Google Cloud NEXT '17 - News and Updates

Building stronger teams with team-based functionality (Google Cloud Next ’17)

NEXT '17
Rate this post
(Video Transcript)
[MUSIC PLAYING] RYAN TABONE: That was good. I got some woots and I got all right, just like straight. I'm glad that there's still some energy at the end of the second day after the keynote. My name is Ryan Tabone. I am Director of Product Management on G Suite. Thank you all for coming here today. I wanted this to be a little bit of a different session than other sessions. A lot of sessions will go deep on one product or go deep on something technically. This session is really meant to piece everything that you've been seeing together over the last year, year and a half, the announcements we had today, and where we're going in the future and give you a little bit of the thought process behind all of this, the data that we use, and just a little peek behind the curtain. First though, I'm going to go a little bit meta. I am going to for the next 30-ish minutes try and get whatever is in my head into all of your heads. That's what a presentation about, right? The speaker is trying to get the content that he has into the audience's mind.

This notion of trying to get what's in my head into all of your heads, it's part of something we like to call capturing and conveying thought. And really, it's what the entire productivity market is based on, trying to get you to be able to capture and convey thought more effectively. But yet, there's this irony of the situation because this is productivity market, which is trying to become more efficient, but there's no real metric, no real number that says, hey, you're doing this more efficiently. Actually by switching to these tools, you're actually making your company more efficient. We don't actually have that. There are tools that say, OK, this is how you're spending your time. In fact, there's also negative metrics, right? So I don't know if you've heard this before but a notification can break you from your flow. And it may take you 20, 30, or even 40 minutes to get back into the state of high throughput production. Now, that's interesting, but it doesn't really help you justify an ROI on productivity software or a switch to certain productivity software.

Now, this is where Google has a unique vantage point. We have some genuine insights here because we spend a lot of time and resources on this. How many of you are familiar with the "New York Times" study on Project Aristotle? OK, well, the people in the front row are going to have to help me inform everyone else. I will give you a high level here. Google did a study starting in 2012 where we actually studied how our teams work. And being a data driven company, we obsessed and analyzed everything that our teams did. And we learned a ton. And we realized that we actually had to take a lot of these processes to the enterprise. And it was the justification for us starting the Early Adopter Program, which some of you are probably familiar at this point because we launch a lot of our products that way. It started about a year ago. And the notion behind this is that if we have use cases that we're trying to solve, how do we make sure that we get them in front of customers and get the feedback in order to actually craft the future solutions?

And that's great from a qualitative perspective. But again, we're a data driven company. We also have the benefit of being a Cloud-native suite meaning that when we deploy a solution to all of you, we actually can see, hey, did the behavior changes that we intended actually happen. We can actually look at usage patterns and look at how our products are actually being used. Now, something interesting came out when we actually looked at this usage data, another tidbit that we didn't quite expect. And that is that most people spend their time not on the actual ideas that you have, right? We all have unique ideas. That is our value. But they don't spend most of their time on that. They spend most of their time learning and manipulating tools for rather mundane tasks. Let's walk through an example of this. Scheduling a meeting, I'm sure none of you are familiar with this process, right? Never had to schedule a meeting before in your life. So I'm going to teach you about how you schedule a meeting.

I got some smiles there. Yes, this is a process we all know and it's really painful. What happens? You ever try and schedule a meeting with I don't know 5, 10 people across different sites, maybe different time zones, try and find an available time that works for everyone, try and find a meeting room. Good luck with that. Then try and find the documents that you need for that discussion. Maybe, it's the meeting notes from last time. It's the presentation that you have. This can take several hours in some instances. And then what happens when the key people in the meeting cancel? You just got to go through that again. Now, this is just one example of a mundane task. There's tons in our everyday life that we probably just don't even think about, email management, formatting documents, finding files, expense reports. It's not hard to think about places where it's not that you're producing unique ideas. You're just doing the same tests on a different data set.

All right. So let's make this real. Going back to data again. This is going to be an audience participation point so just yell out here. When someone creates a document, what percent of time do you think most people spend actually capturing the idea they have relative to other processes in that document? AUDIENCE: 3%. RYAN TABONE: 3%. 16? I'm also part deaf. So you're going to have to project. What's that? AUDIENCE: 35? RYAN TABONE: 35%. Are we doing Price is Right rules? AUDIENCE: 0. RYAN TABONE: 0. AUDIENCE: $1. RYAN TABONE: $1. Yeah, I think we went full Price is Right. That wasn't what was supposed to happen. Well, I think 3% won because it's actually closer to 5%. The rest of it– so 5% spent on your actual ideation, the thing that's valuable, the whole thing that you're trying to convey. And the rest is spent on various forms of research, formatting, refinement. And this is not just unique to documents. I'm sure that as I'm talking about this, you're starting to do this analysis of yourself where you're like how much time am I spending on my actual thought versus just using tools?

It happens all the time. The time that– Oh, should I go back? I just saw a lot of phones go down. Yes? AUDIENCE: Yeah. RYAN TABONE: All right. I'm going to give you five seconds because we're on a clock here people. No, I'm kidding. It's fine. So this time, this time that people spend on learning and manipulating tools for mundane tasks, we call it overhead because of the obvious opportunity loss or opportunity cost. That's the word. But what if we did a better job of building tools, ones that were cognizant of what you needed when and maybe were even cognizant of who you work with and your larger organizational needs? Well, then maybe, just maybe, we could kill overhead. All right. Well, how do we do that? This is where we turn to machine learning and artificial intelligence, or ML and AI. And I'm sure this is a surprise, right? This is the first time you're hearing about these terms? No? Oh, you didn't hear it all this morning? How many of you skipped the keynote?

You've heard a lot about it already. And you'll hear a lot about it more within the industry. Now, machine learning for those of you who are not familiar with it at this point, it may sound intimidating. And the math is pretty involved, but the concept's pretty straightforward. Machine learning algorithms observe examples. And then they make predictions. Simple enough, right? Let me give you an example of how this has worked. So Gmail, a long time ago in a far, far away place, Gmail started studying spam using these machine learning algorithms. And they fed it this algorithm spam emails and non-spam emails. And guess what that algorithm did? When an incoming email would come in, it could predict whether it was a spam email or a non-spam. Layer on many years, and now we get to a point where, hopefully, if you're using Gmail, I'm sure everyone is in some form in this room, you rarely, if ever, see spam. But we're getting to a point where machine learning and artificial intelligence are making a lot more progress.

I'm sure you've seen this in the news that we are very proud as a company of having market leading machine learning and artificial intelligence. There's a lot of press about it. You probably saw how we were able to defeat the go champion. Why does that matter? It matters because it shows that we're now making progress fast enough at a point that we didn't even expect that we could start using machine learning to actually make an improvement directly in our products and help you and assist you in your experience as you're using it, not just in the background like in the case of spam. And we're in a position to actually use this more than others in the industry because we are a Cloud-native platform. So we have these tools sitting on these data centers. We have these products sitting on these data centers. We can actually integrate them more rapidly, make big bets because we don't have any legacy in governments. And then can again use that data, the data that I mentioned before to see how we iterate on that cycle.

Now, and the proof as I mentioned is in our products. It's not just that machine learning has improved the magic of Google Search, which it has. It also transformed Google Translate, transformed translate, changed Google Translate overnight. Now, of course, there was a lot of work that led to that point. But all of the sudden, overnight people couldn't tell the difference between what Translate had done and what a human translator had done. That's pretty incredible. And we're building this into our enterprise flows as well. Now, when we think about how to build this into our enterprise flows, we have to look at what flows exist. And it turns out that there's really two types of flows, the individual and the team. And if you don't solve for both, you don't actually improve efficiency in the enterprise. You need to solve for both. And they're related, but they're very different problems. Let's first look at the individual focused flows and see some of the improvements we've made there.

Smart reply, this feature was launched in inbox. And the idea was, well, let's see how people respond to a bunch of emails again that observe examples, make predictions. And then we could start making predictions about incoming emails, what you're likely to say. And what's interesting about this again, because we want to measure and see how people respond to it, for our power users, the ones who use this day in, day out, every day, they have the ability to save up to 14 minutes a day in typing. That's crazy to me. I still try and wrap my head around that. And Slides Explore. How many of you have ever had to make a presentation? I envy the people who did not raise their hands. And how much time did you spend just pushing pixels around trying to get things to format just to look just right to get your idea done correctly? Well, you're not the only ones. We're not all designers, so it takes a lot of time. And rather than talk through this slide, I'm actually going to go ahead and jump to a demo, if we can switch to the demo machine.

Here you have a sample deck with formatting. It will show up. We'll get there. So how's everyone doing? So we have this deck. Let me close this out. And you can see I spent a lot of work making a sample deck for you all because it's all lorem ipsum. And it's gone again. OK, so we're just going to have to like blink really fast and every fifth blink you'll see what I'm about to do. So I'm just going to create a new slide here. And I have a hypothesis that weather affects– am I spelling it wrong? Affects, I got it. OK, presentations. So if this presentation sucks, I'm blaming it on it being sunny out. And I'm going to hit Explore over here. Now, I don't actually have any data for this, which if you know me drives me insane. And then I'm going to say– I'm going to type in weather map, which you see I've already typed this in before. And I'm going to look for images. And I'm just going to pull an image over here because that one looks pretty.

It's got a lot of colors. And I'm going to type in presentation and that one looks. OK, done, right? How much time would you spend now trying to format this? So you got rid of those ugly bars. And you got it laid out correctly. And it would take a while. I don't even know what's going on behind me, so just pretend like it's working. How's everyone doing? So now, I'm just going to hit Explore again. Guess what? I just have a bunch of layouts that I can use instantly. What did I do? Absolutely nothing. This one I actually really like because not only did it crop and scale the images and got rid of the nastiness, it actually also used the– blank screen. I'm just going to cut all remaining demos. I'm just going to start telling horrible stand-up jokes and maybe that's me. Is it me? Someone in the back, magic person, if you want to come forward and fix my horribleness. That's fine too. What it actually did here is it actually formatted it to look like the template that I have for my company.

So now, I have custom colors. It actually recolored the images themselves, cropped and zoom them, and used the right fonts, the right layouts. So now, I look like a professional designer even though I did absolutely nothing. And that's the way it should be, right? You focus on your idea not on the formatting. Oh, forget it, just go back to the presentation. I'm not going there anymore. You can do it. Come on. If we all click our heels will it happen? Presentation? Maybe? OK, so the actual usage data here showed us that people who use this feature save about 50% of the time in making slides. It's crazy. Really, I honestly want the presentation. I'm not kidding. All right. I'll go to another one. I'm just going to talk to it. Oh, we got the big controller out. Drive quick access, so we talked a little bit about this today. You ever have that instance where you're about to go to this meeting and you need that one file with that piece of data because you– and you know you've seen it before and maybe even last time at this meeting you saw it and you want to just pull it back.

And you just– what folder was it in? It's horrible. Well, drive quick access uses a number of different signals, like your calendar, your email, what you did around this last time– I'm getting yelled at from the screen. It's amazing. I'm going to try and do this while talking to you. This is like my best magic trick ever. Oh, wait, hold on a second. I'm just going to detach this screen and hopefully that will fix it for you. We going to get back there? So what drive quick access does is it actually serves up the files for you based on these signals, being able to say, hey, this is what you're most likely to use. I don't really know what's going on. I think I'm just now going to describe with interpretive dance all the rest of my slides. I don't know what that is. Go for it. Yeah, thank you, appreciate it. OK, so drive quick access– and what that did is it actually also saved people 50% of their time. Now, we started saying, OK, we're throwing out these numbers, seconds, 50% of time, what does that actually mean?

I don't understand. OK, so we try to make it real. Let's say you take all of these three features, just these three features. And you say someone becomes a power user of these three features and uses them every day. How much time does it actually save them? Well, we find that it actually saves them two business– or can save them up to two business weeks per year, just for those three features. OK, here's another question for you. How many of you– I'm just going to here if it's possible. How many of you are spreadsheet ninjas? That's my interpretive dance again. All right. That's probably about right. Less than 30% of enterprise users actually feel comfortable manipulating formulas in a spreadsheet. And yet, spreadsheets are the most– OK, we're here. This wasn't– OK, sure. Spreadsheets are the most relied upon source of data or source of data in an enterprise. So now, how do you rectify those two statements, right? Most people don't know how to manipulate spreadsheets, but everyone uses them for their source of data.

Well, obviously, what's happening is that a lot of this load is being shifted to a small number of people. So we had this goal to democratize this analysis so that everyone can use it. And we called that Explore. So I'm going to– now, that I'm in here, I'm just going to show you this. So here's a wall of data that we commonly call a spreadsheet. And I'm going to click on Explore. And the resolution's going to be a little bit funky here. But I think you get the point, which is I didn't do anything. I don't even know how to use this spreadsheet. But automatically, I'm starting to get something over here on the side. Here's some analysis. Let me go into that a little bit. OK, I have a chart. It's showing me the link of cost of goods and profit. And there's actually this little text insight here, for every increase in $100 profit, cost of goods increases by $113. So it costs me more and more to make the same amount of profit. That's pretty interesting.

But that means someone could have just sent me this spreadsheet and I could just walk into a board meeting and just say, hey, did you know this? I didn't do one set of analysis. But what's interesting about this is that we're not trying to teach you formulas anymore, right? Instead, the spreadsheet app is trying to teach you about your data. And that's the goal. So I could make this easy. And I could just drop that chart in if I want. Now, that's all well and good, but really when people open up spreadsheets they're looking generally to answer a question. So in this case, the title of this, which I think is in the top left, Holiday Week Sales or Holiday Sales. Wrong, laptop. So let's say I wanted to understand more about my holiday sales. In fact, I wanted to just ask the question in this little section here called Answers where I will say, what were the top selling items on Black Friday? And I got a table back. And it said the answer, the top 30 items during 25th of November 2016 and there's a list with a count.

Oh, and there's also– this is my favorite part. And there's also a formula that I would have to write if I wanted to do it myself. I consider myself a semi spreadsheet ninja, maybe like a white belt. And I can't do that. So, wow, you just got me what I intended without actually having to do any work. I just asked the question and you got an answer back. I want to pause on this for a second. We did that so quick that people are like, eh, that's cool. What actually happened during this is that I typed in this question. We used our market leading, natural language processing to break down that question and say, OK, top selling items plural and Black Friday. OK, what does Black Friday mean? OK, it is a term that corresponds to a date. And what date do we actually mean? And, oh, by the way, that date changes every year. If I change that to 2015, it would know as well. And then it takes those pieces. And then it maps them against the data, creates this formula, and generates this answer, not bad.

This is what happens when you have market leading machine intelligence. And what we saved here is instead of someone sending a spreadsheet to an analyst with a list of questions and saying, hey, can you help me out with this? And then going through a few iterations as he prioritizes that work among the 10,000 other spreadsheets that he got. And you go through a few cycles. You may get back an answer at the end of that, but you definitely spent a lot of payroll. In this case, you make analysis again available instantly and for all. OK, can we maybe go back to the presentation again? Yes. Hey, look this is that quick access thing I was talking about. And then there's Sheets Explore. Wow, we're just– we're plowing through here. This is great. So having this market leading machine intelligence is what– or machine learning and artificial intelligence is what allows us to ship these features. And this is also the reason why answers in Sheets Explore and smart reply are unique in the market.

They haven't been able to be duplicated. OK, I mentioned before that in order for an enterprise to be efficient you have to be able to solve, not only for the individual, but for the team. And the team really is a completely different problem. Let's do a thought experiment. OK, think about how you work at home. How often is your work shared with others? AUDIENCE: Never. RYAN TABONE: Never. That was the answer I was going for, except some people will share it with their spouse. Around this time of year, the common answer is accountant. Other than that, yeah, it's pretty much solo work. OK, now think about how you work at work. How much of your success is dependent on your work alone? Any one? AUDIENCE: Very little. RYAN TABONE: Very little. I appreciate that. You kind of gave me a lob there. A lot of people say they realize that really none of it is all– even if you– I talked to a reporter yesterday. And I was like, well, do you– I mean your name's on it. This is you, right?

This is just– She's like, no, I can't do this without a whole team of people. Everyone I've ever talked to says they can't be successful without working with a team. And these teams are a mix of wonderful, hopefully wonderful, people with unique traits and skills. And that's why you're paired together with them because you each have different skills that you can work together to accomplish your task. But what does that actually look like when you're working with these other folks in this team? Does that look somewhat right? I'm sure most of you probably haven't mapped out how your teams work together, but probably this resonates a little bit with you. Meetings, action items, more meetings, balancing other projects, collaborating, scheduling out– oh I have to make sure that I get this done before I get something else done. It's not straightforward. Now, you see why the team problem is a different problem. Now, let me add overhead to each one of these pieces, right?

Before, we talked about meetings and adding overhead of just scheduling one meeting. How about the meeting, the milestone meeting, the sub meetings in between, and then within that everyone's going to have their different sets of tools to do their different tracks as well as be able to prioritize their time within those different tracks. And also I, need to track the action items between these different things. And, oh, where did I store those assets because so and so– it gets to be a mess. In fact, it gets so bad teams break because of this, right? I'm waiting on someone to get my job done. They haven't prioritized the thing accordingly. We haven't had a discussion about it because they can't find time in their calendar. I'm going to move on to something else. And then what happens? Well, now a project that could've taken a day is going to take two weeks because I'm sorry, I'm working on something else. I see some heads nodding. I'm sure none of you can relate to this at all.

All right. Well, the goal here is to not only remove that overhead, but just to really understand how people work together, to really understand their priorities, to start laying out their future for them, to make all of those steps simpler, and fit in with each other so that you work towards a common goal, and that we help you achieve that common goal on a common timeline. This allows you to focus on your task as a team, as a unit, as opposed to the overhead of the process of each individual within the unit. And when you do that, it makes an impact not just on you as an individual, but on your entire team and your entire organization. And this is why we relaunched Google Cloud and G Suite a few months ago. It's because we're focusing on how enterprises work together in teams. Now, this may seem like, oh, well, haven't we been solving this problem for a really long time? No. It's actually a relatively new concept for the market. And, yeah, there are a few point solutions here, but really to solve this, you need a suite working together.

Because as you see, there are many different flows where everyone's using different sets of tools. If they aren't all cognizant of the common goal, you're still going to have those overhead. And you're still going to break the process. So what are we doing within G Suite? Well, it started with the docs editors and real-time collaboration. That may have shifted the entire industry towards that model because they realize it helped teams work closer together. But we wanted to take it to the next level. And we made a few announcements today. For those of you who were at the keynote, consider this somewhat of a refresher. Jamboard, now this is the ability to actually look at a team space and combine virtual with real. I can be in one room, you can be in another room, and what generally happens is instead of working together on a white board, the people in one room work together on a white board and then we go, trust us, it's working. Trust us, we got this going. I'm going to take a picture and send it to you.

Sure, that's going to happen. Or, they write do not erase on it, right? So that when the person does– is in the same room, they can see it because that always works. But instead, now, I can actually join two virtual rooms into one real room where we can all work together and then resume this work later on. And we all get copies of it. Again, rethinking how teams work together. But we also went a step further to lay foundational ground. Now, this is actually pretty important because this sets a journey for us that goes on a ways. When teams work together they really need two core pieces. One is a source of truth and another is a way to communicate about that source of truth in order to work towards a goal. Make sense? Straightforward. This is actually like the complete background for a lot of our launches today. It was that thought process. OK, so let's go to the first one. Source of truth, team drives. The idea behind this is I don't want to think about storage. I have a team.

I want to know where my assets are. Who has ownership? Wait, do you see that file? I don't see that file. Is that there? No, all of that gone. I don't want to worry about permissions. It should just work. I should be able to say I'm going to Team Foo and all of my stuff for Team Foo is there. In addition, because we want to make teams as efficient as individuals, again, we're bringing quick access, which is that thing I showed you before that has 50% savings to teams as well. OK, now you have your source of truth, you want to communicate about your source of truth. When we actually talked to customers, we realized– and to teams, we realized that there's two forms of communication that happen on a source of truth. For those of you who saw the announcements, you know what's coming, but here's the rationale of how we got there. One is a continuous, asynchronous form of communication where constant progress can be made, right? Let's say I go to a meeting. And Bob and Sue on my team still are discussing the problem at hand, pulling relevant content from that source of truth.

They need to be able to have that discussion and move it forward. Now, if that happens in one-off conversations, I'm not going actually know. And I'm going to hope someone tells me. You actually need a way to capture what's going on and being able to reference back those materials as well as the decisions being made. And that was the reason why we launched Hangouts Chat together. This is a way to make teams work together and have one place where they can coexist and make constant decisions and discussions. OK, so now you have– and also be able to easily locate this content. So you see at the top a really quick way to be able to search and find that content. The second form of communication is discreet face-to-face high bandwidth. I need to get five people in the room. Text just isn't cutting it. We got somebody in Taiwan. We got somebody in Europe. I need to get them altogether right now. And it has to be simple, again, because we want to go and kill that overhead. Because if it takes 10 minutes to set up a meeting for 5 people, you just wasted 50 minutes not just 10 for yourself.

And that's why we launched Hangouts Meet. And the idea is two clicks and you're in, dial ins for those who can't make it on the VC, and best in class video and audio. All right. Now, that we have those pieces, we're going to build on them. We're going to layer in things like workflow, being able to predict what you need when, routing content, prioritizing and tracking tasks, automatically scheduling blocks of time. For those of you who saw the demo today with At Meet, that exactly what happened, right? That idea of calendaring now just went away. I was able to take five people in a room and just create an instant meeting for us. We can kill this overhead through these tools. And this is the vision to take world class intelligence and bake it into a suite of tools that work together because that's the only way that we can get teams to actually work efficiently on their different tasks. All right. Quick wrap up. Today, we showed productivity tools actually cost more time in overhead than actually helping you complete a task.

We can address those today for individuals using machine learning. And for teams, combining machine learning with a suite of tools is the only way that we will actually be able to address their needs. And we laid the foundational pieces with team drives, chat, and meet. Also, interpretive dance doesn't work for telling a presentation. But this is the start of a longer journey. And I encourage you that part of the reason why I'm here is hopefully you see where we're headed. And it makes you want to take your teams onto these products now so that as we iterate and build on top of this stack, your team gets the benefit without having to teach them something new because we'll be doing the work for them. [MUSIC PLAYING]


Read the video

How to use G Suite products and their inherent team-based functionality to make teams – not simply individuals – more efficient and productive.

Missed the conference? Watch all the talks here:
Watch more talks about Collaboration & Productivity here:

Leave a Comment

Your email address will not be published. Required fields are marked *

1Code.Blog - Your #1 Code Blog