Peter Coffee  0:04  
The observation was made that when a Tesla drives down the street and it's machine vision detects a pothole and dodges around that's an impressive, stupid technology trick. But the fact that the next Tesla drives down the street doesn't need to see the pothole because it already knew about it. That's the kind of thing that a global computing fabric with immense resilience and scale can deliver without even breaking a sweat.

Jeff DeVerter  0:30  
Hi, this is Jeff DeVerter with Cloud Talk. Now, the voice you just heard was from Peter Coffee. Peter is the Vice President of Strategic Research over at Salesforce.com. Today's conversation, Peter and I discussed many aspects of work in the digital world, and the challenges of it in this COVID-19 reality. We even touch on what it might look like to eventually go back to the office or not. Before I give to much away, let's jump in. Remember, after this episode, we have some information for you, like always, in a preview of our next episode of Cloud Talk, so be sure to stick around.

Intro  1:08  
Welcome to Cloud Talk. Here's your host Jeff DeVerter.

Jeff DeVerter  1:25  
Well, hello, everybody. It's Jeff DeVerter with Cloud Talk again, and boy am I excited about our conversation today. Today I have with me, Peter Coffey, who is the VP of Strategic Research over at Salesforce. And today, we're going to talk about what does, and we're not going to say return to work, because that implies we've been sitting home watching Netflix the entire time. But what we're going to do is we're going to reopen the office. And what all of that actually means. So Peter, I'm so glad you're here. Why don't you introduce yourself, give us a little bit of background on who you are? What you're doing today and the like? So Peter, welcome.

Peter Coffee  2:01  
Thank you, Jeff, it's great to be with you today. I've been with Salesforce now for 13 years. I think I was somewhere around number 2000 at the company back in 07, when we were only doing sales support and had a nascent platform business. And I believe I may have been the first person in the company to have the word platform in my job title, which was originally Head of Platform Research. Now, of course, the Salesforce platform is the numerical majority of the machine cycles that we serve. In fact, the majority of the calls that are made to our backend are made through applications written by others. So we're really a platform company with very strong sales service and marketing business. In the same way that Microsoft is a platform company that also sells a little thing called Office. And understanding that relationship. understanding the nature of being, the most aggressive consumer of our platform is an important understanding for what the relationship of the platform business to our bread-and-butter of sales service, marketing analytics, increasingly b2c, and digital customer engagement is really what we're all about. So, I spend most of my time going out into the world and finding out what's happening in three different buckets of things that people didn't know Salesforce did. That's a communication challenge, or things that they think they need from it, that I have to explain to them are really just things that you used to need in an older operating your stuff, or the things that are just genuinely new transformation opportunities. Digital transformation is a wonderful phrase except for the noun and the attitude, because digital makes...

Jeff DeVerter  3:38  
If it weren't for that it'd be all good.

Peter Coffee  3:40  
Yeah, it makes it sound as if technology adoption is the hard part. When that's really essentially the putting on your running shoes to start running the race part. And the other problem with the word transformation is it implies there's an initial state and an end state. And once you get to the end state, you're good? Well, no, because the end state is a state of continuous transform-ability.

Jeff DeVerter  4:00  
That's right, I spent a lot of time talking to companies around when they think about moving to the cloud. Because there still are companies who really haven't, as far as they understand, done much in that context. And I keep telling him, it's not just a real estate change. You can't think of this as your VMware one to VMware four, or even bare metal to a virtualized environment, or operating system X operating system Y. Where you signed up a project team, you do that work, and then you don't have to think about it for a few years. You move into this, this environment of a truly digital workforce. And you are adopting a methodology that by the way, the very foundation you stand on is transforming underneath you, whether you realize you've got it or not.

Peter Coffee  4:38  
It's also a complete inversion of your relationship to data and IT. Because it used to be the data that mattered was the byproducts of your own activities. you've generated it, you owned it, and so you monetize it at your discretion or not. You could do things like leveling workload across your IT stack. And so typically people would have maybe two or three times the capacity that they needed to cover their average workload for the occasional spikes at the end of the quarter. And when most of the data that matters originates from outside your organization, originates in its own good time and not at your convenience. Now, today, you would have to own something like 1000 times your baseline capacity to handle those surges, which is why the elasticity of the cloud is now absolutely necessary. And why the whole relationship you have with how value is brought to data has to be completely transformed into a completely customer experience driven point of view instead of an operational efficiency point of view. 

Jeff DeVerter  5:44  
That's right. And in no time through this continuum of adoption, has that been more absolutely required than what we've experienced in the past three or so months.

Peter Coffee  5:53  
Well, that's again, the problem is that people think there was a transition to the cloud. In fact, it's an ongoing acceleration, and the acceleration is not over. Pat Gelsinger at VMware said back in 2017, "Try to remind yourself that what's happening right now is the slowest things will be changing for the rest of your career." Nothing is not still in an acceleration phase. And you can cope with high speed when you're sitting in a plane doing 600 miles an hour, the speed is not making itself apparent to you. But acceleration, change of direction, change of the velocity vector. That is what tests organizations, the ability to handle continuous acceleration. Or maybe even more to the point discontinuous and chaotic acceleration is the real test and why the scalability and elasticity of the cloud model are so much more important than the cost efficiencies.

Jeff DeVerter  6:47  
So much now is expected, and it's amazing how much has happened just in the past couple of months. In fact, we had a little call last week and getting ready for this. And I've quoted you like three times since then. And those three times, really around everything that's happening right now. It isn't a surprise, there were plans for this. We're just doing it in two months as opposed to two years.

Peter Coffee  7:05  
Well, Joe McKendrick said 2025 got here 16 months early. And so many of the CIOs I talked to immediately nod their heads. And their whole body language says, 'Oh, yeah, I'm taking binders off the shelf for stuff that I've been telling them for two years we needed to do'. And they've been telling me, 'Maybe in the next two years, we'll get that done.' And now they're saying to me, 'How quickly can you get that stood up? Can you get that done by September? Can you get that done by next week.' And for those who made an appropriate adoption of cloud not only as the physical facility, but also as the manner of operation and behavior of their IT over the last few years. They're able to give very good positive answers to that because for them, this is just turning the knob up to 11.

Jeff DeVerter  7:52  
That's right. And so when you go back to your previous comment around, it's the acceleration that you feel, I don't believe that that leadership is going to be in any fit of a capacity or desire to decelerate.

Peter Coffee  8:06  
Well, no, because there's a briefing deck that's making the rounds this week, Masayoshi Son over at SoftBank did a long discussion of some of the challenges they've been through. And there's this wonderful graphic he's got, of horses climbing up the hill, falling into the valley of Coronavirus. But then one of them grows wings, flies out and grows a unicorn horn. He's saying the next few unicorns are waiting to be born. But they're not going to rise by climbing out of the valley, they're gonna have to grow wings and fly out. And this is such a wonderful image to ask a company that feels not merely challenged, but maybe even existentially threatened. Their whole industry, now the people are saying they're saying they'll never come back. We'll come back is the wrong idea. Fly out of the valley and discover that you've become something better is the more exciting opportunity. Turning yourself from a restaurant brand that is a place where people come to eat in to a dining experience brand. That's just as happy to give you an amazing evening in your own home. These are opportunities that are waiting to be seized by those who are able to see past the limits of what they know how to do and rediscover why anyone cares that they did it. 

Jeff DeVerter  9:20  
That's right. Incredible. And so your title is Research. So, as we think about what does reopening an office look like? And we look at that through a few different paradigms. So first of all, it's right now, everybody in, even the companies that weren't comfortable with work-from-home, now that everyone is 100% working remotely. If we're all in a playing field, we've all been in the conference room where there's been a couple of folks working remotely and nobody's got they're making sure they're always looking at the camera. They've got their back to them. They're not considering them as first class citizens. And so, now we're all in this Zoom or Teams or WebEx-y type of an environment, cameras on, we're all there. So, what happens when part of the workforce goes back to the office, whatever offices remain. (Hate to be in commercial real estate right now.) We think about what that transition starts to look like? And how CEOs are going to start to deal with those things? Those that were more averse to people working remotely, who wanted to see people in the office, build the camaraderie. What are your thoughts in that space?

Peter Coffee  10:18  
I know a few people whose middle managers have been the obstacle to a more flexible work-from-home policy. Because they feel that if I can't see people in their seats, I don't know that they're working. At Salesforce, we have a thing called V2MOM Visions, for values, methods, obstacles, measures. It's a thing we do every year as a company and as individuals. Everyone in the company has a written statement that they agree on with their manager and review several times a year. That says, 'Look, this is why I think my job exists. These are the values that are going to constrain how I pursue it. But these are the methods I'm going to execute, the obstacles I might encounter and the measures. Crucially, the measures at which I want to be judged.' And it never occurred to me before a few months ago that, that framework made going to a work-from-home models so much more straightforward for us. Because we never had V2MOM in which someone said my measure is my number of hours at my desk, ever. What are the measurable outcomes you're going to achieve. And once that's agreed upon, whether you're doing it from home, or from an office, or ideally from a customer's venue. As far as I'm concerned, more time I'm spending on customer's premises than on ours the better. That I think is key. And so companies, that are late to this party of replacing standard top-down, 'I'll tell you, if you're doing a good job' management. And need to move to an, 'I'll tell you what I think my job is. And if you don't agree, tell me what I don't understand. But once we've agreed what measurable outcomes I'm supposed to produce. Get out of my way, and let me get my job done.' And companies that can really live that, are going to do quite well in this environment. And those that measure your accomplishment by 'The number of hours you spent where your manager could see you', are not gonna be able to attract and retain talent. Because now that people have discovered that the arguments against them working from home really didn't hold up. They're not going to work for companies that are out of date in this regard.

Jeff DeVerter  12:20  
That's right. So, I think about certain roles inside of it. So you and I, from a knowledge worker perspective, we can have and will continue to do our job, wherever we're always working. We're always observing worlds doing what we do. But when you think about a role, I'm thinking about my insurance company, I actually used to work for this company. And my brother still actually works over there as an enterprise architect. And one of their roles. Where these folks that we think of him as the customer service agent. They knew they were working, when they could see them, and they could watch the board. And they could see the amount of things that were going on. And they were so concerned also, with the security of that data, they had binders of reasons why that role happened in the office, and they had a lot of offices that did it. But that was what had to happen. Of course, they just like everybody else sent them all home. But with tonnes and tonnes of infrastructure so that they could go do that their own computers. They could still control the routers, so they had put the right mechanisms in place. But here was that one role that was sacrosanct that said, 'It has to be in the office'. And you know what? It didn't have to be in the office at all.

Peter Coffee  13:19  
No, it doesn't. Now, to your point, a lot of the things that are being discussed by people like you and me. Are people who are able to do these things from laptops from home. And this is among the reasons why I militate so aggressively against using the expression back-to-work. Because the only reason we're able to do this, there are people in the power plants, in the telco providers, and the public safety and first responders, and even the grocery stores and everything else, who are making it possible for the rest of us to enjoy this solution. That we don't even show up at the office anymore. They do. And so some of the most important conversations I'm having right now are with people in the field service space. Or doing things like rolling to install new router capacity, or rolling to make sure the electrical system still work. Because we have a huge responsibility to them. They're being forced to expose themselves to contact with people, and contact with otherwise hazardous environments as they always have. And so the ability to use IT to have them know what they're going to encounter before they get there. So, the truck never rolls without knowing what the job is going to be. So, that the first time it rolls, it rolls with people with the right tools, parts and certifications to do the job once. Instead of doing one reconnoitering trip and then another trip to actually do the work. Think about the safety, governance, risk management. And not to be mercenary about it cost reduction opportunities that can result from doubling or tripling the productivity of a field service team. And so much of that is about virtual reality or augmented reality. 

Jeff DeVerter  14:58  
Yeah. That before were science projects, I used to call them. So somebody would say, 'Hey, I can come to your field service organization, and I can improve them by X percent.' Usually, it wasn't a double digit number. But they were, it just sort of works. And we'll just increase subscribers and different reasons. Well, now that they're the only ones who can be out. And safety matters. And efficiency matters. All of these things just bubble up to the top, of some of the most important things that these companies can do, not just to beat a competitor, but to survive, quite literally to survive.

Peter Coffee  15:25  
Well, along the side effects of this is that being able to prove that you were exercising due diligence and the protection of your workforce and your customers is going to be really important going forward. And replacing systems in which people check a box on a paper form, with systems that actually report in the moment of engagement. Yes, this is what happened, this is who did it, this is how it was done. There are manifold benefits that this can have in every area: manufacturing, health care, retail, and, to your point, vendor assurance. Look at the things that can be done. Where now the customer walks around the car with their camera takes a couple of pictures. The metadata of where they were, what's in the background, some of that other stuff gets captured by the customer and reported in. But it's so necessary that these facilities be twentyfour seven, everywhere, and be able to handle the peak loads. If I have a 60 car pileup, I'd better not have my systems crash just when they're needed most. And so this is again, why the elasticity and scalability of these operations as well as their geographic pervasiveness are necessary. It's also why the move to 5G is so important. Because the biggest change from 4G to 5G is not the enormous increase in bandwidth. It's the transformation of the network from dumb connective tissue along computing and storage resources into an intelligent computing fabric, some of whose functions include, what we used to call, networking. That's really what 5G represents. The ability to push processing as far to the edge as possible. When you look at the supercomputer you're carrying, that looks like an iPhone. When you think about the computational capability that represents and being able to push that stuff out there. And also the other way around. The observation was made that when a Tesla drives down the street and it's machine vision detects a pothole and dodges around it. That's an impressive stupid technology trick. But the fact that the next Tesla's drive down the street, doesn't need to see the pothole because it already knew about it. That's the kind of thing that a global computing fabric with immense resilience and scale, can deliver without even breaking a sweat.

Jeff DeVerter  17:40  
And this is how most science fiction movies begin. It all started so altruistically.

Peter Coffee  17:46  
Science fiction, the good science fiction is really more about people and societies, catalyzed by technology. And the bad science fiction is a narrative of people using cool tools. The discovery of fire was interesting. People figuring out that you could make a business out of cooking for other people that came later. 

Jeff DeVerter  18:06  
Indeed. So as we start to think about what this return to the office looks like. We're gonna have to deal with stuff that we didn't have to deal with before. Because we, as were talking earlier, when we mentioned that this isn't like previous globally impacting things like a nine eleven, like a financial crisis. We're not going to see a defined end to this. We're going to see a cycle of things that are going to happen. And it's a thing that is. It's not like we had an earthquake, or a wildfire. Or I'm in Texas, so we think about hurricanes. We think now about communicable type diseases. So, we can say, 'Hey, the office opens at eight and we'll see you then.' When 500 people storm the elevator. I mean, these are challenges we've got to deal with.

Peter Coffee  18:47  
Now the picture of the pandemic tolerant office that focuses on the architectural aspects of people sitting six feet apart with plastic partitions, and so on. That's fine furniture. You've just mentioned the crucial question which is arrivals and departures, bottlenecks, and lobbies, and elevators. Scheduling visitors, warning visitors before arrival of precautions that may be in place. That's an IT capability. It's what we're all about with what we're doing under the work dot com brand. It's all about conductivity and computational capability being pushed out to the edge of the network as far as they can go. So the number of needs for physical access to a centralized office is minimized. And these are all IT capabilities. I believe in many situations. The people who are being charged with the reopening plan are from a more straightforward real estate oriented office design mindset. And yes, there are challenges there. But having a perfectly redesigned, pandemic tolerant office with boundaries that don't have the kind of control and workload management that you've just described, would not be a useful thing to do.

Jeff DeVerter  20:08  
Yeah. And then you take that to the farthest degree. Now you grab a large metropolitan area take a New York, or Chicago, or somewhere that relies on mass transit. That people are arriving to get on the transit. They are in motion. They are downtown. They are at elevators. They're at lunch stands. They are at... What types of controls? Think about the dance that's going to have to happen, even between companies in scheduling when people show up. Think about a block, a city block that would need to coordinate the entire arrival and departure of that workforce, or lunchtimes for that workforce.

Peter Coffee  20:43  
Well, this brings together several themes that you've already raised. The need for coordination among parties who have traditionally been able to operate independently. The question of how much personal data about my movements, and my level of previous virus exposure, may need to become part of the organization. And the choreography, of these systems. These are all things that need to be done thoughtfully. I can't remember if I used this expression already in our conversation. But someone said, 'Be very careful of the things that are being done hastily under the umbrella of urgency, because ad hoc can become bad hoc.'

Jeff DeVerter  21:22  
That's a good one. No, you didn't mention that. That's good.

Peter Coffee  21:24  
Okay, well, once these things become embedded as process, you may be dealing with the legacy of something you did in 48 hours. And you didn't think clearly about HIPAA implications or disadvantaged population, impacts and things. And we want to be sure we don't create a whole new stratum of unfortunate legacies of being frictionless and gravity of our operations. We could have been thinking just a little bit farther ahead.

Jeff DeVerter  21:52  
That's right, you got to think a few clicks farther down the road. So, in your estimation, when we think about. I think you've already just called one out? And what are the hard things we're going to have to do in the next few months? And that is, how do we not adopt something today, that becomes a super big challenge for us in the future. Whether that's a rights or privacy or even a process perspective. What are some of the other things that leaders really need to be thinking about in their organizations, for how to go back to an office? And what that's going to look like?

Peter Coffee  22:20  
The nature of the workplace going forward, is going to be an acceleration of trends that have already been underway for some time. And so I can find you existence proofs of anything I'm going to discuss here. The degree to which, we will expect that collaboration takes place across many channels simultaneously. I saw a really good flowchart the other day that was designed to help people avoid having too many meetings. Because when you think about it, among the side effects of everything being done digitally, is that I can now call the meeting on five minutes notice among people in five different cities. Where before I might have had about a week ahead. And now, the opportunity, 'Well, let's just go ahead and get Europe, North America and Asia on the phone together tomorrow.' Pick a time that's not impossible from a timezone point of view. But that doesn't mean that having a meeting was a good idea. And so having good systems for saying, 'Okay, could that be done by email?' Could that be done asynchronously? Does that actually require real time meeting exposure?' There are companies in which there's some very strong discipline that when you show up for the meeting, you'd better have done your pre-reading. Because the only purpose of the meeting is going to be negotiation and decision. And there's going to be no information sharing. And every time, I've rooted that idea around in any organization. I've been told, 'Yeah, but people don't really feel the pressure to have their data ready, or their arguments worked-out until they're going to be in front of their teams.' Well, I think that's a management challenge. I think the workplace of the future needs to be one in which your responsibilities to your coworkers start to be regarded with a little bit more rigor. Because now you're asking me to sit in the room with you breathing your air. And that's a new social imposition, that wasn't in the way before.

Jeff DeVerter  24:06  
You want to put my life at risk to sit in the meeting and listen to your unprepared presentation?

Peter Coffee  24:10  
Yeah, and let you read your slides to me. Which I should have just seen last night. Also, everything that we know about doing major live marketing events. Obviously, Salesforce is one of the tallest poles in that town, with Dreamforce being the world's largest single vendor event. We are already actively discussing how we don't merely provide something almost as good. Because that'd be very lame goal, in my opinion, but how we say, how can we transcend all of the annoying limitations of live events? Not how do we approximate what was good about them. But how do we go way beyond what was difficult about them? One of the routine problems with Dreamforce, for example, is anticipating the number of people who will show up for a session so that the appropriate sized room can be used. That's maybe not so much of a problem anymore. If it turns out I've got three times the number of people that want to show up. I turn the knob.

Jeff DeVerter  25:01  
Or even having to think about the limitation of the number of sessions that you can have. You had artificial constraints before based on the number of rooms that were in the facility.

Peter Coffee  25:10  
Yes. And now we'll be able to have multiple tracks. Fewer concerns about, 'Oh, these two things are both at the same time'. Because now by definition, anything you can show up for at live event, you will probably be able to see as a recording later on. You may even be able to have two events on your screen at the same time. And when they get to the boring part of one, you can divert your attention to the other. Maybe be watching the chat streams of both of them simultaneously, decide which one you want to participate in. It's going to be an environment that certainly rewards the people who are actually good at multitasking, as opposed to the rest, the larger number of people who think they're good at it.

Jeff DeVerter  25:45  
And really prove if multitasking is a viable solution for us at all. Or, if we actually can train an AI model to care about the things that we care about. And let us know when a certain speaker might be on, they mentioned, a certain topic that we care about becomes something I think would be really interesting to embed into the experience.

Peter Coffee  26:03  
The role of AI. And this is a really interesting one. Because so many people, unfortunately have spent the last five years becoming very excited about a piece of AI, namely machine learning and prediction based on past data. That at the moment may be dangerously good at predicting the wrong thing based on what used to happen. And AI is a very big umbrella, someone observed once that once something actually does something, we call it what it does, and it stops being called AI. And so AI is the residual label for that, which is, and this is a 1983 joke, that which is almost implemented. And once it actually does face recognition, once it actually does understanding of natural language once it actually does predictive analysis, we call it that. We don't call hammers, artificial-arm-extenders, we call them hammers. And in the same way, AI is a a power tool for a person, not a replacement for a person. So using AI well, in this environment is going to be asking it to look at a situation and make predictions without being constrained by our past experience and our assumptions based on what we've tried in the past. AI is good, at not having artificial preconceptions that prevented from considering possibilities. AI is good at suggesting things and then testing them. And we have to use it creatively. And not use it unfortunately, in the way that some companies have gotten very excited about using it, in the last few years, to automate what they've done before but do it more cheaply. And so the the new thing that AI is not supposed to be is automated inflexibility. 

Jeff DeVerter  27:48  
Very good. Excellent point. So kind of with that we're unfortunately hitting sort of the end of our time. Incredibly fascinating conversation, Peter. So if people want to get to know you a little bit more. Where your properties online? How can somebody go learn more about Peter Coffee?

Peter Coffee  28:04  
On Twitter, I'm just plain Peter Coffee, no punctuation, P E T E R CO F F E E. I tell people, Twitter's the logfile of my brain. If you want to know what I'm reading, and what I'm likely to talk about the next time I'm in front of your audience, just go ahead and watch my Twitter feed. And you'll have a pretty good warning. A lot of the things that I've had to share over the last few years are in my Diginomica archives of just Google Diginomica. And my name, you'll get by author page there. That's a lot of what I share there. And then unfortunately, if you just go to YouTube and type in my name, you'll see a lot of talks I've done.

Jeff DeVerter  28:37  
We'll find you there. Excellent. Well, thank you so much for spending the time with us today here on Cloud Talk as part of Rackspace Solve. Incredible conversation. Rackspace and Salesforce have an amazing relationship. And we appreciate that between the two of us. So, we're grateful that you were here today. Love to have you back in a quarters time. And we'll see how many more years we fast forwarded in our digital transformation journey here, since 2025 certainly did get here.

Peter Coffee  29:04  
Salesforce and Rackspace, we're in violent agreement on just about everything that really is important in the world right now. And it's a pleasure to have had the chance to collaborate with you today. Jeff, thank you to you and all of your colleagues.

Jeff DeVerter  29:14  
Absolutely. Thanks, Peter. Hey, everyone. Thanks for listening today to Cloud Talk. We appreciate you taking the time for it and new episodes coming soon. I am Jeff DeVerter, CTO over here at Rackspace. Have a great day.

Closing  29:26  
This has been Cloud Talk, you can find cloud talk wherever you find your favorite podcasts. And be sure to check out more content from Rackspace Solve at Solve dot Rackspace dot com. 

Jeff DeVerter  29:42  
Peter Coffee. What an amazing thinker. Such a unique technology minded viewpoint of the world. Make sure you follow him over on Twitter at Peter coffee. Really appreciate him taking time to share a little with us here at Cloud talk. And now it's time to kill two birds with one stone. You see our next next episode is with the CTO of Rackspace Technology's EMEA region. He's also a contributor and a curator of content here on Solve dot Rackspace dot com. It's CTO calibre folks like Lee, who are responsible for creating and curating the content that we make available for you. We spend a lot of time with our editors to provide valuable, timely content, all focused on helping you in your company be successful. Now, usually, there's more than one path to success. Your path will be unique to your needs and your organization's culture. That's why we've named this hub Solve. It's here to help you identify how each piece of your ideal solution should fit together. Now, I've already told you that Lee James is the guest of our next episode, where we'll talk about of all things, the game Fortnite. And how it's a model for technology team structures of the future, check out this preview.

Lee James  31:04  
One of our customers today, they have a great analogy. Which is, 'Think small and you'll achieve big things'. While it's counterintuitive, I think it's the small pod structures today, the gamification, etc, really leads to companies being a lot more nimble and innovative in terms of getting the best out of their employees too.

Jeff DeVerter  31:26  
And that's in our next episode of Cloud Talk.