Knowledge Sharing at Work with Bill Schonbrun, CarboNet COO

Tapping into shared knowledge for enhanced data flow and team efficiency


Andy Claremont
David Blonski

Andy Claremont, and David Blonski

Tuesday, April 30th

3 pm GMT+0

CarboNet is trying to fix the world’s water crisis through the application of novel and innovative chemistry.

It’s a daunting task for an agile team. CarboNet’s scientists run experiments in R&D while their field crews work on-site with customers. There’s a lot of quantitative data and qualitative information to maneuver in between.

We spoke with Bill Schonbrun, CarboNet’s COO, to learn more about how they’re handling it.

Key takeaways:

  • You don't know how your business is operating if you don't have the information.

  • Start by using tools that your team is already familiar with. What spreadsheet software are you already using? Use that to pull your information together.

  • You need buy-in from your frontline users and workers. They're the ones capturing the information. What's the benefit to them? What do they need? Solve high-value problems that create momentum when they're solved.

  • Internal benefits of centralized knowledge sharing include faster onboarding, faster decision making, and justifying those decisions with real data. External benefits include compiled knowledge and insights that can benefit others.

  • AI is inherently about working with information and knowledge. Use AI to surface relevant insights and guidance from your compiled data.


Here’s the full conversation, edited for clarity:

Hey, everybody. I'm Andy from Glide, and this is Innovators at Work. We’re talking about taking the concepts of innovation and putting them into action for your teams and companies.

We're not just staying in the theory, we're actually getting into the tactics, talking about applying those concepts. And today we're talking about knowledge sharing at work.

How do you gather all of that information from around your organization, get it out of the heads of your team members and actually make it useful and how do you do that effectively?

To have this conversation with us, I'm delighted to have William “Bill” Seanbrun, COO of CarboNet. They help their customers reduce expenses and emissions by applying chemistry to industrial water treatment. Did I get that description right?

Awfully close. It's all good man. Wonderfully done.

Well, how would you describe it? How would you describe what CarboNet does?

We're trying to fix the world's water crisis, right? And it's just the kind of simple way of doing it. We're using novel and innovative chemistries to try to fix it. The number six biggest crisis in the UN’s top 25 is water.

You've been on this really interesting journey. I was creeping your LinkedIn and going back to where you started with science, then going on this really interesting journey through sales, advertising, and software. Now you're at CarboNet. Does this feel like a bit of a homecoming, coming back to science?

It really does. It's funny, I never actually expected to return. Back every time I would watch House on TV, I'd be like, oh, like I'd kind of crave missing it.

When this opportunity presented itself, I was at Oracle. When the opportunity came up, I was like, you know, this is a bit of a full circle. I got really excited about the idea of marrying up all these bits of my career into doing this. So, yeah, it's a return.

I can only imagine, given what you're doing with CarboNet, what that scale and complexity must be. Can you give us an idea of what it's like, what kind of information your teams are dealing with?

It's huge, and you can kind of think in like a continuum, always where this is kind of like R&D or intellectual property and this is the custom around here. 

Out in R&D you're dealing with coming up with formulas, and often the formulas are by testing out thousands – tens of thousands, hundreds and thousands – of different molecular connections. A plus B, A plus C, A plus D, A plus…  you just keep going and going until you find the one that works best.

There's an enormous amount of data out there that they're using. There's a lot of AI and ML tied up in there to figure out what will work best, and then there's the application of that against water, like does it work on the water?

There's a huge amount of data there to go and figure out whether product is or isn't gonna work and if it is gonna work, how and where and why and you could take it all the way to the other side of the equation, which is like your customers, it's how is it working for the customer?

You set an expectation when you go out to work with them, it's gonna be cheaper than what you currently use. It's gonna work better. It's gonna reduce your labor. It’s slices and dices, right?

So gathering all of that data out here to understand whether it is fulfilling its promise and then everything in between. Your employees, how they're working production. Are you producing on the schedule? Is your waste where you thought it would be? Your raw materials, are you getting them for the cost?

QA, QC... you can just go round and around to something as small as expenses, and all the way to as big as dosing. It's parts per million dosing, right? So you think about that there's parts per billion, lots of zeros.

So there's huge amounts of data that you gather and without being able to see it or bring it together, you really have no idea how the business is operating.

When you were coming into CarboNet and seeing the complexity of this and just kind of the vastness… I'm trying to keep up my head, just imagining what that must look like. What was that, a first kind of inclination of priority for you, when you're dealing with this volume of data, and you're coming into lead operations, what's going through your head?

In the early days? It was the belief that you can't fix what you can't measure. And so you don't know what you're doing, right? And you don't know what you're doing wrong.

When you're early in your journey, as an entrepreneur, when you're early, you often can't answer the question, like where are you a good fit? You know, who's your target customer? What's your this, what's your that? 

There's all kinds of things you're trying to answer. And so, and you said, as a management team and it's like, well, how did we do this quarter? And everybody's got a slightly different response to that, right?

So the first thing was codifying that. What are the KPIs, what are the target metrics? How do we run the business? How do we measure the business? What matters to us?

In the early days, to us, it was like what matters is how much water are we cleaning, right? What's the number of barrels of water we're cleaning? How much toxic chemical have we taken out of the industry? How much carbon did we take out of the atmosphere? Things like that. So we started looking, saying, okay, that's the first thing that we want to look at.

So you begin to build just really simplistic data models, and you use what everybody else uses. Excel, right? It's the world's best BI tool. It's the one that everybody uses. Everybody starts with it.

I was in BI for years and it was like 84% of the world uses Excel. And I was like, gee, I wonder why? Well, because it's free, you know, you go by office and everybody's got access to it. Everybody knows it, everybody's got it and it's good. It works for a long time.

So the beginning of it was really just building models around, you know, revenue at one side cost. So you understood that stuff tying that into finance because that's just the basics of running a business. Like we all kind of think, money, it's always there. And then the other important metrics: are we fulfilling our mission? 

My partner Barry, the CEO, has always got this question: is the world a better place because CarboNet’s in it? It's a really important thing to us, and that doesn't mean "did we drive good revenue this quarter? Did we hit our GM targets?" It's "did we pull enough crap out of the world that it, you know, it's better because we're here?"

So we're constantly watching those things. So that was where we started. So you kind of start there and then it's like, okay, what are the legs of those stools? And then you just start to move down those stacks.

And going to Excel? I always joke with my friends, you know, how much of the world runs on Excel and spreadsheets. So this mission, at the high level, when it comes down to the actual tactical bit of "how do we know, how do we know? Are we doing the thing that we're supposed to be doing?" This is all being tracked in spreadsheets, tis is all just kicking around in Excel workbooks?

Day one, that's where it started. I mean, you know, before day one, it was, you know, on notepads. For the first part of my journey here… I've got a couple of different hats over here. But the first part of my journey was to create what was called Bill Suite. 

So at NetSuite, my nickname is Sweet Bill. And so here we turned into Bill Suite.

Bill Suite 1.0 was a series of Excel spreadsheets that were linked together and they updated each other and anybody could go in and look, no one ever did, but you could always go and look at it. And every other Tuesday we get together as a management team.

In fact, today is that Tuesday, and we look at the metrics. Where are we, where, you know where and why and how and it was just Excel.

When we were talking in our prep session, you were talking about you having these disparate systems. So I assume Bill Suite wasn't the solo system running on Excel. Were there a lot of other workbooks kicking around? A lot of other sources of data?

Yeah. Finance started in Excel and then went to zero. The R&D team has tons of different scientific things. The field team, the guys who deliver the product, they have theirs. Slack is heavily used. A lot of our ordering systems are in Slack, WhatsApp.

I mean, there's just data, data everywhere and not a drop to drink, right? There was just data all over the place, and in a lot of cases, not forms that you could easily manipulate and bring together Excel became the simple ubiquitous platform.

But yeah, there is still, to this point, we're constantly closing off bits. But when there's nothing, when people don't have things out there? They grab the tools that they're comfortable with and they start to use what they're comfortable with and they create their own little shadow world.

And so you go ask them a question, they have an answer. Well, where did you get that answer? It's, you know, I've got it written on my game boy. It's like, well, I don't have access to your game boy, so sorry for you. There's lots of that that sits out there.

We hear this a lot at Glide, about these disparate systems, and you're going through the process. Now, when we first spoke earlier in the year, going through that process of trying to bring these systems together, right? What's that process? I assume there's a little bit of tension, a little bit of push back. Why, why rock the boat in your case? Why go through the trouble of bringing these things together?

The more practical answer is you can't grow, you can't scale the business. And this is the kind of thing we sit and we talk about, is when we set out, we started CarboNet, we were building it. We wanted to build a very large company with not a lot of head count.

So the idea was to be very efficient, and to be efficient, you have to have tools and the kind of most basic tools when you think about running a business, it's data, it's mining data, it's understanding data, it's surfacing data. So it's these systems, they have to connect, they have to talk.

It's easier, frankly, to start a company. When we got here, there wasn't a chair, there wasn't a desk, there wasn't a bank account. It's frankly a little easier to start from that than it is to go in.

We've taken over a bunch of companies and like, "oh we've been running this System 34 since 1812, like before electricity was there." And it's like, okay.

So we didn't have to break a lot of bad behaviors, a lot of bad habits. The habits we were breaking is like, "is this a tough thing to do? And how do you do it?" It's not that tough here because people hadn't built it. It was only a couple of years old in their habitual cycle.

When you build an efficient company, it also means that a job that would usually be done by 10 people is being done by one. So that one person, they don't have a free minute to give you. So anything you're going to do to encroach on their time has to add value to them.

So, a big part of this journey and a big part of getting somebody on board and like, why do you do this? This is a bit of Sisyphus, right? You're pushing a rock up the hill.

How do you go tell somebody working 80 hours a week? Do you want them to do more? You can't, you have to tell them "I need you to do less". You have to explain to them that by doing this, you're actually going to do less.

So we kind of designed everything where we would take the burden, the brunt on our shoulders, until it actually was less for them. We built things in a way where minimal amounts of their time and information maximal amounts of hours and then only cut it across. And it's like, is this gonna make your life easier? And if their answer is no, then we go back, we keep drawing, going back to the drawing board until it is.

That's part of how you get momentum, you get the buy in, you get it working as making it easier. I think when IT goes at something and it's "we're doing this because like there's a rule" and "thou must, thou shall"...

It’s like Harry Potter, right? They were the educational decrees that went up on the wall. You can't do it by decreeing it. It has to make their lives easier. Otherwise your teams won't follow.

So at the leadership level, you see the end game, "this is what we need to get into place". But when it comes down to in-the-weeds, when talking to the team, it’s "how do we get buy in here?"

Is it really about that? Reducing friction and making them feel it, like here's the thing that's gonna matter to you. When you’ve got a million things going on for your team-of-one, that we're asking you to do less is a huge lever, sounds like?

We believe very much in an inverted pyramid. Like Barry (CEO) and I are at the bottom of the pyramid. We're the lowest, like we're the whale line, we're the bottom. Everything slides at us.

A lot of companies work the other way. We're like, oh, you're the boss. It's like, no, we're the exact opposite, we're here. So you do your job, not the other way around, and we really do live that.

So when we were designing this and we were doing it, it wasn't "do I think it's a good idea?" It's "do you think it's a good idea?" I don't do your job, right? So you really have to walk that walk. You go out there.

We didn't have a big design doctrine. There wasn't a big design document. There wasn't a big SOP that's, like, here's what we're building. It was, I want to go out and do the following. Here's my first cut at it. What do you think? And the team has to be totally empowered to be like, this is crap, right? This won't work for me. And it's like, ok, great, why not? And they have to feel empowered to tell you why not.

A lot of organizations you're kind of scared to say to somebody like, oh, this is why it doesn't work here. It's like, no, no, no. Tell me. I can't make your job easier if you don't tell me how.

So the true designers of the technology of the software, the platform, whatever you wanna call it… it's the folks who are out there on the front line. They have a huge amount of power in how this works. And that inversion in doing that, I think, helps it pick up steam.

On a daily basis I'll get “hey, could we do the following?” It's like, absolutely, you know, hold my beer and you go. You do this for them, and as soon as they see that they actually have the power to drive where the value is, they become way more invested.

That, I think, is a really key thing in getting your end users invested. It's not an IT-led project, it's not an executive-sponsored project. It's a front line project. "It only works if it works for you."

Do you see yourselves, from a leadership level, more as stewards of getting this thing going? It's really that momentum that's coming from the ground level. You're not gonna be able to push this ahead without their buy in.

We can't do their jobs. We can't do their jobs. If they weren't here and we're not making their lives easier, we don't grow. It sounds silly but you can replace an executive, right? There's lots and lots and lots of people who go out there know how to run an Excel spreadsheet, right?

Go get Bob Petitto, drop him in my spot. He's gonna run the same, or better than it ran, you know, today. But the folks on the front line, the ones that have figured out how to solve the customer problem, they're diamonds, right? You have to protect your core asset and that's the front line. That's the folks out there.

Think about any experience you have, as simple as, like, going to Burger King. It doesn't matter how good the burger is. If the person at the register was a jerk, that was your experience, right? 

So the front line is your experience as a customer, and we care about three things: the best chemistry, the best team, and the best customer experience. We hold that one really, really dear. And so the other bits kind of hold together, and if you don't really live that, it falls apart. So that inverted pyramid? That's the core ethos to us. It always has been.

I'm gonna go back to the name you dropped: Bob Pettito. Bob is one of our Glide Experts, building Glide apps for others. He’s an awesome YouTuber as well. Tons of courses. So let's do a little side rail here. Let's talk a little bit about what you built in Glide, to give folks some context about how you're pulling all this information together.

It goes end to end. I mean, it really is an ERP system end to end. That’s truly the start and end of it. You can start your simplest thing, like a business balance of supply and demand. 

So demand: What do your customers want from you? You supply? What can you provide to them? If you're a services company, you're providing our accountants, lawyers, whatever. If you're a product company, you're providing products. 

So the whole system is really around the supply. So what are the materials we buy? Where do we buy them? What are the materials we make? How do we make them, who makes them?

So employees, what's the machines they go on? How much does it cost? You know? So that's inventory. You're making your inventory, and then the other side of the equation is your customers buying them, and watching how much they're using them. 

So ours is a consumption-based business. So say you buy, you know, a can of soda, and when that can is getting down low, you're gonna replace the can of soda. So everything is really connecting those two worlds and then understanding all the underpinnings of those bits, right?

How fast is he drinking his coffee? When is that coffee going to run out? Did he want more cream? Did he want less cream? Was the sugar good? Was it not good? Right? Is it achieving what you want to achieve? 

So the system is really, top to bottom, in its simplest form. It’s an ERP right? Enterprise Resource Planning. We just wanted to build one that used the 5% of an ERP system that we actually need without the 95% burden that we didn't need. 

So the first application that we put out was one for our field team. Our guys and gals who go and visit our customers in the field. "How is it working for you?" They gather notes, they take pictures, they do all kinds of measurements and stuff when they're out there.

First thing was to make their lives really easy while they're on the field, because they'll be wearing, you know, it's cold, they're wearing gloves, right? So how do I dictate? How do I take a photograph without doing it? And the back was production, called a batch log, which is a work order. Anybody who's  manufacturing to work order, and just taking all the work orders and moving it from being paper to moving it to being in a system. So you're kind of connecting the back like the work orders here.

The other end of it is the customer, right? So that's the two sides. So I kind of started the two sides and then we're, we're working towards the middle basically.

And then Bob was helping with some of that build?

So I reached out to Bob early early in my journey just to be like, so I went, I got certified into the whole thing and I was like, yeah, I'm certified and I'm an idiot compared to this guy. So I need to go and make some friends. I reached out to him, I reached out to Marco from Loqode, reached out to a bunch of different Experts and was like, I just wanna know what's the best way for me to raise my level?

Do you, do you want to look at my apps and tear them apart? Go tell me, you know, what part I'm doing wrong? Do you wanna educate me? Do you want to write some? 

So Bob and I have created a great relationship. He's a bit of a sounding board and I give him access to the apps, and I'm like, what would you have done differently?

It's the "I don't know what I don't know". He's got these great videos out there, and I can follow him sometimes, but other times it's like I’m so over my head, it's not even funny. I've never written a line of code in my life, right? And so I'm like, dumb it down for me, Bob.  I'm not a smart man. Tell me what to do. How did you do this? Why did you do this?

He and I work together like that. He's awesome. I love the guy. He's just amazing.

You mentioned AI and machine learning in your intro to CarboNet. So where does that fit into the picture here? Are you using AI, or how are you thinking about AI for what you're building?

We first introduced AI as a fun tool because we wanted to get people comfortable with it. So I actually called the module “Your AI Overlord”. The joke of it was because we wanted to get people who didn't really understand it to be comfortable.

Our company is interesting. It's built from young PhDs, right? So, you know, 280 IQs to folks that we have in the field, who it's a totally different skill set. It's not that one is smarter or less smart, but like truly academic to truly not academic, right? Grade in the field and you couldn't put these guys here and these guys there. It wouldn't work.

But for the guys out here we did this overlord thing, and the idea was, they took notes, right? They're always taking notes. And so we changed the first to be able to dictate.  Glide’s very good at text-to-speech, speech-to-text, and moving it out there and then summarizing it. I built a little prompt engineer platform for them, and be like “I want to summarize it".

Then I started to do fun things, like what language do you want it in? English and Spanish were very common, but I put German, I put French, I put Dutch, I put a bunch of really stupid ones and then I started to put tons in there. This is where we started to use AI to be fun, to get folks comfortable with what your “overlords” could do, and it got them hooked.

Your tone was Internal or External. So for Inside, you're gonna say anything about the customer. External is you gotta be nice. And then I did Texan, because a lot of our teams are out of Texas, and so Texan would convert what they wrote and it was very funny to come back. "I was out on your site fixing to look at the pump." They kind of loved it, and that it broke the ice around AI.

Now they kind of understand that it can do fun things. So now we're beginning to add bits to it where it's like, go back and look at all the information I've had before and give me a suggestion. 

So the first bits were just breaking the ice, like simplifying AI, humanizing it down to where it's like, yeah, it can be fun, and then showing them where the value is.

That was the first bit, then in the back side with R&D, we already do a lot of AI and ML. So it's embedding those machine learning models and then taking some of that information from machine learning. So if you learn that these molecules work well here, how do you apply that information out to the front line? It's beginning to proliferate that model.

You're almost taking your own LLM and moving it out here so that the early data starts to grow with your real life data, and so on and so forth. So there's a lot to it.

We're in the first inning of this AI baseball game. It's not entirely clear how to use it long term. There's some great stuff, like Copilot’s killing it in my mind, the way that Microsoft does it. But it's bringing those kinds of things into the business tools.

If anybody doesn't use Copilot: we record all of our company meetings on Teams, and an hour later we get a transcript, it's summarized and you can ask it questions.

We want the same kind of thing in the product. We wanna be able to say, hey, what customers are doing well, who's not doing well? Where's the product working? So you get to that, that bit nice. So we have big plans for it, but we're really early in it.

So if we zoom out and we look at this path from where things started… you have multiple data types, data sources, different teams, things living in spreadsheets. When you got started, you were connecting the dots between these different spreadsheets. 

Now you're bringing these different sources of data together, and you're building a custom ERP, the fraction of what you would typically get in an ERP suite, only the functionality you need to pull that data together and do useful things with it.

So now we have this centralized access to information and knowledge. You're extending that with AI to make it useful for your team, and then the different stakeholders, they're gonna be interacting with this system. Do I have that right?

Yeah. And the last bit to it, Andy: think about your growth and when you add new employees, who do you want to train them, right? If you have this huge amount of information in there, and you think about onboarding an employee onboarding a client, you eventually get to know, you know?

The topic here is knowledge transfer knowledge management, right? AI at its core is around information, it's around knowledge, it's around using that, kind of making that an advantage. 

Eventually, I would see us using that as a knowledge management tool, right? AI is gathering the knowledge, it's summarizing the knowledge, it's figuring out what bits to pay attention to and it's evolving and it's learning over time, which is something we humans don't do a great job of. 

We write a document and we leave it, and then you go to the next document. You don't think a year later, "oh, I need to go update that one that I wrote because my hypothesis was wrong". AI is really good at doing that.

So eventually when we're onboarding clients, when we're onboarding users, when we're onboarding employees, new scientists, whatever, I think that gives us the foundational platform for raising their level of knowledge at any time.

It really comes full circle with this idea of knowledge management. And that's where I think AI really is gonna play a huge role for us.

Yeah, I think you've really hit on the core of, when we talk about AI and AI and practical use, what we're talking about is knowledge. Now, in our prep call, you also mentioned this impact that you've seen from going from seven days to 45 minutes in end of month reporting as one example. Do you want to elaborate on that or are there other impacts that you're seeing as well that you'd like to talk about?

Oh, for sure. I mean, that was a really simple one. 

We manufacture our product in multiple locations. So at the end of every month, if you think about how finance works, you have your balance sheet and you have to value your inventory. So how much inventory do you have? And it's a combination of, it's like how much did you start with? How much did you make or consume, what you buy, what you sell, what you use and what's left and you value what's left.

All of that was being done on pieces of paper. So imagine multiple locations with multiple pieces of paper. You'd get photographs of these pieces of paper that got dropped into Slack.

Piecing that together at the end of a month, and then you go and you check it like, okay, somebody will walk through the warehouse. It says you have 32 of these, so go count. "I only see 28." And this happens every single month, so where did the other four go?

As soon as you codify that, as soon as you put that into a table, and there was a question? It's all Glide native tables. As soon as you put that into a table, now you're watching it in real time. So I know we know at any given time what's happening.

The end of the month used to be the operation team and the finance team got together. We went through all the photos, all the documents, all the everything, all the sales, all this and it was like, okay, how much do you think we have? Okay, how much do you think we have? Yeah, those numbers don't match. Let's go find out why. And we spent days and days and days just unraveling and unraveling and unraveling. 

In the end it was like, what's everybody comfortable with? What's your best guess? Because things get forgotten. "Oh, that one spilled. Don't you remember that day? That there was the the hurricane and the cow tipped it over?" You forget when you have the day. That was a true story, actually.

When you have the data sitting there, you can literally just look and at the end of the month. “It says we have seven, I'm gonna go check this.” You go down to the warehouse, they start up FaceTime, they walk around, you look up, there's a seven. Everybody good? Check. And you move on and we just, that number moves across to zero is the value. 

So that was the seven days down to 45 minutes. Frankly it's less than 45. It's only 45 because we’re still getting to trust in the info. Eventually it'll be one minute, like the info is right and you just push it across. That was that bit of the journey.

Gotcha. Following up on the AI note about "what's the actual practical application here?", knowledge management for the sake of knowledge management isn't useful. But when it's knowledge management, for the sake of something like this, where it's reporting, tracking, being able to go back and validate assumptions, or just double checking things. That's where the value comes through?

It's the internal value. Think of the external value though. So let's say we as a species are united that solving the world's water crisis is important to us all. Well, if certain customers are figuring things out and you can redact the data, but people have figured out how to do this, you eventually can surface this information out and say if you want clean water, do A B and C and here's all the underpinnings of how and why.

So you're now starting to gather huge amounts of data so you could open source. That is one of the ideas, but information is what drives us all forward. It doesn't have to be the formula. So often it's the "what are you doing?" and then there's the "how are you doing it?"

This is the how part. Putting the how out there helps people understand "in this region of the world for this type of water to remove this constituent, whether it's dirt or iron or whatever that's constituent. Here's what you do, here's what worked."

Surfacing that as AI, if you eventually put that out there, you provide that knowledge. It’s old knowledge is power. Information is power. 

Why should everybody have to start this journey, you know, from zero and learn? You have to stand on the shoulders of those who did it before you, in something like this, where we don't look at this as being competitive, right? This isn't a zero sum game. We all win and we all lose. If the world runs out of water, we all lose. So if we can use this information as a way of providing it so others can get on and you're accelerating your timeline, right?

The first voyage took them five years to figure it out, and then it became this and this and this. Eventually it should take you five minutes because all of the how is there that AI really powers. That’s, to me, the long game in AI for something like this.

It’s like a compounding interest of knowledge.

It really, really, really is. It is.

Sticking with that external application of capturing knowledge and sharing that knowledge... another nugget I loved that you dropped in our prep call was “equipping yourself with the data for negotiations and decision making”. That's just stuck in my head since we last spoke. Can you give us some examples of that?

Yeah, the mission is called weaponizing data, and you know, weaponizing is fun.

Like you think of that as a military term, as a bad term. It isn't necessarily weaponizing, it means giving it power and importance, right? Strength.

We started with "you can't fix what you can't measure". So the customer comes back and they're like, the price, you know, the product isn't doing A B and C and the price needs to be, you know, blah. 

If you don't have good information, it's really no way to have an intelligent conversation. So because we have such good data, when our clients come to us, it happens often, and you think about it.

You get the folks in the ivory tower and then the folks who are doing the real work. So the folks in the ivory tower call you up and they're like “your products doing blah”. And it's like, actually here's what's going on, and because you can back that up with huge amounts of information, you can help educate them for things that they're not able to see.

We often have better data than our customers will have. And so we're able to show them like, hey, here's what we see, here's why, here's how we're doing it. Here's why we priced it. So that becomes a very important thing.

It's really shifting from opinion to truth and backing that up with the data. So when we talk about weaponizing data, it's often gathering so much that we state an opinion, it's really more fact based and you can back that up. We try to do that everywhere, right? When we get into these meetings and someone says, you know, oh this molecule works better. It's like, okay, let's go look at the data.

Barry, my partner, worked at Bain and Company. The saying there was “in God we trust, but show me the data”. We very much hold on to that. And it's a truthful thing, it’s not negative to religion.

This is kind of the joke that like everything is really about the data. And so everything we do starts and ends with the data. Opinion is just opinion. It goes on the side when you have good data, and that's what we mean by it.

There have been a lot of great insights in our conversation so far. But I want to be explicit with this next question. You've been through this process so far with CarboNet, plus a lot of experience dealing with other companies. What advice do you have for others who are trying to get started with this? Maybe they're a startup, maybe they've been established for a while, but they're in that situation you were talking about. There's information everywhere. What advice do you have for them, for improving knowledge sharing within their teams?

Start small, right? Often people get stuck in analysis paralysis because they try to bake and they want to boil the ocean. They want to think about everything all at once and it's just not feasible to do a big company or small company. I would say, think agile, right? Start diving in, get in there. Do it. Don't spend your life planning; actually get in. You won't really know what's wrong or right until you start it. So just dive in with both feet. Don't be afraid, just go at it.

I would pick users and problems that are high value things, that when you solve they're going to create momentum. So don't “go fix a problem”. Like finance has all the tools they need. They'll be the last ones that I go and try to do anything for. And it's not a knock to them, but like there's a gazillion finance tools out there. But there's no field service tools we need, and there wasn't a production tool. 

So go and find folks who might not know that they're clamoring for it, but you see the pain they're going through, and you know that you can help them. So to me, it's really that simplistic. Just start, don't wait, don't worry about failing. And like the old joke with Abraham Lincoln. He failed 20 times before he succeeded.Get in there and try and don't worry about what might go wrong. That can all be solved. 

People spend too much time sitting on the sidelines trying to plan it out, and game it, and wonder and wonder and wonder and then eventually they give up because it's like, ah, I got my day job.

Just get in there.

It's funny. Glide became my happy place. I joke around the company with this. I have all these other bits, but it became the thing I did on weekends and nights because I ended up enjoying doing it so much. Once you dive in and you start to deliver this value and you go for the small wins. Go for the quick small wins. So you can build that momentum, right? Build it up. So it's things like that, I would say as a starter, it's small little wins and then you just build upon that.

My final question. And I'm gonna assume that you've taken the Innovators at Work assessment, if Glide is your fun place. What's your innovator type?

So, it shocked me. I got Scientist, and I didn't think, if you know of the four choices, it was the one I probably would have the least on. But I came up a Scientist.

I love that. Okay. We're both scientists. Its original iteration was Mad Scientist, which I love, because what you described about data… going in, getting started, just do it? I think it’s so much in the vein of like that, that scientist ethos.

And I think that's a great point to leave off on. Bill, thank you so much for the time today. So many great insights, so many little nuggets that I can't wait to pull out for everybody else. 

Thanks everyone. See you all in the community. Take care guys. Thanks Andy.

If you have more questions, more thoughts, let's keep the conversation going in the Glide community.

Don't forget to subscribe for more sessions like this. We’ve got a bunch lined up. Hit the Glide Events page where you can also find all the events that we've hosted over the past year.

Share this article