Fireside with Mike Clayville of Amazon Web Services | Greyscale

You.

In this episode, Jerry Chen sits down with Amazon Web Services Vice President of Worldwide Commercial Sales Mike Clayville to talk about sales strategy, how Mike spots industry trends and aws. This fireside chat was recorded at Grayscale, an event hosted by Greylock Partners. For more podcasts please please visit news.greylock.com.

I want to invite Mike Claville up on stage. We're going to do a fireside chat for the next half hour. Then Tim has a question. So by introduction, Mike Clayville is the VP of Commercial sales for Amazon Web Services. So Mike runs an entire go to market field for the non government business at Amazon's cloud.

That's right.

I'll embarrass Mike a little bit. So Mike and I know each other, we work together at VMware and we talk about the whole theme today is this hyperscaling, blitzscaling. But you been at Amazon how long?

So three years.

Three years. So you've seen it go from what it's like a $10 billion run rate now what was when you started I.

Guess somewhere north of 2 billion probably.

So in three years has gone from a 2 billion run rate to $10 billion run rate 5x. So growing that fast at scale is kind of amazing. But not the first time you've seen this rodeo. Being Mike, being a professional cowboy on the side VMware, you were there for.

Six years as years. Can you believe it?

Nine years.

Nine years.

And so you saw that grow from how many?

So it went from, I think it closed the previous year before I got there on about 100 million to 5 billion.

5 billion. And before that you were at a small company, BA Systems.

BA Systems. And it had gone from zero to about a billion in six years.

And before that you were at Tivoli.

As part of IBM Tivoli and Tivoli had gone from about 50 million to about a billion and a half.

And then before that a small company called Markham Systems in the mainframe space.

And they'd gone from like 0 to 300 million in just a couple of years.

So Mike is either the luckiest guy.

On the planet, but I started at IBM, so I didn't, I wasn't. It's not like I chose everyone, right? I chose IBM when it was the most profitable company in the history and I watched it dove just to the least profitable company in history in just a few years.

There's probably some learning in that. But what VCs like to say are like things like the IBM versus other ones. There's many ways to screw up a company, there's only a handful of ways to do it. Right. And so the thing about investing, thinking about doing startups, like, you got to figure, what are the two or three fact patterns to do it. Right. So the first question I want to talk about, there's a whole bunch of stuff I want to ask about on Amazon is what do these companies do? Right. What do you look for when you join these companies? Was there like, they all had a blue logo? I mean, what was about the culture, the product that said or the mark said, like, okay, Markham, Tivoli, BEA VMware AWS.

Yeah. So the first thing I think that I do, the first, I guess there's one characteristic that's I think really important, and that is you have to be curious. So I'm always asking customers about what's interesting, what's new, what do you think about this? What do you think about that? And in almost every case, I found customers making a movement to those platforms early, and I found that early. And I can tell you about Clyde, who told me about Markham. I can tell you about Dan, who told me about Aw. And these guys, you know, the senior executives. You know, my specialty is enterprise senior executives that I'm constantly querying to see what's the next interesting thing. And so if you're always asking questions, you'll end up hearing the darndest answers.

Did you notice what was the early adopter? Verticals. Right. And. Or size? And did you find like these customers you talked to cluster around a vertical, like tech, financial services, retail. Do they cluster around size of customer? Big, small?

So I look for use case because what I'm looking for is an inefficiency in the market.

Yeah.

And so I, I talk to a lot of customers every day and they tell me a lot of interesting things. But there are only a few things that you come across from time to time that you go, oh, that's it.

Yeah.

Right. And so constantly being curious is critical. But then looking for the use cases, I like use cases that can emerge into a platform and I use as kind of a definition that I use, is that they don't have a Gartner quadrant yet. Right. If they've got a Gartner quadrant, it's too late.

Yeah. You've been defined by somebody else.

Correct. And so you look for something that hasn't been contextualized by the market, and then you look for significant pain point. Right.

Yeah.

VMware is like falling off a log to see that one. Because the average utilization of an x86 server was 3 to 5%. And you know, virtualization had been around for decades. I sold it on mainframes back in the mid-80s. And so putting those two together was an easy connect once you found the technology. Right. And so it's those kind of connections that you look, look for that huge inefficiency in the marketplace.

It goes to show that I remember the early days of VMware. We had these issues with Microsoft, Oracle, supporting their products on a different platform. And you know, AWS had the same early. But your point about inefficiencies, if you create economic value for Your buyer and VMware was a 10 to 1 server consolidation, Amazon was so much of a cost saving for everybody out there. If you create economic value for your customers, you can't stop it. Right?

You can't stop it.

And I always say drivers and drags, there's reasons to use the product, drivers the drags, reasons not to use the product. Before drivers outweigh the drags, then you have something inside the tornado. Now if you look at cloud virtualization, app servers, management products, once you're inside these companies, Mike, because you've been in this rodeo four or five times, what's similar across these companies?

Every company is exactly the same on the inside. Exactly the same. I call it the tornado. And they're all from the type of people that you find you're working with, their motivation, the types of customers that you have. There's a journey for customers in every major technology cycle. I've been able to map a journey. I've written a white paper on how customer goes through exactly the same journey. And in every new technology innovation there's an adoption cycle and it starts in a certain place, then it moves to the next place. And it's all super consistent in almost every one of these major technologies.

Yeah, the people thing's interesting. I think you said this to the employees that get attracted to these tornadoes. They're in this blitzscaling or hyperscaling phase or all similar DNA. And oftentimes I characterize them like you have barbarians versus villagers. And the people that are in these companies in this hyperscale, they're quite frank. They're, they're barbarians, no offense, barbarians, because like they don't play by the rules, they're aggressive and they get shit done by sheer force of will.

But they're motivated differently, correct? Yeah, I mean it's religion.

Yeah.

They're not here for a job, they're not here to make a dollar or two. They're here to change the world. And you can see. And the reason, I mean, when I leave companies, it's when the religion leaves, when I sense the religion leaving, it's like it's time to go someplace else. Because where you want to be is around a whole group of people that fundamentally are changing the world. And they strongly believe that and they're going to drive towards that, right? They're not going to let the, you know, the naysayers get in their way. They're not going to be dissuaded by all of the no's because what they've got is something that's going to be hugely impactful to the marketplace and they're gonna go out and prove to the marketplace they're right.

So maybe I want to go back to that and talk about kind of the experience you're having now and being sort of the tornado that is Amazon. And this is an enterprise audience. So we're all keenly aware of Amazon as a friend and a foe and as a partner. But before we get there, I want to talk to some mechanics around building teams, building go to market, building channels, because we have a bunch of founders in the audience and you've done this several times. So may, let's talk about first kind of the go to market. Then I'd love to talk about how you hire your team when you started Amazon, we started vm, we started bae, the average deal sizes were small.

I remember well, I noticed that I was in his dead zone in almost every one of my tornadoes. But the interesting thing about the dead zone, super important is that at BEA Systems, 90%, 95% of the revenue started as a download self service. Now the average deal size was around 7k, but it was self service. So it was really hugging very tightly on the self service. It was more highly indexed on self service than lowly indexed on revenue on the direct sell side. And you'll find that the development community is most attracted first to these really interesting new technologies, right? And whether, you know, whether it was, you know, BEA or VMware or AWS, the development community immediately saw how empowering this technology was and they flocked to it. And the development community doesn't need to be handheld, right?

Just we're talking to the audience developers.

So yes, let me see the technology, then get out of my way and let me leverage it, right? And so what you find in these cycles is that you want to play in a community where the developers can get access to the technology in a way that is very self service. And then can create something super interesting off of that, something that's business value, and then from there launch into greater use cases and the likes.

How do you walk that journey? So we talked about, you start in the dead zone, you're selling 7,000 deal sizes. Early days, VMware was like five or 10.

God, it was the exact same 10,000.

20,000 is a big deal for us back then. How do you, how do you scale them up?

I remember the first $50,000 deal, champagne.

Bottles were being popped.

High five. Look at that. Wow.

Yeah. So how did you scale the $50,000 deals? $10,000 deals. So quite frankly, you're doing 10, 20, $15 million deals.

So this is a really important characteristic of new technology adoption. What you need to do is you need to start a lot of fires.

Okay?

So Those, you know, seven to $10,000 deals are projects. That's what they are, is projects. You know, you got a developer that's got something interesting that he wants to or she wants to deliver. They start a project and they spend that much money. And then what you want to do is get as many of those projects started as you can. And how you get the big deals as you begin to aggregate projects into more, bigger and bigger, more sophisticated deployments. And so in every single case, the way it all started was lighting up as many projects as you can in as many accounts as you can, and then everything else takes care of itself.

How do you manage that as executive? How do you create a system to encourage many fires? How do you track it? How do you nurture it? Because it's one thing to say it.

But practically this is the most important discipline in your go to market. And by the way, I strongly believe that you have to make intentional choices, not accidental choices. And one of the most important things you can do is intentionally choose so that you can iterate on that choice. Well, so the core part of the go to market is in the selling motion, okay? And the selling motion has to be super consistent. Super consistent. You can't iterate on your selling motion if everybody's doing something different.

Define that, because that's both. You say it's a use case, it's a Persona you're talking to.

What I'm going to do is I'm going to come at it from a business process. So inside of your company, your go to market is made up of a series of sales teams or partner sales teams. And they're doing two things. They're either prospecting or they're selling. And oftentimes the sales teams mix those two up. They start prospecting when they should be selling, or they start selling when they should be prospecting. And they're very different things. And so the first thing you have to do is line up the notion of prospecting, what prospecting means. Each of these two process can be broke down. In this series, it's steps. And there are five steps to the prospecting cycle and there's about six steps so you can break those out much further, more fine grained if you want, in the selling cycle. And so you got to make sure that you spend enough time prospecting and enough time selling. It's the balance of the two that lights all the fires.

Okay?

You over index on one or the other, you're not going to deliver on your, your business.

Is there a rule of thumb on how you guide the split between prospecting versus selling? It varies by customer, by market.

So it depends on the first thing you have to do is early, in the early stages, you've got to settle in a standard process.

Okay?

Right. You get your first five customers. In those first five customers, you get to sort through how you went about creating, identifying those customers prospecting, and then how do you went about defending your value selling, right? And so you know, the prospecting cycle starts with who do I target, what company, what person in the company do I target? Right. How do I get access to them? Right. How when I pick up the phone, will they pick up the phone? When I send them an email, will they send me an email back? Right. And there's, there's a whole methodology around what I call making contact. And then once you make contact, the next thing you have to do is you have to earn the right to have a meaningful dialogue. You get 30 seconds to three minutes when you make contact to defend your value, to earn the right to have that meaningful dialogue. And then in that meaningful dialogue, you get 30 minutes to prove that they need to make a big investment in this technology. And that's all in the prospecting cycle. You haven't even started to sell yet, right? And so having that super consistent is critical because your sales force is just like a manufacturing plant. You have a certain amount of capacity and you have a certain amount of yield. And you have to manage both your capacity and your yield. Right? And your yield goes up by consistent selling motions. And what is also happening in the marketplace is the marketplace is evolving. So you need to understand that there is this notion of the Darwinianism of sells. If you're not evolving, other people are evolving past you in your selling motion. And you're not going to get the access that you need in those prospecting cycles if you aren't constantly evolving. And you can't evolve something that's not consistent. So you have to, in those first five customers you gather, then the next, you know, 45 together to hit that 50 customer mark, you have to really decide what is your prospecting motion that is the most efficient and what is the selling motion. And then you got to standardize your teams on that.

Are the teams. Is is prospecting historically was like, you know, inside sales reps or BDRs. And then you know, everybody or is.

It CEO should be prospecting. So I tell everybody, look, prospecting is the calisthenics of sales. It's how you stay in selling shape, right? And everybody should be prospecting. Everybody.

Yeah. I was talking to founders.

I found most founders are the best salespeople out there, right? The thing that people undervalue is there are moments of truth in every day. And those moments of truth can be critical to your business. And that's when you step onto an elevator and you see that client. That's when you walk down the street and you run into and you've got 30 seconds to make that connection. Now, prospecting is, it's a muscle memory sport, right? It's a muscle memory sport. And what people make this mistake all the time, right? They walk in and they say, I gotta go call my best prospect today. I'm going to prioritize them first. And that's the absolute wrong thing to do. Your best prospect you should prioritize is last, dead last. Because it's a muscle memory sport. If you're on your 10th call, you have practiced that pitch 10 times. And so your best pitch will be made to your best prospect. What you don't want is your worst pitch. You know, you're still warming up to your best prospect, right? And so you know, and you'll hit these moments of truth that if you're warmed up, if you're, you know, you've got your swing down, you're ready to go and you that moment of truth hits, you're going to be in shape.

To take it because you're probably getting a bunch of objections early from other customers that you've nailed. You're probably having similar conversations. It's kind of counterintuitive too, that your best practice prospect is last, but it forces you to always fill the funnel. Because if you only go for the good leads, whatever, like Len Guri would call it, right? You're ignoring the bad leads, which quite frankly, it's like eating your veggies or doing your exercises. You're not widening the top of the funnel. So in order to get to the prospect the good leads, you got to do the bad ones first. It's interesting because you never know when you find gems there. Maybe. What's super fascinating on the process, I want to go back to it, but you mentioned the right team in place, and a great question I want to now touch on is hiring people in this team. So you've hired, I know, thousands in sales execs and fire them. Maybe talk through the process of how do you identify what does a great sales exec look like and when do you know? When is it simple as missing the number, making the number, when do you fire them?

Yeah. So the first thing I would say is don't think about your selling team as a bunch of individuals. Think about it as an organism. Your selling team is a group of people out to accomplish this specific mission. So the first thing you shouldn't be thinking is, gee, I wonder how many sales reps I need to hire. And that's where people go first. And that's wrong thinking.

Yeah.

What you need to be thinking is what is the makeup of a team that can prove the value of my solution? Right. And so I have a notion that I call a minimum deployable unit. What is the minimum deployable unit and what is it made up of? So at VMware, my minimum deployable unit is I needed a channel person because we were heavily channel focused. You needed a rep and you needed an se. And so I would only think about adding resource in minimum deployable units. I still think that way today. Right. Because if you add a rep, but you don't add des, then you're not going to be able to actually make the cell. And adding the rep was a waste of resource. Right. So you got to figure out what your minimum deployable unit is first and then think about adding resource based off number of minimum deployable units. And then within that minimum deployable unit, you have to decide what's the makeup and what's the, you know, not only the skill set, but the psychographics of that. Right. And so you may be looking for, you know, depending upon the cycle. I often use this analogy. You've got to recognize what you're trying to do on the field. There are people that can run the 40 and 4 3, and there are people that are 350 pounds. There's very few people that are 350 pounds and can run 4 3. And it's the same thing with selling. It's the exact same thing. It's a muscle memory sport. So you got to understand what it is you're trying to accomplish. Do you want the guys that are going to block out on the field right now? Well, let's go hire the 350 pound guys or do you need the receivers that can get down the field in a hurry? Then let's go hire the four, three guys. Right. And you'll recognize that as you evolve your go to market that you're going to need to bring on different teams and the makeup of that minimum deployable unit will shift over time.

When you think about the team, the big 350 pound lineman or the sprinter. A question we had this morning in hiring was domain experience versus functional experience or experience. So if you're a security company, storage company, CRM company, those are different domains. And sometimes your boss like oh, I need someone out of Salesforce, I need someone out of VMware, I need someone out of Oracle because I'm selling databases. When you hire reps or executives or ses, how much does demand experience matter versus like this person is the.

I gotta tell you, I had hired Curry to do anything.

Yeah, Steph Curry, probably.

I'm not worried about domain experience because I guarantee you that if I put him out on the football field, he'll do a heck of a job for me, you know, if I give him a freaking battle. Better hit that ball. Right. So you got to recognize that you're looking for raw talent.

Okay?

Right. And even the. Really, I mean there's a certain requirement of your talent to be able to consume the technology that you're delivering. Right. So there's. You got to make sure that the folks you're hiring have the capacity to understand the value proposition. But you can train people on domain, but it's hard to train people to shoot a three pointer like he shoots a three pointer. That stuff, some of that stuff just comes naturally. So I would always choose the athlete, you know, the person that's really technically competent or really can simplify a conversation to have, you know, a really meaningful conversation to a customer over domain expertise.

Even though the domain person might have a book of business or contacts in the field, you take the athlete with the personality first.

I would. So there's a bump that you get, but it's an unnatural bump associated with hiring a. They call it the golden Rolodex, but it's an unnatural bump. It's you're relying on something that's not long term sustainable because there's, there's a capacity to that Rolodex and so you get that benefit of that and then it's over expires.

There's expertise.

So you, what you really want is you want a really repeatable process that can get custom, you know, that these people can use to get customers engaged today, six months from today, 12 months from today, 24 months from today. Because the biggest cost associated with Salesforce is a restart of that minimum deployable unit. Right. It costs you time in market, it costs you a ramp up time for the new team that comes in. And so for me I index high on what's their ability and index lower on the golden Rolodex. Although I'll take a golden Rolodex all day long if it comes along.

Things like golden LinkedIn account or something these days. But yeah, yeah, so you mentioned something that's interesting is like the right aptitude at a field rep, either inside whatever can learn the domain, can learn how to sell the value, which is powerful. And you've done this across multiple technical companies. How do you like to work with technical founders or the engineering, the product side? Because we've been able to make this kind of business side and technical side marriage work several times and in the audience we have mostly technical founders that come from the engineering background, product background and you know, TV shows or movies aside. There's people think there's this divide between sales and engineering. What are some tips or suggestions? How can technical founders work well with their sales organization?

Well, so the first thing I'd say recognize that the sales organization needs you for prospecting. They need you for prospecting. They need to be calling on customers that have never looked at the technology before. They'll learn a ton from that. And so I highly recommend you get involved in prospecting cycles with your teams. You know, let them cold call and you're there along having the conversation with a customer that had never heard of you before. Not the most comfortable thing to do, but hugely valuable. I've always been a strong believer in, you know, the minimum viable product and then iterate quickly.

Yeah.

And you have to have what I call a strong voice of the customer in that model and hearing directly from the customers and prospects. Super important. The second thing is, you know I talked about there's five steps to the prospecting cycle and six steps to the sell cycle. Right. So whether it's the qualification phase that sales reps get right, but as you get into the technical validation phase of a sales cycle. You know, you should be deeply involved with the technical validation and know how customers are deciding whether this technology is relevant for their pain point. And there are various steps in the technical validation phase, everything from POCS demos. There's a bunch of steps in that technical validation stage that you guys should own, understand and constantly be getting feedback on. Yeah, we did a bunch of demos. None of them came across. Why didn't they come across? Are we off message or are we off product? Right. There's a ton to learn in that part of the selling cycle, and I would highly recommend you be deeply involved with that.

When you think about that, I mean, it's hard to boil down kind of best practices in a short fireside chat, but what mistakes or traps do you see? And all throughout, one is selective hearing. Right. A lot of founds log exactly the.

You know, we call it happy years.

Happy years. Right. You know, I, I've done this mental. We're sitting in the same meeting with the same customer, same cio. We come out with two different set of notes on priorities. Because you're listening for what you want to hear. I'd be curious, what other traps do you see when you're. When you're doing this prospecting? And do you have tricks or ways to kind of neutralize those biases?

Yeah. So the way I think about it, again, coming back to the manufacturing metaphor, it's your capacity, and you get to use your capacity in the most efficient way. And the biggest challenge in or I'd consider a waste of capacity is calling on people that aren't going to buy anything. I have a notion that I teach the team is what you want to do is only sell to the people that are going to buy. I call it fish where the fish are. Let's go fishing where there are fish. Let's don't fish where there aren't any fish. The qualification of the prospecting cycle is where people make the most mistakes, you waste the most. The largest waste of selling capacity is because we're not targeting the right people, the people that this solution really fits. Right. We're targeting people that will answer the phone. We don't want to target people that answer the phone. We want to target people that will buy the product.

You get adverse selection that way.

And so highly targeted selling is core to protecting selling capacity, then the second part of that is to be able to qualify whether or not the customer. So if you get the targeting right, which is what will protect the bulk of your capacity we're only going to go call on people we know we're going to buy. And you do pattern matching to get to that. Right. You got to get in that first 50 customers, you'll do pattern matching and you'll recognize a series of patterns that you can begin to target with. And then it's about how to defend the value. And so people aren't going to buy just because they like you. Right. People aren't going to buy because they think it's cool technology. People you're going to build a business around are those people that are getting business value from your offering. And so what that means is you've got to recognize some critical. You call it. I call it the headwinds, but you call it something else around. The things that keep people from buying drags.

You call it drivers and drags.

Headwinds.

Everything has to have a framework. For me, drivers and drags.

Yeah. So for me, I have this notion that I call selling triggers. And there are critical selling triggers in every sell cycle. For example, virtualization. The critical selling trigger for virtualization was server refresh. Why? Because somebody's not going to buy something new, put it on the floor if they can get rid of something. Yeah, right. And nobody's going to write off that server. They're going to fully depreciate that server. Nobody's got extra expense sitting around on the balance sheet to just write a bunch of shit off. They're not going to do it. Right. So what you have to do is you got to recognize there's a selling trigger here. So in order for me to sell virtualization, I need to know the refresh rate and the refresh timing. And if I come in and I know that the refresh for this set of servers is Q3 of 2017. Two things I can do. First thing is I can know that there are no fish in that pond right now. Quit phishing there. Right. The second thing I know I can do is, you know, one quarter before I can engage with a good conversation about the value of getting rid of that server and not replacing, but. But instead having a VMware solution in. And so that's the notion of triggers, and there's associated that with that is the notion of selling intersections. So you have a number of triggers. And triggers are server refresh triggers are migrating data, triggers are new security or compliance regimes. There's a bunch of triggers, and then selling intersection is when they have to be delivered. So Q3 of 2017 is the selling intersection. And the two of those make up your ability to target and deliver a customer in the most efficient way possible.

Interesting. Maybe I want to migrate the conversation. Talk a little about aws because you have an audience that's very curious about it. And talk a little about.

And thank you all for AWS customers.

Raise your hand if you're on Amazon or Amazon customer.

I'm absolutely delighted to have you guys as customers. So thank you very much.

Raise your hand if you're using a different cloud beside Amazon.

Maybe can we talk a little later?

Maybe in addition to. Not exclusively, but there you go. Oh, yeah, there's Matt, as you know. So Amazon in my mind is I said this years ago. The biggest storage company started the past 10 years is S3.

Right.

The largest server company the past five years is EC2. And quite frankly now the large security company started in the past few years is Amazon as well. And it's fascinating to watch, kind of what's happened there. So first question to you is, are we looking at a world where there's like one or two clouds and that's it? Is it Amazon and Microsoft and maybe Google? Do you think it's a winner take all market or do you think there's plenty of clouds for everybody?

Well, it is a big market to be on. It's a huge market. This is probably a $4 trillion space. There's never been a market that is this horizontal, I think, in the history of it. So this is a really, really big space. Very unique. And it's a unique time in history where we're going through a technology transformation that is unprecedented. And I think it's the biggest one that we'll all see and maybe even our kids will ever see. It's such a big transition. So there'll be multiple players. There will be multiple players. I know that Gartner thinks that there's only going to end up with three. I mean, I saw them, they were at, I had a CIO conference the other day and they said, you know what, I think this thing narrows down to about three. I don't know if that's true, but I do know that the fundamental economics of cloud, all of the technology responds well to economies of scale. That's true. And I do know that if you're focused like we are on taking cost out, that's a core component of Amazon is we have what we call a virtuous cycle, which is aggregate volume drive cost out and relentlessly give that cost back to customers in the form of price reductions. Right. We've reduced our price 50 times in the last six years with absolutely no competitive pressure to do so. Virtually none. We're doing that because we recognize that that will drive more volume and if we can drive more volume, we can drive more efficiency. And Jeff's always talked about this flywheel that you get set up. And so that's what drives us. We're building a long term business here. We're building a business that we hope outlasts all of us. So driving that volume is going to be critical long term.

So do that two ways. On the bottom end, you've dropped price like 50 times the past couple years, which has been great. And we're thankful for you. And as investors, we're thankful for that cost reduction. At the same time, you've launched on a 400 new features or services.

722.

722 last year alone, right?

Yes. 516 the year before, 281 the year before that.

So we're continuing to always be prospecting. You know, there you go.

And I think there's some, probably good, some features there for you if we talk later.

So if you think, if you think about that, what's the best way for companies, let's say if your app company building something, Amazon, take advantage of the databases, Aurora Redshift, et cetera. If you were selling some like monitoring security or storage, what's the best way to work with Amazon? Right? Because from the outside we're like, holy crap, okay, I know Jerry, he can call Mike, but you know, I'm not going to call Mike for every little thing. What's the best way to work with you?

I've got three programs that I've set up. I've got an ISV program and I've got a team of hundreds that all they do is go out and work with ISVs, that's SaaS, providers, people that have applications that they wanna just deploy on AWS as a platform and you can work with them to get into our partner program and that'll give you a lot of interesting things. You can get access to something that we call a test drive, which is, allows your prospects to deploy your software on AWS and trial it and then decide if they like it or not. You can get access to a lot of things that can be accelerants in that selling cycle, the third phase of that selling cycle. See, you should do that. We also have a bunch of partners that are sis, that are there to help deploy this technology into big enterprises and small enterprises, but enterprises in general. And so you can get connected in with the big sis that are delivering solutions out there on behalf of our customers. And then you can get help from our, what we call our solution architects, the guys that, whose job it is to have hands on keyboards. So AWS has a different philosophy than any other IT company that I've ever worked for in that we hire really technically competent solution architects. Instead of say a systems engineer, it's somebody that can come in and sit right beside you and say, let's write some of this code together and let me show you how I would architect it. So then you can get some really deep expertise from my team. And we're delighted to do it. We've got a loft here in town if you want to stop by the loft at any point in time. We hold office hours there and our essays are there and you can come in and you can ask any questions you want there. But if you got something that you feel like you need, you know, Jerry's got a team assigned to him. He can connect my VC team into you. So there's a lot of paths forward just to get technical help as well.

Let me ask you. So that's on. So there's help on selling to customers, there's help on tactical solutions to make sure things work together. There's education stuff because you're launching 722 new features last year alone to understand what's new, because the playgrounds kind of semi facetiously, how do you want us to compete with Amazon? Right? Because there's a bunch of startups out here that either both benefit being on Amazon, but quite frankly may offer technology solutions that compete with some of the things that you offer. And I know that you probably just want some more Amazon, even though it means maybe not using Redshift, maybe using something else, what's the best way to communicate with.

So we have a completely different philosophy than you might imagine.

Okay.

On this. So, and this goes all the way back to Jeff in a decision probably in 99, where we learned something really important. So Jeff has this mantra that he talks about strategic focus and Amazon, Strategic focus is to be the most customer obsessed company in the world. And so everything we do is to achieve that goal. There's nobody that's asking about being the most profitable, there's nobody that's asking about being the most successful this quarter. None of that stuff matters if you're the most customer obsessed. Right. And so early on in the development of Amazon, Amazon had its own products. They were selling from their warehouses, but they also started offering Things from everybody else's warehouse. You could just come in if you had some stuff, you could post it and you could sell it through Amazon. And they kept the websites separate because some of the stuff, you know, Amazon had already paid for sitting in their warehouses and some of the stuff it would be like they were building, completely competing. And he said, you know what, this needs to be changed. If we're truly customer obsessed, what we're doing is we're helping customers make decisions. We're not trying to sell something to customers, we're helping them make decisions. And so that's where we come from. We're delighted if you provide something that is completely competitive to us. In fact, we have our EMR product and we've got, you know, some of our biggest customers are Hadoop based Cloudera.

Customers working on Amazon or Doc.

And we're delighted to support all of that because we feel like it's our job to help a customer make the decision. And if it's partially aws, partially our partners, we're delighted in the end to have that. Because you won't build a long term relationship with a customer if you come in and you're trying to force something on the customer that doesn't satisfy their needs.

It's classic. Customers don't want to be sold to. Right. They want to make decisions. I'm going to turn on to questions from the audience. I have a million other questions for Mike if no one does, but I get the sense for probably a bunch of questions in the audience.

Hey Mike, thank you for the Insight. I'm ex VMware too.

So at VMware we often talk about.

A number about like, you know, every VMware sales, there is a $1 sales, there is a $8, $10 drag of the ecosystem, storage or whatnot, you name it. What's the equivalent picture on the AWS side? I'm curious, you know, what's your analysis? What's the insight on that? You know, every time a customer chooses to implement EC2, there's a lot of stuff that goes around that I haven't actually haven't done the analysis. But they have to deploy software on top of that because they're not going to spin up a server to do nothing with it. So there's got to be something that runs on top of that. There's oftentimes somebody that's there supporting that. So there's probably an SI that's implementing on top of that there's also a new and emerging market around managed SO services and so There's a bunch of our partners that have stood up managed services where they run those systems on behalf of the customers. So we get a whole new kind of born in the cloud movement around managed services in these. I haven't actually looked at VMware. They claimed it was between 1 to 10 and 1 to 12. I don't think it's that high here, but it's certainly a ratio because every cell there's enough.

It could be via subscription versus license. Right. So you think of VMware's 1 to 12, a lot of the 12 was CapEx storage for example.

Yep, that's right.

And so in Amazon might be one of five, but the five is recurrent. Subscription could be of the lifetime longer. Interesting other questions.

Typically with the paradox of choice, pricing is a critical vector. So there could be a situation where Amazon competes with an existing player on the marketplace, severely undercuts on pricing, which could lead to the death of that particular participant in the marketplace. So how do you maintain that gentle balance between providing value to the participant in the marketplace and also conducting business? Yeah, so for me, price isn't the driver, but business value is the driver. Right. And so if you're creating significant business value, then price isn't going to go away. And if you're creating unique business value, then you get a bigger piece, you can have a higher price. Right. And that's kind of the core tenet of the technology business that we, you know, as a, as a community run in. Now the reality is the cloud based business is, are run at high volume, low margin businesses. Right. That's just the nature of that part of the business. And so if you're planning to build infrastructure that's going to compete in a low margin way, you're going to be prepared to have low margins. But if you're going to add unique business value on top of that, then you're going to be able to price to that value. And, and that's the way it makes sense to me. If your technology you're deploying is still within that low margin stack, you gotta innovate above that and get some unique business value going. And you have to continue to innovate quickly. Right. That's the reason why this notion of minimum viable product, continuous customer listening systems and fast iteration, that's a recipe for success in this next technology transition, in my opinion. And if you're doing all three of those, you're going to be creating unique business value every day. And that's what's going to defend your business Model.

That's great.

So you made a comment about for.

New technology that you need to light a lot of fires. And I saw a lot of people.

Nodding and jotting that down.

And I was just curious as I've.

Been thinking about that more, there's a.

Lot of different ways you can do that.

And so I'd just be curious to.

Hear you extrapolate a little more about.

What you mean by it. Yeah. So what I really mean by in that context, a fire is an analogy to a project. And every project has a specific thing that's trying to get accomplished and a business value that they're expecting from it. Right. And so what you want to be able to do is contextualize a project which hopefully will be short running and defends the value quickly. Right. The faster the time to value, the better off you were. As an example, Bea's WebLogic, you could create a Java app inside of 30 days and you could be ready to deploy. Right. And so you went from gee, I wonder if this technology works to deploying a Java app in less than 30 days that allows you to have a very rapid turnover of your selling motion. But it's that project, a defined time frame, a very specific set of things you're trying to get accomplished and a recipe for accomplishing those that has a tight business value associated with it. Right.

Java app in less than 30 days. Just. Just that phrase compared to 30 seconds to spin up like an EC2 like instance launch some. It's amazing. Just within the span of a few years. But like Java, 30 days and you.

Know, we've just come out with a new technology called Lambda.

Yeah.

Which now spins up servers in milliseconds.

Yeah.

And you're charged by the millisecond with Lambda. It's a whole new generation of compute that's going to fundamentally change the world. Because no longer you committed making long term commitments to compute. Like whole hour long.

Sure. Right.

You can now make microsecond commitments to commute. It's just the world is moving so fast right now, it's stunning.

So we're laughing at java app in three days. In two years we laughing at Lambda app in 30 milliseconds. Oh God. Remember those days when it took you milliseconds? Interesting. Other questions. I don't know how much more time we have. Maybe sour. Oh, we're good. Two things. One, you never got to answer when you decide to fire a salesperson or.

How you decide, number one and number.

Two, how does one find someone like you early in Your career? Career.

Well, so the second question is you have to go out to a ranch in Idaho. That's where I came from. I don't know that I would cold call ranches in Idaho though. But what I do is I look at, you know, basic behaviors, right? What you want to do is you, you don't want to look at outcomes, focus on the outcomes. What you want to do is focus on the input, right? And so you look at the inputs and their ability to deliver on the inputs. And if they've got a unique ability to deliver on the inputs, then you have to recognize and they're not getting the outputs. There's one or two conclusions. One, you haven't given enough time or two, the processes that you're using are broken down. It could be that, you know, you're asking this 350 pound guy to go out for a path. Hopefully you didn't do that. If you did, you gotta recognize that and make a change to your process. But more often what happens is, you know, there are five stages of the prospecting cycle, right? You know, the people that are really good at targeting the thoughtful. Let me go figure out not only which segment, which company, but the person may not be the best at picking up the phone and getting past the secretary or getting through the voicemail to the customer. Those are two different skill sets. And so what you may do is you may break those five steps down and you look across them and they just may be deficient in one of those steps. And all you have to do is give them some coaching in that one step and you can save what could be a very valuable person. And so you think about each of those five steps as five different skills. Look at the characteristics of their delivery in each of the five skills compared to the average. Based upon that, coaching could end up saving that person. If it turns out they're deficient in too many skills and or they won't take coaching, that's when you have to say it's time to move on. But people oftentimes move too quickly or move for reasons that aren't data driven. And all that does is create churn.

One of the first SES I worked with was marvelous at mapping out the org of any large company. Like he had the psych ability to create this org chart, but to call, he was like, oh no, I don't do the calls. Sarah, in the back you talk about lambda and 3 milliseconds of commitment, but how do you talk to customers about the different commitment that is doing Lambda versus EC2 or S3 or services that kind of have an equivalent.

Right.

How do you talk to them about lock in?

The funny thing is there's a lot of talk in the market today about lock in, and I think it's probably because there are some people up and down the valley over the last 20 years that have taken advantage of their technology to force customers to do things that they simply don't find valuable. Right. I don't know a single customer that isn't looking to flee Oracle simply because of the licensing kinds of models that they're using. And they're forcing customers to do things that, you know, that aren't customer obsessed. Right. So those kinds of things have created this real high concern around lock in. Now, you put an application in a Docker container on aws, you can pick it up and put it right over there in your on prem or in any other, you know, cloud in a way that's dramatically simpler than anything that we're worried about in lock in today. So we've got this world that we've been, we're dealing with that has got us locked into things that we just don't like. But the world that we're going to, it's way more open by far. More open by far. Now there's going to be things that are unique and valuable that you're going to say, well, I'm going to commit to this just because there's no technology that delivers anything like this and I can create unique competitive advantage for my customers on this. And since it's the only thing that does it this way, I'm locked in. But there isn't any other choice, you know, and you're still going to build it because it will defend value for you in the marketplace. But the lock in story is, in my opinion, is kind of a hangover of the last 20 years of us not having, you know, containerization, microservice, all of the things that are really freeing us up to do things that we simply couldn't afford.

So Docker is the key to everything.

Apparently, according to Jerry. That's what he keeps telling me gets. What's their Docker?

Docker? One more question. Okay, then maybe my last question for you, Mike, is the following. What are the two or three things either technology solutions that you would like to see Silicon Valley build on Amazon? It could be the next great app. Like, you know, Benioff just put his Iot thing on your cloud, Right. What are two or three things?

Well, I get it. So there's a couple of mega trends that are super interesting to me right now. We're calling it IoT, but it's not really IoT because IoT doesn't really mean anything but what the world is exploding around sensors and there's going to be sensor data that's going to change the world. We're working with John Deere to put Iot on their tractors, and we're going to be able to space the seeds out so that how far apart each seed is and the depth of it based upon the soil correct characteristics of that square inch. And by doing that, we're going to improve productivity of agricultural land today by 20%, which will feed the 9 billion people that are on Earth in 2050 without breaking out any new land, by the way, we're going to protect our most valuable resource on Earth, fresh water, by minimizing the amount of fresh water we need. And we're going to reduce runoff, so it's going to end up reducing pollution, and it's all coming from IoT. We're going to end up in a world of wearables where we're working with a couple of major healthcare companies to help them build ICUs at home. Because if you're really sick, the last place you want to be is around a bunch of sick people, right? And with wearables at home, you can build an ICU in your own. In your own bedroom. And that's where you get better faster. Right. We're headed to precision medicine that will do deep sequencing, and we'll be able to track the early introduction of, you know, the four majors, right? Cancer, lung disease, diabetes, and heart disease. You'll be able to see that in the sequence way earlier and be able to address it. All of this is coming about from IoT, so that whole place is going to explode. Associated with IoT is going to be enormous amount of work around data. The amount of data that's going to be in the world tomorrow is going to be dramatically higher than it is today. And all of the tools that you need to be able to manage that data, they all have to mature. That's a whole new revolution that is just around the corner. All of our customers are seeking to migrate to the cloud. Any tool that helps them migrate to the cloud, you know, in a. In a reasonable fashion, is super valuable. Today I've got customers. I have a customer that I was talking to yesterday, one of the largest utilities in the world that's migrated 6,000 servers in five months. It's amazing, and it's just no GE stood on stage and said they're moving 9,000 servers. Capital One stood on stage and said they're moving all of theirs. Right. And so the reality is these enterprises are looking for tools that can facilitate kind of the next generation of IT and add value around that. So there's a few just megatrends occurring right now where the right investments with the right smart technology will end up delivering substantial value.

Amazing. Well, the last thing I want to say is thank you very much for the time, Mike. Mike is both a cowboy. He literally is a cowboy and he's also a philanthropist. He's got a foundation nonprofit at Stanford Hospital focused on cancer research. So if there's a way you guys want to get closer to Amazon and through Mike, that's a great way to do it because he and I share this interest around using technology for solving things like that. Anyway, Mike, thank you very much for tuning in.

Certainly my pleasure.