrast is a very different model it says look instead of us trying to be the smartest people in the world trying to predict the future why don’t we just assume we can’t predict the future and why don’t we assume that if we’re going to try to build something and deliver it all on mass three years or ten years later that perhaps we ought to have some serious intermediate steps not tests but intermediate steps where we could test our assumptions because if you really think about it all we have is a series of untested hypotheses on day one now for buying the next rifle those assumptions probably are understood but if we’re doing something big and complex that’s new we might want to think about how do we build a process that’s agile that allows us to build things incrementally and iteratively and at every step testing our assumptions whether it’s does this work or does it fit or is it still needed or are these features in the right order with the users and customers and warfighters and everybody a piece at a time and that’s called agile and there’s so there’s known ways now to do this for hardware software complex systems etc that basically reduces the amount of money and time spent to get the product is delivered correct right and again the biggest examples and they’re so beaten up i’m sorry to pick on them but you know the four-class carriers and they F35 right i mean you would think those requirements and acquisition programs were run by the Chinese because i think that nothing could have set our country back so much about spending time money and resources eventually of course they’re going to be great systems but if we actually were building things with a different methodology about figuring out which pieces of these things have the higher risk and iteratively testing and even for the most complex systems we could have done this and again the bigger assumptions are or is this the right system to that solves a problem that’s still going to be current when we eventually deliver it now i uh I understand much like in large companies there are external forces that have nothing to do with waterfall versus agile and how we deliver so for example if you do waterfall and you’re a prime contractor you want to make sure you have fixed requirements and fix systems because every new change is a new major contract being able to you know upgrade a a system with a touch of a button like your tesla can or some other systems we build or amazon can or google can you know is an anathema to a prime whose business model is predicated on low cost bidding for the initial system and then make the money on the upgrades right that business model doesn’t allow you to kind of build agile and upgradeable systems because you’ll go out you’ll go broke or more importantly you’ll lose the deal you know the other part is that like um every prime and at least the large ones take the top ten have jobs in the right congressional districts that you know make sure that these systems take years and decades no one is interested in something that you could build in months and then you have to re-compete and maybe lose those jobs in that district and so therefore we have large lobbying dollars in fact the lobbying dollars for you know a prime exceeds you know what it cost a startup to build a you know a new company did i answer a question
Eric: yeah pretty much. you know it’s interesting you’d think that — and this was from oliver williamson he kind of said this you know if you actually chunk those down into smaller efforts and we saw with spacex nasa actually in the early years they had a other transaction they had payable milestones so there was like these you know incremental steps where they would actually get funding for different accomplishments that were outcomes based right so i guess you would think that there would be sales stabilization actually if you had smaller more frequent contracts rather than these big one time take all and then you know once that thing ramps down you have to have a follow-on or you just get rid of this entire production facility so why is modular contracting are smaller contracts more iteratively done aligned with an agile methodology you know why do we not do that? why are we stuck with this kind of waterfall overall process in the funding mechanism and the requirements mechanism and the contract mechanism? why do we have agile fall?
Steve: well you know if if you think about it let’s go all the way up to the top i mean the benign view is the dod itself internally has done a bad job educating itself and its requirements writers and its acquisition i mean defense acquisition university is slowly getting it but if you want to start in a place that might be useful i think from top to bottom you know people now need to understand you know the nature of agile and lean and you know i still talk to numbers of staff people when you use the word lean they think you’re talking about head count seriously it’s like oh yes we’re reducing the number of billets you know uh no that’s not what we mean so what is an internal language and education problem inside of requirements and acquisition inside the dod and then to senior staff inside the dod about what’s possible it’s a big idea I don’t think we’ve done a good job educating ourselves about the possible and the possible is not here’s the religion the possible is here’s the outcome of the religion here’s what it will let us do to keep pace with our adversaries who have already adopted this this is a big idea the second part is to and you live this is to educate the congressional staffs first the staffs about the benefits and the outcomes and the whatever and the impact on the safety and security of the country and then finally educate the congressmen themselves who have different issues about and and you obviously the dod needs to deal with the political realities of i mean it’s like you know when kennedy says you know about the famous rice speech he also started out about why does rice play texas well you know the same thing is why do we have sls right there’s no rational reason to have the space launch system other than that there’s a senator who controls the budget and then there’s no you know um you know the nasa buck rogers quote no bucks no buck rogers i mean no sls no nasa so you know dod obviously understands the politics of that but i think everybody ought to understand that we could be developing systems faster cheaper better more timely more needed more agile being able to get this ambidextrous path going with a different mindset different methodology but we’ve done a bad job and i think your blog and others clearly are the beginnings of a set of people who get it i also think it’s a generation gap i think um you now enlisted men and officers coming in today you know most of them have you know used github in the stack exchange and whatever and go run that test on 06’s and above versus 03s and below or e3’s and below for those who kind of you know work in those areas i mean i I truly think it’s what you grew up being used to i just hope it’s not too late that as those as those officers and enlisted growing ranks that’s not too late for us to kind of catch up to a both russia and china who and even you know regional threats like you know the other two out of the two plus three who have figured out uh how to use this stuff does that answer your question
Eric: yeah definitely i want to kind of go back to what you’re talking a little bit about earlier on in terms of customer interaction and just like we have this big waterfall process and then we don’t talk to the customer and so you’ve given several examples just in the commercial world where startups try to build these products and then ship it and then they just find out later uh oh the customers aren’t behaving as expected right and this is kind of the opposite of what john kenneth galbraith used to think you know back in in the 1950s and 60s he thought that firms could just you know ram these products down the throats of consumers using advertising or otherwise but it seems like galbraith’s model is actually kind of appropriate for the department of defense because users don’t really seem to have this choice right they don’t have their own acquisition money. their inputs way at the front and then a new generation of users is kind of getting the product that comes out of that system later. so most of the people in leadership positions today they seem to get what you’re saying right they when they talk on their webinars or whatever. they’re using the language you’re using and you’ve been involved in this to a lot much greater degree than a lot of people. so can you just talk about have you seen the users getting more involved in the process in a more iterative fashion or is that changing things in the department of defense? how do you see that actually playing out? or is it just this is a lot of more innovation theater and we’re not actually getting the users involved that much?
Steve: well you know i think if you look at the head of acquisition of the air force roper and jetty from you know he used to run the rapid equipping force and army and some of the other surveillance look at the marine corps new operational concepts and what they’re asking for i think we’re starting to see that change but you know there’s all these moving parts to kind of you know turn that super tiger in this case during that battleship you know and a good part of it you kind of articulated earlier is we’re kind of restricted by what laws and acquisition processes are embedded in in those regulations starting with title 10 and moving down and so it’s not just what do we wish I mean socom was already doing this a couple of years ago right and they were kind of really now afwerks and others but i think the test for me will and i’ll go back to what i said earlier is that a lot of great you know talk but until some of those become programs of records that are measured in billions and the folks who you know are still buying things because they like to play with them rather than buying services and outcomes i think until that mindset becomes dominant this is still going to be a long struggle and i do believe that the folks who get it could do a much better job in education it’s when people stop talking about lean as like for example as i said it’s not a head count problem it’s a way to work does that answer the question that is I i’m hopeful i just think it’s been a series of unorganized efforts from if not the and i won’t say from the bottom up i think in for decades there have always been heroic innovation activities at the bottom the good news is there’s now some heroic innovation activities occurring on the top but they’re not connected together in an integrated way to say this is the new model of how the dod works and here are the laws we’re changing here’s how we’re revamping the entire you know eisenhower curriculum and supply chain and procurement and how we’re revamping the uh acquisition university from top to bottom not just adding another thing about agile here are the laws and things that need to be changed my favorite one is oh my god you know if security is like again and i i think i love our security folks because of what they’re trying to do but how they do it it just seems to be again sometimes so counterproductive to get low side to high side stuff in from where the innovations are happening in the commercial world it’s you know we’ve run off some of the best companies and people who want to help I don’t know if you ever can imagine that there’s a great cartoon of a castle under siege in the middle ages and there’s a king looking over the battlements and he’s surrounded by all these enemy troops and there’s someone knocking on the on the back door and you can see it’s a salesman and he’s carrying a machine gun and the king goes to his courtier you know tell the salesman to go away can’t you see him fighting a war and of course the you know if he would have let the guy in it would help encounter that all the time when uh on the security side when commercial companies try to you know figure out how to knock on the doors of 20th century security models so when we talk about acquisition we also have to talk about you know how do we bridge this partnership gap with a commercial world for me that’s truly a major failing so far of the dod i mean we talk about otas and we talk about mid-tier procurement and i think ellen lord has done a good job in trying to change that model but again to be honest from where i sit it’s still pasting solutions on an existing system rather than going no guys we’re now buying we’re buying outcomes so why don’t we start with that kind of system or no you know most of our stuff ought to be commercial this is what perry did back in the 80s you know and you know there ought to be a version of the last supper again that you know i don’t know if your listeners are familiar with when bill perry told the con do the primes that there’s look around the table half are you going to be gone and you ought to figure out which half yourselves before we do it i think the same conversation needs to be had with the sheet metal vendors that says guys you’re you know your hardware guys trying to play software no you guys need to be software companies who make hardware why don’t you figure out who you’re acquiring and who you’re partnering with or else you’re not getting any more contracts boy that would be a really interesting conversation and yeah they’ll all go to their lobbyists and go to congress and whatever so you’ve got to play that out as a chess game but I think that’s what’s necessary for us to be able to compete you know the best book that um and um i don’t mean to pitch this but it was a great um read was called kill chain by mccain’s staffer who’s now at anduril i thought it was a great summation of the problems the country faces and the real eye-opener for me was the observation that we might not win the next war and the next war might actually involve attacks in the u.s homeland might not be physical troops but you know cyber and other things i mean things that are unthinkable for most americans losing a war we don’t lose wars we sometimes bail out on our adversary and our allies but we don’t lose them you know this might be the first time when we have to deal with that and I don’t think the american public and certainly congress has not gotten their heads around what’s it going to take to win the next wars yeah i mean i definitely hear you on that the the future wars are being fought in today’s you know program offices right like the developments that are happening now are going to be affecting those and that’s why i think there’s a lot of urgency as to you know what we’re doing today in acquisition and in the technology side of getting these programs out and i completely agree that it seems like we have this huge edifice and then we’re just tacking on little agile you know leaflets on on top of that and one example is for example the agile earned value management guide which is the current instantiation of pert but that’s just like it seems like an oxymoron right where it’s like you I want you to do an agile waterfall process what does that mean with an emphasis on the second syllable right it’s not right and i think that’s what’s stopping some of these
Eric: i’m with Marc Andreessen on this idea that you know software’s heating the world and the corollary is that software firms are actually going to be better at doing hardware than hardware firms are going to be able to pick up software. but when you have to do these earned value management things and get into a program of record which is still defined in kind of a waterfall way it’s almost like boxing those guys out and i think that we almost need to reimagine it’s not just can we transition more things into a program of record we have to re-imagine what that program of record is and i think roper was trying to do that with advanced battle management system which is the new kind of command and control system for the air force and he got a lot of pushback from the general accountability office on that and you said something interesting that you said well government’s not really like it’s just not government’s not a large startup it’s not like that there’s lots of reasons for it but some people are calling abms and maybe some of these other programs like kessel run they’re calling them something like a startup within government. so how do you see this? can and should government carve out these small organizations with a startup mentality or are there just too many contradictions in the incentives and the way it works for that to actually happen?
Steve: so i’ll give you a business school answer which is yes to both and so what i think you know the advanced management system is and kessel run, afwerks, or whatever and all the incubators and experiments i think what they’re doing right now is education i think they’re educating leadership and congress and what we’ll almost call the old guard i mean think of them as the battleship you know admirals that that there’s a different way of doing things but what this lacks is it doesn’t fit into an integrated innovation doctrine that all services dod acquisition congress have bought into of here’s the way we’re going to be doing business and what that means is we’re going to spend if we’re not careful the next five or ten years having these piecemeal battles without having an integrated view of where we want to get how do we get to it etc and we’re going to be arguing about specific systems rather than reforming an entire acquisition system an entire requirements writing system and acquiring an entire pro how we deal with programs of record we need a meta view here and that’s why i keep going back to an innovation doctrine that like everybody kind of agrees on or at least i don’t care if you agree on it let’s just have one and i think okay it’s now because we don’t have one and we haven’t had visionary leadership that kind of that is i think we let me just go back i think what mattis did with the national defense strategy to wake up the entire country to say look guys it’s two plus three you know we spent the last 20 years fighting you know non-nation states they’re still one of the three but we really ought to wake up we need a document like that so broad reaching but with a with a detailed annex of what we want to do for acquisition and weapon systems and concepts and I think we’re missing that equivalent does that make sense
Eric: yeah you know when i look back on like the 1950s and before you know it seemed like defense technology and weapon systems were just more agile and innovative in those years world war ii 1950s and then in the 60s and maybe 70s you know pure teal kind of march 1970s kind of like this breaking point where you know a lot not just defense but like a lot of a lot of sectors in in the economy kind of became less innovative especially on the hardware side do you think that’s true are there lessons from the 50s in world war ii eras that we should be looking at to model this or should we just completely reimagine it and say that was that worked then, we’re now in a new paradigm. how you see that?
Steve: you know unless you’re a historian of innovation you kind of forget that in world war ii we invented civilian military fusion there was something called the osrd that basically drafted professors but kept them out of uniform and gave them military weapon systems problems and allowed them to staff the weapons labs inside of universities it’s how we did radar and mit electronic warfare at harvard under fred turman who ended up a stanford professor ended up creating the innovation ecosystem at stanford post war you know the the physics problem started at berkeley was spun out of osrd it was called the manhattan project and its outcome you know had some interesting consequences for the next uh 50 years and so there was a a model that said we did that um we kind of ended that with world war ii and then went on to a contractor model but the but the dod was capable of picking winners and losers and that kind of is what stopped in the70s there’s a bunch of us in the startup world who think that one of the big problems with these incubators and accelerators is giving out you know million dollar grants is actually detrimental to getting effective weapon systems is that no you don’t want to hurt anybody’s feelings so now every everybody gets a prize for participation rather than the goal isn’t you showed up on the soccer field the goal is to win the f in game we’re kind of forgetting that that’s also the goal of these incubators and accelerators and um and if we’re not prepared to pick winners and losers then we’re not prepared to actually you know move what we need to move forward that’s the other problem with what we’re doing so to answer your question i think we could come back to a way where civilian and commercial world not only could help the dod but i think the dod in the last 30 or so years has been so blinded by their primes that they’ve just forgotten to look outside the building and seen what’s going on the commercial world i think with the jaic [joint artificial intelligence center] we now understand the power of ai i think machine learning and vision techniques i think we’re still confusing shiny object and technology with weapon systems and operational concepts but okay at least we’re starting to look outside the building again my favorite horse to beat is you know i think the 10-year struggle to get spacex recognized as a resource for the country and the dod is just illustrative of the problem the rest of the dod has in building a new alliance between the military and civilian world that just doesn’t exist today and i know this because i and joe felter and pete newell built a program called hacking for defense which we now through a defense innovation unit run in 30 to 40 universities where students at research universities work on dod problems these are students that the dod would never have seen i mean they would go to facebook or google or whatever but the post-class connect rate is over 40 you know it’s a small token of what the dod ought to be doing in funding startups picking winners you know engaging with the venture capital world there’s an underground venture capital group in silicon valley now started by us not the dod called the defense investors network there’s about 75 venture capitalists who invest in dual use technologies and the dod is still trying to figure out what the heck it is so we decided to kind of do it so there’s a series of things that could be done if the dod appreciated the commercial world a bit more to build a acquisition system that embraced them rather than stood them off at arm’s length and incented the in current primes to figure out where the future is all those companies that you brought up earlier the palantirs the spacex’s the andurils and then what you’re talking about with the defense innovation network it seems like all of those kind of came in for almost ideological reasons on their own not because it seemed like a good business opportunity they knew it was going to take many years and potentially a failure and it’s going to be a really hard hack to get in there but they stuck with it anyway for your listeners and required a billionaire with idealistic exactly because no startup could afford to deal with the department of defense this is a big idea for your listeners in the dod no startup in silicon valley can spend you know like the first pile of paper they run the other way you know oh otas are better you got to be kidding you know where startup could raise 25 million dollars in an afternoon the dod just has no conception about the speed that the commercial world operates on now and again that’s a real failing for the country because our adversaries have figured that out you know if you want the most visible look just look at the number of generations of ships china has turned out while we’re still arguing about you know the arresting gear on the on the ford class carrier and again it’s not a technology problem it’s a people in process problem and it’s people recognizing that there’s an a major impedance mismatch between what the commercial world has and how they operate and how the dod needs to partner with them speed urgency funding i mentioned earlier security issues etc and the only people who as now back to your point who could overcome them were billionaires with time patience and a fanatical interest in helping their country in spite of itself i want to say this again in spite of itself and that’s just you know it’s just hard to wrap your head again around that
Eric: i completely agree with the emphasis on the people and the organizations and it seems like everything that we do is focused on the program what’s the program life cycle how are we going to like predict that and manage that rather than how do we manage these organizations and you know you brought this up a little bit ago with uh after world war ii you know we basically uh neutered the bureau and arsenal systems which had all these great in-house technical abilities that allowed the people in government to pick winners and losers rather than…
Steve: you have a kind of business person or someone that’s not really proficient in that area coming in and having to rely on these processes and now other transactions which are supposed to be commercial-like and quick actually involve a great deal of paperwork in an avalanche of other types of processes and to be fair a good chunk of this is the fault of congress is that for every corner case of a scandal instead of saying this is a corner case we just layered on another set of rules that actually blanketed everybody rather than just saying hey you know what like in every 100 you know acquisitions
there’s going to be one criminal activity um yeah so instead of shutting down the other 99 why don’t we just focus on how to how to weed out the one what people don’t understand particularly in congress and also the dod is when you do innovation most of them fail this is a culture problem right if you don’t embrace failure and risk in the more innovative the ideas and more disruptive this is a funnel not a pipeline when you’re acquiring a new gun gee we kind of know how to build guns and therefore we should not accept failure in those types of acquisitions but we’re when we’re trying new concepts and new ideas if we confuse that with execution rather than innovation this goes back to and risk ambidextrous for those types of activities and given that our current culture plays a gotcha game rather than what startups do is we you know you know we have a special word for a failed entrepreneur in silicon valley you know what it is experienced try that in congress let’s go oh there’s an investigation of why did this fail how come you didn’t predict whatever if we don’t educate people that says look the innovation pipeline has different risk profiles than the execution pipeline those things that require imagination and vision most of them are actually hallucinations but we’ll spend money on them but the ones or twos that exist those will be game changers and we ought to have the stomach and the inside and the vision and the passion to be able to do both to be able to acquire things that that we need to sustain our current capabilities but also have the vision to kind of see the future and take the risk to acquire things that that will get us ahead but a good number of them might not work
Eric: that’s a great place to end. steve blank thanks for joining me on acquisition talk
Steve: this was great thank you so much for your time
Leave a Reply