I’m so glad you joined us because we have Phil Komarny and Dr. Ben Waber here. Phil is the Vice-President of Innovation at Salesforce.com. He’s working with blockchain technology and education. Dr. Ben Waber is the President and CEO at Humanyze and he’s known as the worldwide expert in people analytics, collaboration and wearable technology. This is going to be high tech, very interesting discussion with both of them and I hope you enjoy it because I’m looking forward to it.
Listen to the podcast here
Chainscript: Reinventing Education with Phil Komarny
I am here with Phil Komarny, who is the VP of Innovation at Salesforce.com. He’s the former Chief Digital Officer at the Institute for Transformational Learning at the University of Texas System. He has over a 25-year career and C-Suite roles that have led innovation, transformation and technology. It’s so exciting to have you here. Phil, welcome.
I appreciate that. Thanks so much.
I’m happy to have you here because I’m very interested in the work you do. You’re very into blockchain and technology that a lot of the people may not have a strong background as much as some other shows because this is more business-related. You can talk it down if you need to. I’ve had some blockchain experts on and they’ve been interesting. What fascinated me about your work is the focus on education because I’ve taught thousand online courses and I know all the issues with higher education. I’m curious about what you’re doing. You’re coming to Arizona to talk to a university here. Do you want to start with what you’re working on?
When I was at the University of Texas System, we were building a new stock of services. At the center of that system for services, fourteen institutions were a new way to address learning or higher education. Technology is limited the way they can address the business models. They want to attain complete subscription model for education, for instance, precluded by the technology that they’ve already implemented inside the university or inside the business to run the transactional day-to-day. What we were able to do there is back pitting or underpinning this new way to address the market with a new way to trust the data. We looked at the blockchain and built something called the chain script. It would allow not an A in English, but one of the marketable skills that came out of that learning environment that will be invaluable to a learner in these new color rules and not blue or white color. The new economy needs new color rules that have some type of certificate buying them.
It might be something lightweight like some small certificate or stocked up to a whole degree. That work can move forward with a very innovative work at Arizona State. They’re taking this mental up with Michael Crow’s big positioning around universal learning. We need a universal learning network that has trust built into it to help power this idea of universal learning that takes place over a lifetime instead of four years when you’re in an institution or your first twelve years of developmental primary and secondary school. It starts to create a new way to operate a new language. I’m calling it digital Rosetta stone between academia and industry, both to understand what we need to develop talent-wise and what are the signals that the industry can give us back to what’s needed in our workforce.
Right now, it’s very hard for both for the industry to single back the academia and them to react in time to be able to deliver those type of learning out to the industry. Trying to create this lingua franca between the two and being able to build systems on top of this trusted data and student data, being resonant with the learner makes it motivational for them. It makes them be able to find the next part of their learning. That type of unbounded system is what we’re working on now with Arizona State. We’re happy to work with them. We’re going to be able to welcome more people into that universal learning network, be it other schools or industry partners that want to help develop pallet into this new economy we have.
I’m a graduate of ASU and I talked to a lot of people about the future of education. We were talking about maybe à la carte learning where you don’t necessarily have formal degrees. There are certificate programs, I was working on those when I was working at Forbes School of Business and some of that type of thing. In universal learning, how is that different than people can test out of things and get past certain things? Is it just a way to keep track of all the universities or within one university and your life skills? How do you keep track of all that?
That’s part of this existing protocol and it’s due. Allowing not having the business with the university capture all of that learning and then be able to transmit the validation of that back out to the market through a transcript is very friction full. When we looked at just in the UT System alone, 68% of the students go to three or more universities in our system to get their degree over four years. We have a system that’s built to take that student through a journey in four years in one institution. That little bit of friction comes out to almost $132 million worth of friction when you look at just transfer credits in a system like that. Taking these new ways to operate and allowing this new methodology around trust and around the blockchain where we store that information in service of a person. The person can bring it forward instead of bringing their own device or BYOD. The BYOD 2.0 is going to bring your own data. That resonates data with someone that’s captured and trusted allows us to build new ways to bring that person into a learning environment and not just keep them there but keep them engaged.
It will keep them looking for the next thing that’s going to build them whatever their future state does. Trying to make it more like a Netflix model of that for education, wherever they can take the learning they need instead of this very large size degree that take four years to garner. To some people, that’s a great pathway. If you look at that in the market. Look at the product market set that we’re seeing between university presidents, 98% will say that these kids are already for the industry. When you talk to industry leaders, 11% of those degree students are ready for their roles. There’s a big product market fit question here and we could answer it in different ways. Having somebody bring their learning with them or a representation of a valid trusted representation of what they’ve been able to do in their academic life with them to the system will let that react to what they could do next or might want to do next. It’s what the future looks like and that’s what we’re constructing now with Arizona State.
The whole way that you look at four-year degrees is just going to be changing. As people are picking and choosing what they think they should take or shouldn’t take or what employers think they want, how is that going to impact? Just the glue that held things together that humanities and those things. A lot of the skills that they’re not maybe learning like soft skills where everybody’s getting fired for their soft skills. How are you going to track that and where does that fit in all this?
[bctt tweet=”In the digital space where we have damaged relationships, we prefer to be anonymous so we can just start over tomorrow.” via=”no”]The million-dollar question is how do we make what has inherently been the hardest thing to assess are the soft skills? Everybody’s asking for it. Economic forum says by 2022, the skills look exactly like what liberal arts education gives you and that’s what’s hard to assess. The way we do it electronically or digitally with what we do in computer education or things that are very easy to grade. The new ones that are where it’s super interesting with technology and AI that we have. We can start allowing some passes of that to happen with these new machine learning algorithms to look at the learning in a different way. Newer technologies will get us to a place where we’ll start to see that type of education being able to be assessed at the scale not in a way that we’ve seen to date.
That is the hardest thing to do. If we can find a way to signal to education what’s needed in our industries, we’re going to be able to build the curricula that are a lot easier to deliver in this very modular format. At Salesforce, we talk about the fourth industrial revolution and how it’s going to power all this automation and all this disruption is going to take place. We have over 200,000 customers we work with business level. These industries or these customers trust us to manage their whole businesses in our platform. What I’ve been talking about is how are we going to help them protect their number one investment and their people. How are we going to help them re-skill or unlearn and relearn in this future where the new definition of illiteracy isn’t about you can’t read or write in the 21st century? It’s about learning, unlearning and relearning. If you can’t do that pattern, you’re illiterate because this is going to change so much.
At Salesforce, we’re trying to deliver modular curricula through our system directly to these industries that are being so disrupted so they can learn in the context of their businesses. Re-skilling their own workforces with very modular educational offerings from great places like Northeastern Arizona State or Harvard, delivered right in the context of their business to help re-skill them. That’s what another path of education is going to look like. You’re in the near future as we start to see all this disruption taking place from automation. Trying to model something that could fit that role is what Arizona State’s up to. That very innovative university isn’t looking for something just for Arizona State either. Dr. Crow’s looking is for something that’s going to change the industry.
It’s going to leave a legacy. It’s going to show us how to empower our industries and our academic institutions to meet this call for the future. Right now, I don’t think we’re even structured in a way where we can meet it. We’re getting left behind by other countries around some of these forward-looking skills that we need to develop artificial intelligence, learning, and programming. These things are being done in China and India at a different rate and different level. We need to take this next step and change the way we operate here so we can keep up and build a bright future for everybody.
This is all ties into the stuff I’ve researched with curiosity and anything that’s how we started talking, to begin with. I’m very interested in how we’re going to reeducate people. Part of it starts with recognizing the things that hold people back from being curious and that’s what I was researching in that work. At technology and AI, a lot of the jobs are going to change. People need to be reassigned to different areas and find new things that interest them. By doing that, that’s going to be a big step. It’s interesting that you are so ahead of technology. This blockchain is so complicated and all that and yet soft skills. Some of these other things that seem squishier and easier are the harder things to work with. I find that ironic. The blockchain is very complicated for a lot of people to understand and I noticed you gave a little intro to one of the talks you were giving online. You were talking about the difference between private and public blockchains. A lot of people could benefit just from your knowledge. Some people don’t understand ledgers or even the basics of it. Can you give an overview of how you use blockchain in education and how is that different than what everybody in blockchain thinks Bitcoin sometimes?
I start there when I do talks. I call it crypto confusion between what Bitcoin is and what it was proven that we can solve the double spend problem with a cryptographic algorithm. That should have been kept in the lab. I don’t think that should have been ever left out of the lab because it’s inefficient. It cannot work at scale. It’s already consuming more energy than the country of Denmark on a daily basis just to run the Bitcoin network, which solves an important thing. Can we prove double spend without an intermediary? They proved that and that’s awesome and cool, but it takes the focus away from what this distributed ledger technology can do. We can prove double spending with Bitcoin and have this anonymous way to transact for whatever, which is super interesting. If we take that way that we can trust in the data, let’s keep it at that level because the technology is a bit complex.
Everybody on our trusted network, we call it the universal learning network. All of these different schools have a trusted version of that data they all believe in and they all trust. That information can now be resonant with their customer or their learner. That learner or that customer can bring it to the experience with them and it’s trust at that level. Being able to do that unbounds us from a login and a password. For every system we’ve ever created, we have a thousand of me in every one of these systems. From Facebook to Instagram to Twitter, there’s me. This gives us a way to collapse that into one record that’s mine that I can use in all of these different systems. A digital reputation like a true digital reputation where we’re not all just cats on the internet where we can create another account that I’m a different color cat the next day. This gives us a way to value the digital reputation. The power is a big part of our economy globally. When we started with this network, I got back from the celebration of the 70th anniversary of the signing of the Declaration of Human Rights and connecting 50% of the planet. Vint Cerf and Tim Berners-Lee were there. It’s so interesting to reflect on.
They started this thing that created this thing that powers most of our economies that never had a trust layer in it. We never thought about, “How are we going to trust each other in this?” Now we’re coming back to that. If they go back in time and look at that, we should have thought about a bit more if we’re going to use it for this much of our world. That data is starting to automate things for us. The fuel of these machines is coming from the people who are creating it. Trying to use this new protocol to allow people to value their digital selves like they value our analog selves. I’ve never met you in person, but we’ve been on the phone now and I want to make sure that I don’t damage our relationship by acting in a way that’s not nice. In digital space, we’ve all been able to be anonymous because we can just start over tomorrow. This gives us a way to value that digital reputation. We’ll start to see digital data not as about privacy. We see GDPR and all these things happening all over the planet and about how are we going to say this is privacy around the data. That’s gone.
We’re never going to get to a privacy state. There was no way to do it. What we can do is start to look at digital data as property and start to use law and things that we do have for people to value their digital selves as they do their analog selves. If we can do that, we have a whole new way to address it. We have a new way to value it. We have a new way to build systems on top of it and that’s the future. That would be differentiated compared to what we’re seeing coming out of China where it’s social profiling. They’re using this information to shape their social beings, which is super scary. If we can do it in a different way and have a digital reputation that we value just like real loved ones, we’re going to start to see a reason why people would want to keep their data. We want to see how it helps them in their lives. We’re just taking an academic path with this universal learning network because the data that powers the academic realm is something that’s not strategic to the business. If you look at finance, if you look at supply chains or healthcare, all that information that they store at companies are strategic to how they run them. That’s something that education doesn’t have. It’s a boat anchor for them.
They have to store, secure and share it. There was never anything positive. They don’t sell it. They don’t make money from it. It’s something that holds their businesses back a bit. That’s the first place I thought the blockchain would unlock some new potential that we would all be able to see because it’s not a business driver. It’s not strategic, there’s no reason to say no to it from the businesses side but there are so many value propositions from the customer side or from the learner side. That’s the one realm I thought would be the first to see them. I’m so glad Arizona State sees it that way too and we’ll move that forward. We’ll start to build this valuable digital reputation self in the digital space for people. We’re going to start to realize that. I’m admitting this data and doing these things, why shouldn’t I be able to capture it and be able to promote myself in the digital space and not have Facebook and other folks just scrape it and do it for me and make billions of dollars doing it? We’re at an inflection point or digital space and the academic realm will be the first one to show this reputation as something of value.
This is all coming at a time. I’ve been hearing that everybody’s been hacked. I’ve never heard so much hacking going on as I have. You’ve got HIPAA with the medical and you got FERPA in education. What’s the impact of all that is going to be if you start keeping this digital imprint? How does that impact FERPA?
It completely changes it. That’s why we were working on it with UT when we looked at it that way. It completely changes the way we address FERPA on. It erases that thing. It’s the same thing with HIPAA. It works with a company called Hu-manity. They are Salesforce customer, but they have this great idea of how to instantiate on the IBM blockchain a way for somebody to claim their 31st human rights, which is an interesting play on the blockchain. When I’m talking about blockchain, we’re talking about human data’s property and that should be the 31st human rights. This has already been promoted to the UN. There are eighteen states that have this writing legislation around this to say your digital data is property. That’s a big deal because they can tax for it then. I was starting to see these things come up. It’s going to take policy and law. We’re going to start there to be a law governing society. We should start there and look at where we’re at with our current policies and laws and why they’re written that way.
It’s written that way because of the way our system captures that data and are very insecure just as you said. I have a graphic that I do in my talks that starts in 2004, but it shows in a bubble chart that rolls vertically from 2004. All the data breaches over 50,000 people. There are a lot of wide spaces with a couple of breaches here and there, but then 2009 comes and the iPhone comes out and you start to see mobility happen. It starts to get a lot denser and then you see IOT happen like ‘14, ‘16 completely covered. There’s no way we’re getting the genie out of the bottle. We’re never going to secure this data any better way than we do now. It’s just going to get worse because we’re connecting more and more things to the network so there’s no way we can get better at this. We have to address it in a different way. That’s where this whole protocol comes in and thinks about how we store, secure and share data and that’s where it starts to affect the law and policies, which govern that data.
If we start at the law and policy level like one of UT, that’s why we start are our shared governance committee. All the CIOs of the system coming in saying, “I’m going to take all your student records.” That’s one way to look at it. The other way to look at is all the money you spend to store, secure and share that data. I know what those numbers look like, what they do to build moats around that data, and how they monitor it 24/7. All are just for PII data. That all gets deviated. What would you do with $10 million back in your budget? What would you address? Those are the conversations we need to have and not the high-level conversations about taking some of these data. We need very low level about how we address this through policy and law and how we govern this data over time.
It changes dramatically. Doing those work sessions with folks around what this new future holds for us if we can think differently, that’s where I start and end every one of these conversations. It’s not about tokens, blockchain and cryptocurrencies. It’s about how do we address the more secure and shared data and how do we cooperate on that data with other industry partners? If there’s something there to have a conversation around, it’s something valid and worth talking about and they result in changing a law or two like FERPA and HIPAA.
Like HIPAA, people don’t want to have certain things shared medically. How much of that is going to get shared if you go into a system like this?
It’s Hu-manity.co if you want to go look at it. The app is out. It’s free. Go download it and claim your 31st human rights. Once you’ve claimed your right and you create this little basic list title for your data, after that there’s a nuanced way to set up the way you want your data used. Do you want to use it for this or that? It lets the contract be nuanced. We can do this now because we’re at an individual level. We don’t have to have everybody in one company have this one contract for everybody. Every human is going to act the exact same way. Let’s say I have diabetes. My data is valuable to some research. I can allow my data to be reused for diabetes research but not for cancer research and be given away not renumerated back to me. The money comes back to me. That’s how that system is working. We’re going to start to see data marts. When people could reside or own their data, how are they going to be able to monetize that or be able to use it in ways that it should be used from pharma?
Pharma’s telling that company that they’ll pay three X for the data that they’re getting from aggregators. If it was a consensual relationship back with the human to be able to have, they’ll pay three times as much for that information to run studies with. They could ask more information, they can get more information from that person, but that person gets renumerated for that. It’s not the aggregator and not these big data companies. That’s the beauty of it and that’s where we can show it. The value proposition we’re giving to learn is that you have a way to be motivated into a role in this new economy we have. Everybody was like, “It doesn’t mean anything.” This way with your medical data, there’s a way for you to get paid. It’s plain and simple. We can get down to that level with some of these conversations in the data we see in that space.
These guys who created this Hu-manity company came from the pharma space and the aggregation space. They understand how the data is being used and how it could be used. Once we unlock value propositions across that, that continuum around your personal reputation or your personal data, what it means and how it can be valued will start to see more and more people understand. It’s not about blockchain. It’s about our own digital sells and how it’s powering our digital world, but we’re not being renumerated for it at all as the creators of that. It’s a hard thing to talk about that’s why we have to prove it out and show people how that data is valuable to them. I don’t think anybody’s going to give up the Facebook, Twitter and Instagram just to have a profile on phone of themselves just to be a way to value that. We have some great models to start to unlock that value.
You don’t think of Salesforce.com sometimes in this setting. How did you get involved in that?
[bctt tweet=”It’s about learning and learning and relearning. If you can’t do that pattern, you’re basically illiterate.” via=”no”]This role that I came into was looking at all of our couple of hundred thousand customers. The ones that want to operate differently or use our platform in ways that are more innovative like the way we were using it at UT wasn’t just using the platform or sales cloud, marketing cloud, and service cloud. It was creating a big experience and leveraging that platform in ways that it does so well through service cloud. It’s creating a 360-degree view of that learner across their lifetime, not while they’re just at UT for the state of Texas has a way to see their local economies in a different way. That was why it was super important in how it worked there. Moving forward, we’ll be able to see that in a much bigger sense here as we start to build this universal learning network out and have more and more people participate in it.
This is going to be exciting to see what you do here at ASU. This was interesting and a lot of people would love to know more. If they wanted to find out more information from you or what you’re working on, how would they do that?
It’s Salesforce.com. The blogs and things are up there, but the Arizona State is going to put up a site here before they go on break about this proof of concept that we’re building. I’m coming to Arizona to be with Dr. Crow and Lev Gonick, their CIO and go over where we’ve gotten to at the pilot. We’ll have a website out there for them. We’ll promote that from our Salesforce.com blogs. If you want to know more, just stop by there.
Thank you, Phil. This is so fascinating.
Thank you so much. I appreciate the time.
You’re welcome.
Wearable Technology with Dr. Ben Waber
I am here with Ben Waber who is the President and CEO of Humanyze. He’s recognized as the worldwide expert in people analytics, collaboration and wearable technology. He is the author of People Analytics. I am anxious to talk to you. Ben, welcome.
Thanks for having me.
Louis Carter recommended you when he was on the show. I can’t remember how we got onto the topic of that he thought you’d be so good. I know we were talking about my work with curiosity and some of the research I’ve done and how challenging it was to come up with certain ways to measure things. It was through that discussion. I’m interested in the work you do at Humanyze, but I want to start with your background because you’re an interesting guy. Just when I think this guy can’t get any smarter, I find a video of you explaining analytics in Japanese. Tell me how did you learn Japanese and what’s your background just a little bit?
It’s my BA and MA in Computer Science and I minored in Japanese. When I went to MIT Media Lab for my PhD, the whole lab is about 90% by corporate sponsors. It’s not directed research. A third of the sponsors were Japanese companies. Once I found that I spoke Japanese every single week for five years I was there, I had to translate all my slides. I present multiple times about how do you use data to understand how people interact and collaborate at work and then you’ve got to change the way the companies are managed. My book was the best there for 37 weeks. I was just over there and now we have a bunch of customers over there. We just opened an office. It’s been more useful than I thought.
You have an easy way of explaining things that are very complicated and I like watching your videos because you’re dealing with complicated stuff like analytics and wearable technology. A lot of people don’t talk about that stuff. In fact, my son-in-law works at Apple and he was in the wearable area for a while there, but he can’t tell me anything. I never get to hear anything so I’m anxious to hear more about what you’re doing. I watched your video where you described how baseball teams keep track of data and I thought that was an interesting conversation. Can you just give a little Moneyball story? I thought that was fascinating.
What’s fascinating is that you can go to any organization in the world and ask basic questions about what goes on internally that they can’t answer and how much does management talk to the sales team. Nobody knows. How many hours do people work? Think about how simple those questions are and how critical they are. The reason people can’t answer them is that they don’t have data. You might use surveys, you might use human observers. That data is useful for a lot of things, but we also all understand the limitations. It’s cloudy here in Boston. If you asked me, how much I like working with the company or who I talked to yesterday, I’ll only answer differently if it’s funny. Humans are not recording devices. Now we have all this data about what people do at work.
We have email, chat, meeting data, analysis sensor data from things like cell phones and company ID badges and also sensors. There are lots of ethical questions when it comes to what data should you look at and how should you analyze it. Beyond that, there are these other questions about how do you derive insight from that? The reason I bring up Moneyball is that it’s fascinating in a similar way for 150 years the way you build a baseball team. If you had a bunch of old guys who love about baseball, watch people play baseball. Based on their subjective evaluations, they build a team. Sometimes they’re right, sometimes they’re wrong and it was just viewed as the best way to do things. One day you get Billy Beane, general manager of the Oakland Athletics, the guy in the movie. He said, “What should we do if you use data to build our team?”
In the case of baseball, this is the average ERA and the statistics that we’ve become familiar with baseball. Fundamentally it bends data about workplace behavior, their workplaces in the fields. Everyone thought Billy Beane was crazy. If you followed baseball, if you watch the movie, if you read the book, you know what happened. It went on the record winning streak. They made the playoffs and now every single team builds the organization this way. Taking that analogy then to other industries, you can understand why a lot of the big people decisions we make consistently fails. These things are considered to be soft because we don’t have hard data around them. That’s no longer an excuse. That data does exist and a lot of the work that we first started doing it on PhDs and into our company as well. It showed the outsize impact of even doing simple things with that data.
I remember working as a VAR for IBM in ‘85. It’s been a while but we had certain ways that they tracked coming and going in different things even in the day. Now you’re using sensors and microphones and you’re seeing how people move around, but you’re not listening to them. You’re just tracking what they do. What exactly do you do to monitor their behavior? It’s interesting to have a picture of it.
There are three ways that we collect data. The first is from digital communication systems. This is email, chat, navigate to that thing. We’re not looking at content, we don’t even collect individual data, so there are no names or email addresses or anything like that. It’s just point-to-point communication. When did those communications happen? Just one point for all the metrics we provide back to our customers, we only provide them at an aggregate level team and higher level. Partially it’s for privacy reasons, also partially because what companies should care about are these big macro level things and not what does person A do treatment on Tuesday. It’s not great if they see cases like that. When it comes to sensor data, there are two ways we collect that. One is with environmental sensors. For example, our office in Boston here where our lab out in California, we have Bluetooth beacons throughout the office. Millions of the company ID badges that people already have, enables you to figure out where in the office people are. It’s just some other sensor data collection as well, like movements, overall noise levels.
It’s hard from that to detect interactions although we’ve been working on that. We still provide that data in aggregate level. What that does is enable you to figure out where are people in the office. Do people from this team tend to go over to the area that belongs to this other team? Is it more likely that they’re interacting by combining that with digital communication data? You do start to get a macro level picture of what’s going on about how people are collaborating. At the same time, we want to get much more accuracy around understanding communication patterns. Who is talking to who after we had these next-generation IDs that we roll out with our customers? Typically, those are for a short period of time. You’ll essentially get a new ID badge and you wear it for about a month. It will rotate to another division or another team about what they do. They have microphones on them which don’t record what we say, but in real time are doing voice analysis. What percentage of the time are you speaking? What’s your tone of voice? How quickly you’re speaking? It also has Bluetooth and infrared.
By measuring the strength of signals of other people’s badges, you’re able to figure out how far away they are. By combining that, you can figure out who’s talking to who and what that network is within the company. That’s the most important thing. We also have accelerometers that can look at the motion and there’s a lot of detail about all the things we can do with all the sensors. At this point, we have on a daily basis data from millions of people coming through our platform. A lot of the micro level things you can do with that data in terms of looking at even interruptions and single conversations. It’s hard for them to be predictive of the big macro outcomes that companies care about because the context is so important. What we’re doing with that is being able to create a face-to-face communication network within the company and see how much of the team are talking to each other. How much focus work do different teams gets done? Are you able to spend fifteen minutes or more of uninterrupted time sitting at your desk and work it? What that balance should be? It varies in different parts of the organization.
What we’re able to do with all that data is essentially provide you things. With an unprecedented level of accuracy, the companies can now see what’s going on internally. You can use that to whether you’re going to plan a reorg, you’re going to build a new headquarters or a new training program. Rather than do more or less the standard thing and read an article in HBR about what Google does and say, “Google is cool.” That doesn’t make sense. You can say, “Here’s what’s going on internally. Here’s what seems to be working here and seems to be not working.” You can use that to plan interventions. What is even more important than that is the speed that this technology provides. Now when we see our customers roll out without this technology. It takes you months or years to see the impact of that change on how people work on the organization as a whole. You still want to give those changes time to take shape. Now, the next day you can start to see that behavioral impact. What that enables you to do is just admit. We don’t know what initiatives are going to work.
We have hypotheses like, “This new process is going to work. This new office where I was is going to work. Installing Slack is going to help us collaborate better.” Those are hypotheses. Rather than rolling that out across the whole company, sending a lot of money and a lot of time making it happen, we can test those things out on a smaller scale because we can much more rapidly say, “Just a couple of weeks, I did what we wanted to do. Let’s roll it out everywhere and if it didn’t, that’s okay. We can try something else.” Taking the time for companies to get there. Combining all this data gives you this very detailed, very predictive, holistic perspective on what’s going on internally and it enables you to do the same things we do online but do that with the organizations.
[bctt tweet=”You meet people all the time who think that they’re working so much. They’re working hard but they’re not working smart.” via=”no”]It’s interesting to see when you were displaying an example in one of your talks. You can picture a scatterplot with the dots on the screen of the communications. You showed three different examples of a team that was doing well and one that wasn’t doing so well and so on. The biggest problem you saw in one of the situations is the data was in two groups. Those are because they weren’t communicating well because they were on just different floors from one another. It’s little things like this that people don’t think about like how much of a problem walking ten steps can make one little change. How do you deal with virtual tracking when people work virtually? Can you?
When people are working remotely, there’s data like email, chat, phone calls and calendar data that we use universally. That does give you a picture on some of how people are working. One of the things we estimate, for example, is our workloads. That’s important for our customers in Asia because they have real problems with overwork. The way that most companies try to reduce overwork is some consultant would come in and say, “Roll out this new program.” They’ll spend millions of dollars doing that, a lot of time. Their success metric is, if in the next year no one commits suicide, then it worked. That’s what they do because they have no near-term metrics and we can measure that stuff now. It’s not 100% accurate, but at a macro level, just from digital data. I know when you sent an email or when you had a meeting scheduled, you can estimate how many hours people are working. We intentionally have a very conservative estimate. Imagine you’re working on a PowerPoint deck for two hours and you never check your email. You don’t have anything blocked out your calendar. It’s rare but it does happen. We don’t get that.
One of our customers over there, we know that about 15% of their employees work in more than 70 hours a week. That’s at least 15% and we know that 50%, they’re working more than 70 hours. What that enables you to do is they can focus your attention on what are the biggest problems and then see if your programs work. Now, if you work remotely, in a similar way, you’re then able to see when I moved someone remotely, how does that change how they work? I can even just focus on things like email and chat. We see that if you’re not physically co-present with your coworkers, it does tend to mean that you will interact with a smaller number of people, less time even over chat and email than if we’re in the same office. Even when we first started doing this research, I assumed that it didn’t matter where you sat because I could email anyone. I could chat and call anyone, but we don’t. The tools we have aren’t very good at especially serendipitous interaction. You’ll still have the meetings if you work remotely, but you have to work hard to make remote work effective and there’s a lot of that.
It is measurable if you look at those channels, which luckily generate that data by default, but you don’t capture all of it. With all this data, it’s not just about trying to collect whatever quantity to do it, you can run some analysis on it and then blindly follow whatever the algorithm says is going to be most effective. You need to collect that qualitative data, you need to do interviews because you don’t understand the context. When people are working remotely, it has a negative impact on the way they’re collaborating because an office environment sucks. These are things which some of which are measurable, some which aren’t. What we’re able to do is narrow things down about what was important and enable people to get very targeted with what specific issues that they want to address first.
It’s interesting to look at the idea of what’s efficient and how people work. I meet people all the time that are working hard. They’re not working smart as you’re saying. In their perception, it’s completely off. Once you find out this one’s working 70 hours, but the other is doing the same job in 40 hours, how do you fix it? Is it up to them to fix it at that point or you’ve just supplied them with the hours and fix it?
We’re providing a technology platform that enables companies but also consultants that they hire to not just identify a fix but then see if that fix work. As we collect more and more data, we are able to see when you rolled out this program or put people in these places, here is the likely magnitude of the impact that we’ll have on these behaviors. I need to see those specific change roughly twenty times before I start to feel confident in the predictive power of our algorithms. We’re starting to see that with something particularly with just the location of workforces. Which teams are on the same floor, same city, same country, time zone effects? We can start to do things like simulating that once you’ve collected data internally to see well. If you move different parts of the workforce around or you hire more in this place, how would that likely impact how people collaborate? Mostly because we’ve seen that now a lot of times. If you roll out a new training program though to reduce your homework or to improve inclusion, for example, women in the company. I have some examples of that, but I need a specific program to be rolled out a number of times for me to see the impact. We estimate the impact with a decent degree of confidence. Contracts are always going to matter.
We’re doing some end of the year stuff and I can see our engineering team is starting to work hard right now. If that continued over a year-long period, that would be bad. Rather than automatically flag that and say, “Something terrible is going on,” you can look into what’s going on here because something weird is going on. “I know your deadline and they’re pushing this week, but it’s all okay.” If you see that there are two teams that need to work closely with each other and for some reason haven’t talked once. Looking at a high level, it could be that the point person on one team got fired and was the only person who connected these two different teams. That’s another not communicating and it needs to be fixed. It could also be, “We implemented some new automated system. It now means we don’t have to work with each other very much.” It’s fine. This is something that now you can know and you can update that. There are always those contextual issues, but without this data, you would just never understand that those things are even going on.
I could see how this adds a lot of value to avoid silos and the teams and all that. I’m curious about individual salespeople, not the sales team so much. I’m thinking of how just being in pharmaceutical sales forever and how they started to be able to track where you were just on your GPS in your car or your computer. Just making sure somebody is out there. It used to be set outside our houses to make sure you didn’t come home early. That was low tech time. We had notepad and bond paper back then. On salespeople, they can get creative and try to get around things and started to GPS. They just drove around and did stuff to make it look like you’re doing some. They would do that rather than work and you see some of that. Do you deal with that thing or are you dealing more with teams and silos?
Everything is aggregated at the company level. What you’re talking about demonstrates the dangers of particularly trying to measure the effectiveness of an individual person based on their behaviors, which is not what we’re doing. What we’re doing is showing at an organizational level what’s happening. We do provide to individuals, we’re the only ones who get to see this. You essentially get for yourself if it fits for your career. Imagine you’re a salesperson. What can you do? You can share yourself with your team, but you could also say, “I want to sell more. We’ll make more money. What do the most effective salespeople do differently than me?” You can see that like, “When they’re on calls, they talk 40% of the time and I talked 50% of the time. Maybe I could just pull back and try to listen a little bit more.” There are all these different examples.
Here’s what I can focus on. It starts to get interesting as we get data from more and more companies and even at this point, we have by far the largest data set of workplace interaction in the world. As you get more coverage, you could even start to say, “I want to get promoted. What do I do? How do I become a manager? What do they do differently than me?” You could even eventually say, “No, I want to go work at this other company. Would I be a good behavioral fit for them?” There are all sorts of ethical issues. For example, If I have to submit behavioral data to work at a company that’s concerning, there are ways that it could be done right in terms of using your data to search yourself or where you’d be a good cultural behavioral fit. We’re a long way from doing that and we’re also cautious about the ethical implications of that. There is eventually an application that’s for the individuals to improve themselves.
A reason that we need regulation around this space, is we do things in a certain way. It’s like I won’t share the data you’re talking about, but there’s nothing in the US that prevents anyone to look at data. You could essentially collect that data about people for a long time. It leads to your point, all sorts of undesirable behaviors in a crappy environment. As the power of this technology grows, there needs to be the rule for the road even things like GDPR in the EU. They don’t explicitly deal with workplace data. A lot of people are focusing on consumer data, which makes total sense. If you think about people complaining about Facebook or whatever, they just stop using Facebook and it’s not a huge deal. Your employer controls your life. If they tell you, “You need to give me this data or I’m going to fire you,” that’s a big deal. The responsibility on companies like ours and on employers as well is just that much higher. There needs to be work done on this to make sure that people are doing things in the right way.
If people wanted to know more about how the importance of analytics, they can find that in your book, People Analytics.
I’ve seen this kind of technology. There is a lot in my book. There’s plenty of work out there on what data you can collect, and what analysis you can run. There is now also more and more of literature around the ethics of this. Especially as this is a nascent field, all those things are extremely important for people who are either starting to do this or who want to do this in the longer term.
How can people get the book and reach you and find out more if they want to do that?
You can always go to Humanyze.com. That’s got links to my book and lots of the work that we’ve done. If you Google my name, you’ll see the book on Amazon or my MIT page there. If you’re so inclined, you could read a bunch of my academic papers to get into the meat of this as well.
You have done some amazing work and I was looking forward to this. Thank you so much for being on the show, Ben.
It was great to great to be here.
You’re welcome.
—
Thank you to Phil and to Ben for being my guests. We get so many great guests. You can also find out more about the Cracking the Curiosity Code book and Curiosity Code Index at CuriosityCode.com. I hope you join us for the next episode of Take The Lead radio.
Important Links:
- Salesforce.com
- Institute for Transformational Learning
- Hu-manity.co
- #My31 app
- Humanyze
- People Analytics
- Louis Carter – previous episode
- Humanyze.com
- People Analytics on Amazon
- Curiosity Code Index
- CuriosityCode.com
About Phil Komarny
Phil Komarny is the Vice President of Innovation at Salesforce.com. He is the former Chief Digital Officer at the Institute for Transformational Learning at the University of Texas System. Over his 25-year career, he has held C-suite roles and has led innovation and transformational applications of technology at each level.
About Dr. Ben Waber
Dr. Ben Waber is the President and CEO of Humanyze and is recognized worldwide as an expert in people analytics, collaboration, and wearable technology. He is particularly passionate about the power of behavioral data and analytics to improve organizations and how people work in general. He has been featured in Wired, CNN, and The New York Times, among other outlets, and his work was selected for the Harvard Business Review’s List of Breakthrough Ideas and the Technology Review’s Top 10 Emerging Technologies. His book People Analytics, is an international bestseller.