Todd Gagne

You making the time. It's good to see your face, and excited about this conversation.

Brian trable

Yeah. I miss you, buddy. It's good to to be able to get together and and kinda talk about some things again. It's been a while, so It

Todd Gagne

has been a while. So why don't you start maybe, Brian, just talking a little bit about, kind of your background and stuff. So, you know, today, your title, I think, is vice president of products at Avid Exchange. You and I worked at Concur. You came up through an acquisition, with TRX that was doing some travel data.

And so not only did we get to kind of work on that, but then you got to work more of kind of the data analytics across kind of concur and then longer term SAP. So you have kind of a rich history in, looking at data and then figuring out ways to drive value back to customers, which I do think is kind of a theme in some

Brian trable

of your career from a

Todd Gagne

product, product management standpoint.

Brian trable

Yeah. For sure. I think it started out early being on the other side of it. Right? Being on the customer side, starting at Oracle and having to kind of run the business side of things.

Right? Within procurement and global travel and really understanding, like, how do we run a business? Right? How does data, how does analytics fall into that? How do we think about data throughout the whole process?

So I think that the 6 or 7 years that I had there really helped me understand and have some empathy of the customer side of a lot of these problems to really say, okay, now I can jump to the other side, if you will. Right? And then the TRX piece was the first starting point. And now now knowing a little bit about the customer side and what they need to succeed, how can I help, you know, bring along companies, bring along teams to kinda have that same mindset? And, you know, obviously, TRX was very data and and analytics focused completely, which which wasn't a whole breadth of of how does that fit in the products and services.

But I think with the Concur acquisition, obviously, you know, bringing that more into a product management world, kinda looking at broader integrations and solutions and how we think about data across the ecosystem. You know, working for you for that period of time was really good. And then it really blew up after the SAP acquisition when you start having to think about, okay, now I was responsible with with Jason and others in in in building, you know, in in in ecosystem and a a data alignment and a constant experience for, you know, a concurrent Ariba, Fieldglass, the success factors, the legacy SAP. And now, you know, the problems or the outcomes you're trying to get aren't really any different. I mean, at at its at its most basic, it's pretty basic stuff.

But the complexity of what you have to do to get there is where, oh, holy hell, that that becomes a challenge. Right? So, you know, I think that the SAP experience and and at the big company was was really good for a while. I was ready for a for a bit of a change. I wanted to go try this the small size company, the start up, see if I could bring some of that knowledge to, to a different industry and a different idea.

So I went to a startup. I was really focused on we were trying to build basically, like, a a master supplier repository. A supplier platform, really, where we could integrate with whether it's the ERP and the vendor master side or the procurement side. But, like, how do we use AI and data to create the best single data set to understand companies across the world, and and be the most trusted. And then, you know, you you you landed on Avid Exchange, and I decided, okay, how can I take all those experiences and go work with a great group of people with a good company that has, a very unique and interesting problem to solve, and it kinda tie those together?

So that's that's what brought me to where I'm at now.

Todd Gagne

Yeah. Well, that's cool. It's been a cool journey. So maybe talk to me a little bit, you know, like, I think, so our audience is generally kinda startup founders and and kinda people in startups. And so, I think that product management is a little bit different from the scale and stuff that you're talking about.

And so you've got kind of where wear both hats. Right? You've kind of been in a startup where it's been super small. And I think that 0 to 1 product management is very different than what you just described at, you know, SAP Concur, where you basically have, you know, multiple product lines, you're trying to do a data lake, you're trying to do a bunch of different things to build products. The level of complexity in that is kind of interesting.

And so maybe just talk to me a little bit about, like, what's 0 to 1 look like versus, like, the more mature pieces of it. And so I think the people are not as interested probably in the mature components of it, but, like, how do you really simplify to 0 to 1 when you have no product, you're trying to figure out, you're talking to customers, and you're trying to say, like, what is the most important essence that I'm trying to do to create value? Like, how that's a hard problem to go do. Yeah.

Brian trable

I I think the framework is the same in both the complexity and the number of steps and what's included in it is probably different. Right? So when you think about the process, you know, going from ideation, discovery, design, build, release, operate, and and follow that through, like, that's a a foundational framework that, you know, everybody should be following to some level of complexity or not. But I think where you talk about the different types of work, right, yeah, you you're not gonna have as much feature or targeted fix stuff because you're starting from scratch. But, like, what's that major investment or that transformative change that you're trying to build?

And to me, it all starts really in the ideation start. Right? How do you get the ideation start right? Like, everybody can have a process and a process is linear and end to end, but, like, making progress within that process is messy. Right?

It's like a bunch of spaghetti strings that are tied together. But the ideation part of that, I I think, especially in that 0 to 1 is the most critical. Right? Like, how do you how do you gather enough insights and how do you use the right methodologies to turn that research into insights? Right?

How do I get the qualitative, the quantitative, the structure, the construct, all the information to really understand what insights I'm trying to find, whether that's in my industry, whether that's in my, you know, particular vertical, whether that's in a particular product piece. And then using those, you know, synthesizing methods to really craft the problem statement. Right? And and that getting that problem statement right, that drives everything else. Right?

Because then, great. If you got a problem statement correct and and I have strong conviction about this problem statement I wanna go solve or that I think the industry needs or that's that's missing, now you can go quantify it. Right? Because because now, great, everybody can have an idea of a problem statement, but is it worth doing? Right?

Can we quantify that? Is that is is is the same and the tam and the lamb and the all the a b c's worth it? Right? You can do a quick analysis of, like, is this worth doing it? Yes or no?

And then you craft the hypothesis quickly on, hey, this is this is my solution hypothesis on how I wanna go go about it. Now a lot of people would jump from there to let's go build, let's go design. But if the the most critical component to validate, like, that problem statement or that your solution is gonna be effective once you deliver it, it's really all about the outcome. Right? Success criteria.

And that's not just like, did I build it in a time frame that I was aligned to? But that's like, can I operationalize that? Can we can we sell it? Can we bill for it? What data and analytics do I need to prove that the the maturity is there?

Right? So you kind of think of it as like that that outcome pyramid a lot of people talk about. Right? Where you got your business and strategy and your outcomes and your outputs and the activities, and then the data falls into it, and it becomes a singular loop. So to to me, like, the the single most important thing once you have a problem statement is defining success.

Like, what's the goal and how you're gonna measure that that's successful? And then go run with it. Go go run with that hypothesis. Learn based on those metrics. If you if you need to fail quickly and start over or if you're validating throughout those steps.

And then you can kinda you can kinda build down. Like, you got the ultimate goal. I wanna drive revenue or I wanna solve this customer problem. But, like, what's the user behavior change I'm trying to drive? What what really shows that feature or product success within that?

What are the activities that that fall below that and what inputs do I need to make sure that that ongoing success is happening? I

Todd Gagne

think Brian, sometimes sometimes this is like even maybe even more base level though. Right? Like, I mean, think about like starting a company, you got 0 revenue. Let's say you got a little bit of cash to like go, you know, create some runway for you. If this first product that comes out doesn't work, you don't have a company.

Right? And so, you know, like there's a lot of iterations on, you know, can I find those early adopters? What are the core features they need? They're asking me to do 10 things. I can do 3 things well.

Which 3 things I'm gonna do? You know, is there a beachhead behind this, where I can basically say I've got some differentiation. I'm a small startup company with an idea of how things need to be done differently, but there's entrenched people in this in this industry. I gotta chip away at some corner of this to actually make it happen. And so, like, I think I think everything you're saying is right.

But I think, like, the immediacy of 0 to 1 is, like, you know, it's kinda like Elon Musk's where it's, like, I got enough money to, like, send up 4 rockets. And if the 4th one goes wrong, like, I'm done. Right? And so I think it's an extreme scenario of that.

Brian trable

Yeah. And and that's why I'm saying that the the problem statement and understanding and validating that with people early is the most critical part. You gotta make sure that if you got a dollar to spend, you're spending that on the right bet. Right? It's a gamble.

Right? Yep. And and you need to have as much of that, those data points and that feedback to to be willing to bet on that piece.

Todd Gagne

Yeah. I think it's good. I think it's good. So let's kinda pivot a little bit to maybe, how AI is kinda shaping some of the product strategy stuff. I mean, I work in a lot with the startups and so and again, we don't have any, legacy code base.

We don't have any technical debt. Right? You're starting from ground 0. And so what's interesting to me is as these large language models and stuff have come out over the last couple of years, you've kind of gone from this idea where it's kind of kind of a copilot where it's basically it's an it's a model that's basically trying to assist a human To, you know, we're starting to already see early stuff about, like, individual, autonomous agents that'll actually do they're just an AI agent that are basically geared to do one thing and one thing well. And then what you would expect over time is this kind of mesh idea where it's like, I've got these different agents that basically will start talking to each other and passing information.

And, you know, I think one of the ones that, I think about from our old previous life was, you know, like when a business trip is canceled or like delayed and all the ramifications, right? That customer meeting that you were that dinner you were to have, the hotel, the, you know, the the Uber that was supposed to pick you up. Like, all that stuff just needs to get shifted. And I I'm looking forward to the day where it's like then basically that just like you see that information and it passes it to, you know, Uber and it passes it to the Marriott and it passes it to, you know, Delta. And then all of a sudden, you know, your open table reservation got rescheduled too.

Like, that is the ideal. And I think over time, we'll probably see that. But I'm curious if if you see it differently over, like, the next 10 years or if you think that that model kinda holds.

Brian trable

I think it holds. I mean, I think you've seen you've seen a lot of progress of, like, you know, even in in engineering and development, the concept of the monolith of microservices. Right? And now we're just talking about how do we make those microservices smarter. Right?

So layering in AI within each of those services so they can make, informed decisions based on the context around them versus doing a very specific individualized, task, which which is what a lot of those microservices do today. I think the challenge, you know, the challenge is is all where a company is on its journey and what the existing legacy and infrastructure is. And it's it's it's a lot easier, you know, to think about leveraging AI throughout that entire process when you're starting from, a blank slate. Right? When you're starting from, not legacy expectations of the customers or not legacy infrastructure to have to work through, it can really think about what should happen, not what are the constraints of what we're able to allow to happen, which, you know, to your example that said before, you could have a lot of breaks in that chain.

Right? With with leveraging a bunch of existing expectations and infrastructure and experience.

Todd Gagne

I guess maybe drill down that even further. I guess I guess my personal opinion is that, product management almost needs to go through a rev with some of this technology, right? Because and I'm gonna pick on Salesforce because it's easy to pick on Salesforce. But I mean, you know, like anybody even even SAP and Concur, right, where we were, there was an established workflow, there's an established UI and there's an entire code base to support a whole bunch of customers. And so we thought about that problem in a very linear way.

If you could start again and think about it from a first principle standpoint around UI, like, what would you do differently? And I think what I'm starting to see with a lot of legacy customers is we're going to bolt on an AI component of this. We're going to put some lipstick on it. We're going to charge more. And and I I think there's a great opportunity for a lot of small companies to basically rethink some of these problems in a way that basically makes it much simpler.

And maybe pricing becomes, like, outcome based instead of just the transactional that we have. And and and and so let me just stop there and see if, like, that that premise makes sense or if you if you push back on it.

Brian trable

No. I I think it makes sense. I think the experience and how you integrate that within the workflow is critical. Right? Because I think what I see a lot of companies doing is, to your point, we're we're bolting on AI to say we have AI to have a a higher value, whether that's a higher value to the market or a higher perceived value to the to the customer.

But you can't you can't lose sight of what the ultimate outcome you're trying to drive for that customer is just by infusing AI. Right? Like, how does that integrate to the existing workflow? How does that user benefit from the enhancements without having to change behaviors significantly? Right?

Or, like, how do you embrace a a process or a flow or an industry that's already working by leveraging AI in different areas versus trying to completely disrupt what might already be working correctly by layering some of that in.

Todd Gagne

But do you think like I guess, I still think that there's a lot of processes in business that basically were developed because of a of just that's how we always did it. And I

Brian trable

Oh, yeah. 100%.

Todd Gagne

I wonder if, like, if we basically had the tools that we have today, would you reengineer that in a different way from the user perspective to make it easier? And so, again, this is goes back to, like, it's super hard when you've got 250,000 customers or 25,000 customers or whatever it happens to be to change that behavior. But I think what a lot of product management really should be is a rewiring to say, how do we use these tools to better serve the outcome that we want? Not the necessarily the the workflow process that we've always used.

Brian trable

Yeah. It's a it's a change in mindset. I mean, if you go back to the example you used of of our previous life. Right? Think about it, travel and expense reports.

Right? What's the ultimate goal? I need to get to and from a place. I need to get my money back so I'm not out of pocket, and the business needs to be able to forecast effectively. Well, that whole process was built because I have to get some picture of a receipt, because that's the only way I can validate data information.

I have to send that to somebody who has to review it. That has to get it put into some system. That has to go to some bank account. That has to get packaged up into some tight little time constraint or box constraint. Like, we've talked about this a lot.

It's like, no. You saw the transaction happened. You knew that I already booked the trip that I went on. You know that that's in a time frame of the location that I was going to for the upfront approval I had. If you see transactions coming in that are related to that place that time, assume that those are approved, do they pass policy measures?

Do they have anomaly and fraud detections? And then just pay me in my bank account. Like, I shouldn't have to do anything. Right? They're big.

Ding. Ding. Yeah. So that whole process changed. And, you know, we're we're trying to look at that now in in the world that I I I live in today with, you know, invoice processing and payments.

It's the same kind of concept. It's everything's been so linear because there's always a step by step approach and that the balances and controls were a lot of the driving force behind that. Right? Like, the only way to validate a lot of these things and a lot of these processes and a lot of these procedures, historically, across a lot of industries is somebody has to look at it. Somebody has to physically touch it because there wasn't this technology that would allow that.

So if you rethink and and take the people process out of it and think of the outcome, then a lot of the ways that we think about anything really change or should change.

Todd Gagne

So maybe let's go back to kind of where you were we were talking about this earlier about, just legacy companies in general. And so, I think they've got some advantages and they got some disadvantages. Right? We've just outlined a disadvantage, which is it's hard to rethink your code base when you have existing customers. The flip side to that is they have data.

Right? And I think sometimes we think, you know, and I think sometimes that's a panacea and sometimes that's like reality. Right? I mean, I think we for a while there, we've been saying data is the new gold. And I I personally think that like sometimes it is and sometimes it isn't.

Right? And so I think, I think if you can, if it's really unique and it's not something else you can get in the market, then obviously if you can turn that into insight that benefits the user experience and makes something easier, great. But I think in you know, you and I have been on this journey for a while where it's like we've been saying that for the last 10 years and how many companies have you really thought have used their data to massively differentiate? I can't think of enough, right? Like maybe there's a handful of them, but the majority of them and even our time at Concur.

I mean, I think there were some unique things that we did and we probably had some unique insights, and we monetized some of it. But, like, I think we've never really delivered on that promise. I mean, even what SAP is trying to do where they're trying to pull all that data together, they're still a ways away from actually, un uncheckling that that, insight to, like, really driving value into their business.

Brian trable

Yeah. I mean, some of the constraint you're you're right. The 2 big things that that legacy enterprise customers or or, you know, legacy systems in general have as a benefit is years of data and information to better build models if if they're gonna build their own internal models. It's they they typically are gonna have more money out of the gate. But the constraints to me, a lot of the times, outweigh that.

Right? You have the rigid systems. Right? Difficult to modify or upgrade without disruption to existing customers. You have existing customers, so that brings cultural and structural challenges.

Right? Like, this is the way we've done it. This is the way people expect us to do it. So changing that mindset. The integration complexity, because you know from the everybody's got technical debt.

Right? Like, we don't have modern data pipelines in a lot of these legacy customs or custom or products and systems. So the integration to just get it to work. And then, like, skill gaps. Right?

Like, we a lot of the legacy customers didn't hire and build teams around both the mindset mindset shift to think about the problem differently, but, like, the skills to execute it. So you really almost have to rethink all of those things from a legacy system to be successful. And then, you know, ask yourself at the beginning, do the customers that are using your systems and driving your revenue today, do they even have a a desire to adopt? And that changes by industry. Like, I I work in an industry now that's heavily tied to banking.

Right? So the banks have a very, very adverse approach as does government to AI anywhere within the systems. Right?

Todd Gagne

Or change in general.

Brian trable

How does it yes. Change in general. But so, you know, if you're trying to if you're trying to retrofit AI into an infrastructure in market that you're already good in dominating, you have to have the customers come along with you when, you know, you're kind of in that startup phase. You're unburdened by that. Right?

You you have an advantage. You're unburdened by that. You have the flexibility and agility to experiment. You can evolve without needing to worry about maintaining backwards compatibility in old systems or the the process ways you have. You know, typically, you're using better technology or or,

Todd Gagne

newer stack.

Brian trable

Newer newer newer tooling and and whatnot. So

Todd Gagne

And and I think a lot of times data.

Brian trable

I I

Todd Gagne

think a lot of times you can get the data. Right? There's other public sources that you can buy it. And so I just don't think it's an insurmountable thing. I mean, I think there are some companies that do it well, but I think when startups are looking at this, I do think that there's very little that's super proprietary that you can't find in some other method about buying it.

Brian trable

Well, in in in the other way to go, right, is when you think about what's the advantage of of or what's the advantage mindset when you don't have all that data? Quality. Quality. Quality. Quality.

Focus on quality. Focus on high highly targeted. Right? Use active learning to to to take the most useful data points, and then, you know, depending on the industry and the market, find unique data sources that people aren't using. Right?

If you're in the health care industry, right, go figure out how can you get the data from the local clinics that are here and here and here that are part of a big network that might not have or partner. Right? Strategic partnerships early on are good. Or or just find find us go go start something up in a in a in a niche area. Right?

Go help farmers use AI to figure out what soil health is like. Like, that's not targeted like an Amazon is. Right? There's there's different niche things that Well, let's

Todd Gagne

let's talk about that a little bit further because I think this one's interesting too. It's like the international markets are becoming more and more interesting from this perspective. Right? So you and I lived through some of the data privacy issues and GDPR from a European standpoint. In a lot of cases, a lot of enterprise organizations ended up adopting that for their US as well from just a data security perspective.

But now, you know, with those rules in place, AI is becoming something that's not easily used in Europe. And so to the point where, you know, you've got these new iPhones, the sixteens that have come out with Apple Intelligence, and you can't use it. You they're not even gonna ship it in in Europe. You look at everything that Android's done. And so those are 2 major examples of big tech that basically are now giving an inferior experience in Europe because of regulatory issues and liability.

And so, again, most startups are not focused on Europe to start with, but I do think it's interesting to think about, like, how do you design, GDPR allowed us to just do the least common denominator. But I think if you really want to innovate with a lot of this AI technology, you're you're going to be at a disadvantage, and you're maybe going to be competing products that are not as feature rich in European areas where this is an issue.

Brian trable

Yeah. It's not just Europe. I think the biggest, the biggest thing that they research is critical. Right? Staying on top of the regulations, but 2, it's the fragmentation as you grow.

Right? It's one thing, like, if all of Europe or if all of Asia Pac or had the same regulatory requirements across like, okay, that's challenging, but to your point, like, you can get feature less prod. But the fragmentation where every single country or nowadays, even every single state. Right? Like, I live in California.

We have things that come up every day that are new regulatory requirements that you're not gonna see in any of the other 49 states that come up all the time. And it's like, if you're trying to grow and expand as a new company, you gotta have a a flexible framework and a a a product that can adjust and scale to those regulatory changes. Because to to me, the the the the uncertainty and the fragmentation is is harder to overcome than just, like, here's a hard requirement for the entire market that they're going to. And and those are changing, like, the evolving rights and the consumer expectations. Right?

It comes down to trust. Right? People don't trust that that data is being used ethically in AI. People don't trust that the outcomes are gonna be, you know, correct in those really critical markets.

Todd Gagne

Yeah. I think it's a good point. I I I guess that to even drive your point even further. I think even within the, you know, European Union, there are nuances about GDPR and how it's impacted. So I think your your point is dead on, and then, you know, just how do you manage that at at what level?

And for startups, that's like pretty crazy. Right? It's just difficult to basically say I'm gonna put enough regulatory scrutiny so that I can basically do this. And I think what a lot of them are doing is basically just asking for get forgive this on the backside. Right?

I'm gonna do what I'm gonna do, make best guess, and then, you know, deal with the outcomes, which, you know, that's the startup mentality a little bit in some cases.

Brian trable

I I I think the other thing that has to be taken into consideration when you start thinking about global global marketplace, global expansion is, you know, take all the regulatory pieces out of it. But just think about the cultural nuances. Right? Local representation, localizing the experience, or when you think about adapting a AI, like, the cultural context are vastly different everywhere you go. Right?

So, like, the data or the even the questions that you have to frame to an AI model could be very nuanced different depending on where you're going and have a very specific, you know, difference in the outcome. So being able to understand culturally specific training data for some of those models and how those those are important as well.

Todd Gagne

Yeah. The example that I thought was pretty interesting even from our concur days was just how fraud is done from, like, Norway to Germany to, Brazil to to the US. Like the way fraud patterns happen were different in every single one of those. And they were a function of both your culture and your regulatory arrangement, right? And so if you remember all the travel allowance crap that was in Norway, you know, it was ridiculous.

And so people had found, like, areas where they could pad their expense reports and basically that's where fraud was. Whereas the Japanese had a totally different thing. And so it was just interesting to even see that cultural difference on a fraud perspective where you just think, okay, I'm just trying to cheat on an expense report and add, you know, pat a little bit. But culturally, it was super different and, like, I think training models had to be different by locale.

Brian trable

Yeah. And then when you add into the factor on top of that, like, the different technological technological infrastructure, like mobile first from only mobile only markets like China vastly different on how you have to solve some of those same things than, you know, somewhere up.

Todd Gagne

So let's kind of pivot a little bit to maybe pricing models. Again again, maybe contrasting kind of legacy companies and, and and startups. I mean, maybe Salesforce is one that's maybe kind of trying to bridge this, right, where they're saying, yes, maybe I'm bolting on some agents onto some of my existing stuff. They've kind of had an acquirer type of mentality to kind of grow their top line. And now they've started to say, maybe some of this agent stuff is outcome based.

And I think this is a really big opportunity for a lot of startups, where it's like saying, if your cost to develop is lower, you're going to have fixed costs with LLMs or whatever you're using. But the number of people to get you to some of these products may be cheaper than it has been historically. And I think if you want to have, an advantage from an incumbent perspective, maybe some outcome pricing is a way that you start to get market share. And so I'm curious on on how you think about that. If you think that's something from especially in the b to b world, that you might end up seeing.

Brian trable

Yeah. I think it's gonna evolve just like, you know, just like a lot of companies evolve from, you know, margin based pricing only to value based pricing. And the AI stuff is is gonna be, in the similar vein. Right? It it's gonna more actively reflect the value that the AI is bringing to the customer, not the tool that it's using itself or the the technology.

So I think it's well suited. I think, you know, there's there's a there's a a prerequisite for it, obviously, right, from a value based pricing perspective. And it's not like anything else in any other industry. I mean, look at something as basic as NFL football. Right?

Like, you got 32 starting quarterback. 32 starting quarterbacks are not getting paid the same amount of money. Right? They're getting paid on the number of wins they have and their success in the playoffs. So, you know, you pay for the value that you get out, whether that's a cost saving share.

Right? Like, okay. Great. We reduced cost by 15%. Now we're gonna get 2% of that, or whether that's a revenue share uplift.

Like, I always think back to the thing that you and I wanted to do for a long time of, like, optimizing the sales and travel piece. Right? Like, okay. If if I can help you optimize your ability to close more revenue because we've streamlined that entire process and we're maximizing the amount of people you can touch in a given trip and automating all of those processes, great. Give me if I'm increasing your average sales by a million annually, I want a portion of that uplift based on what what we're doing.

I think the the performance based pricing becomes a little bit harder just because there's there's longer feedback loops for measuring success. Right? So, like, like, if we wanna use AI to help reduce attrition, right, or on the customer side. Or think about it from an HR perspective. We wanna reduce, customer churn or employee churn, I mean.

Like, there's a long pole to understand whether we actually had success, and then, was that the only thing that drove to that success of that I

Todd Gagne

think that's the biggest thing

Brian trable

to me. A lot of other things that contributed. We changed policies. We changed processes. The the economy changed, the market changed.

Like, so those those are some of the impacts to getting there, but I think the basic, like it probably ends up needing to be a hybrid. Right? So similar to, like, a sales quota. Right? You get your base salary, you close this much deal, you're gonna get some quota on top of it.

Like, I think that's the only way it's actually gonna be scalable over time. It's like, there's some flat rate subscription type model. And then based on the value outcomes that you get for whatever industry you're doing, there's some variable price model on top of it.

Todd Gagne

I think you're right. I think that's the right model. I think, like, basically, it's basically a cost plus on your operations and then basically performance based. And I think if we go all the way back from, you know, thinking about, like, on premise going to SaaS, really, what it was trying to do was align utilization to, you know, basically what you're paying for. And so and I think to me, this is another alignment of saying, how do you get to the point where it's outcome based?

And I agree with you. Not all of them are gonna be that. I think if you look at the Salesforce example, you know, I think it's around support. Right? So if we can close a ticket and we can work through that and that was resolved in a way that was positive and with a good experience, then you'll pay us for it.

So that's much more transactional where, you know, like you were talking about, retaining the customer. Well, there's just so many more variables that are outside of the agent's control to go do that. But I do think some of these things, there's a lot of software that I think is relatively transactional that this may be cost plus plus a performance based model might really work. And I, you know, and I I'm always looking at this from a a startup standpoint where you're just like saying, how do you try to provide them some incentive or some way to, like, get in the door? Because there's already an incumbent.

There's already somebody that's sitting there that basically says, I'm happy with my solution. And you have to jog them through to say, yes, but there's a better way.

Brian trable

Yeah. And change management is the hardest part of any is is, you know, whether it's change management internally, externally, changing an industry, changing a concept. And and that's why new ways like, I I always look at, like, when I think of AI and and trying to disrupt the way people think in an industry, I I like to think about Spotify. Right? Like, I'm a Spotify user.

I use Spotify. They came out of, like, there was plenty of things. Apple Music service, there was Pandora, there was Iheartradio, all these things. Right? Like, whatever it was.

Right? But, like, Spotify was the first one to me that came out of the gate with, like, really good AI driven customization of the next song that I wanna hear based on these patterns, and, you know, that caught on quickly. It's a new way of thinking about this. A new it is like, well, great. I can go search through a Pandora radio list where I might get to the the type of songs that I want on some date.

Or I can go get the exact song with the exact I I don't know. Like, Spotify was one of those that, like, changed the way by coming out of the gate with that AI mentality.

Todd Gagne

Yeah, I think it's good. Is there any other ones that you could think of that maybe are early that are basically taking an interesting aspect to doing this? I know it's early, right? And it's like you got all these kind of independent you know, chat gbt bots, that are, you know, kinda geared towards specific things. But is there any other good kind of use case examples or things that you think have been interesting?

Brian trable

Yeah. I mean, I I was reading it. I I brought it up earlier, but like the whole farmer with soil health. I don't know anything about farming. I live in a big farming area in Northern California, but I don't know anything about farming.

But I read an article about, you know, how this company, the startup was coming out to really use AI to really evaluate soil health in the right times, in the right temperature, in the right water, in the right, like, to have their their crops grow and like, that was something that was I don't know enough about that industry to know, like, is that really disruptive? Is it, you know but that was something I would have never thought about kinda using AI for. It's like, hey, I I the farmer out there has done it the same way a 100 times, and they they seem to be having a lot of success. So I I tend to get excited by some of those niche areas that, like, oh, man. I never even thought that that was a problem.

And you've got these very smart people and companies that are coming up and trying to leverage the latest technology to provide that solution. I mean, you're always gonna have the people that are gonna go try to find the best new way to do an expense report or process a payment or, you know, file a medical claim or an insurance claim that I had to do this morning because I got into an accident yesterday. Like, there's always gonna be people that try to solve those problems better, and and and I think there's some companies that are that are doing it in in great different ways, but the niche ones are the ones that kinda get me a little bit excited.

Todd Gagne

Well, you're on the same page. So, I mean, that's basically all we do is kind of, you know, vertical niches. And we do have one that's called AgSense, which is in the ag space, and it's working with ranchers. And it's basically taking kind of what you would think about quants do on Wall Street and applying it for ranchers. Right?

So same sort of concept where they're saying like the rancher maybe isn't as sophisticated about like how to manage their cattle inventory through the market and when to sell and when to hedge. And really what they're doing is trying to bring tools down to that level to provide something that's easy for them to understand how to do it. And so like, again, same sort of concept, super nichey. That we they got a customer that like did, I don't know, 5, 10000 head of cattle, a 1000000 acres, he leased. And basically he was just looking for a better way to go do this because he was like, this is a lot of beef at risk.

And so, they so they started a company to do this. And so and and and now they're doing things like, they're looking at the, the quality of the cattle herd and they're trying to say, if you bought this sperm and this embryos, you could increase it. And then what they're doing is they're looking for the arbitrage, which is the better sperm embryo, but it's cheaper than what like the market price would be for that. So they're doing all the statistics to basically give the rancher the ideas to go do and increase their herd without increasing the spend. Same sort of deal.

Brian trable

Yeah. And when you talk about, you know, we talked earlier about data and not having data or what's the right data. Like, that's the type of niche data that you can really start building a data moat around that competitive advantage. Right? Because that I I I'm sure there's there's a lot of that ranch farmers in in the country and in the world.

Right? So you can start to get large high quality datasets. You can start making that relevance to the domain. You're basically creating a domain that wasn't there before, and now you become, you know, that moat is around your unique niche. Like, that that's a really cool example.

Todd Gagne

But we're finding more and more of these. I mean, we got something in wildland fire. Right? You never would have thought wildland fire, but it's like, you know, it's a niche. It's not price sensitive when there's a fire.

They want equipment. They want people. They want everything. And you know, like a wood chipper from the federal government is like $35100 a day. And so it's big money.

And so if you got your equipment in the right region for the fire, lots of thing good things happen for you. And so it's interesting how, like, that's another little niche and you're, like, saying, okay, but there's way more complexity in it than you thought to start with.

Brian trable

I was just thinking about that. Like, think about that complexity without AI. You've got weather patterns. You've got, you've got logistics, you've got wear and tear, you've got the people resources, you've got mobility. Location Yeah.

Todd Gagne

Like, how do you get them there? How quickly?

Brian trable

Health and safety. Yep. I mean, there's so much like, I was just trying to think through as you were talking through that scenario, like, how challenging that is today without leveraging better technology. Okay. More power to those of you that are working on that.

Todd Gagne

Yeah. Well, it it's kinda fun because there's, like, you know, you you you know, from our world we've seen patterns, we know it works. And then it's like you have somebody that's basically super excited about this problem and they have so much domain expertise. Right? They can go deep in it.

And and you know that like combination of the technology and like the business patterns that you've seen, there is a better way. And so, you know And then then

Brian trable

you have that that expertise like now you can find the right people to really bring that model accuracy to fruition. Like how do you really Yeah, that's exciting.

Todd Gagne

So we got about a couple of minutes left. We got maybe 4 or 5 minutes left. And so maybe Brian, like to wrap this whole thing up, maybe talk to me a little bit about what your recommendations would be for somebody doing a startup. Right? You know, kind of what this is, right?

Where it's like, it's a small team. You're being product manager, you're being CEO, you're being CFO. You're you and sometimes you're the guy that's doing the support. You're doing all of it. And so, you know, taking a lot of what you're doing and then bringing it down and reminding them it's about the customer, it's about value.

But like where would you like, what what parting words would you give to somebody that's like starting on this journey? Maybe they're not product people, but they're trying to figure out that 0 to 1 to basically just find that foothold. Right? Just to find the first couple of customers to validate that they they have a business.

Brian trable

Man, that's a that's a loaded question to have. So first, more power to those folks. Like, I, you know, I've worked in startups, but I've never started a startup, and I think that's one of the hardest and most fulfilling thing that I've ever seen people do when they're successful in that. I think the challenge to your point is they're, they're very smart, successful, ambitious people driving that, but don't necessarily have all the specific domains of all the hats they have to wear. I mean, I think at the end of the day, it starts with the customer.

Like I said at the beginning, like, what's the problem we're trying to solve? And and have conviction around that. Right? Like, you're gonna make one bet. Like, let's have conviction that that's a problem.

Let's have conviction that it's a a a

Todd Gagne

quantum problem. Crossing your fingers going, hopefully.

Brian trable

Yeah. And I think data is the key there, the the upfront research to to get that problem statement. And I think it's easier nowadays than it ever has been. And and coming back to the AI topic, like, I encourage folks, like, use AI the best you can through every step of that process. Right?

Leverage AI to gather data. Right? Leverage AI to synthesize insights. Leverage AI to come to a problem statement that's meaty, that's specific, that's focused. Right?

It's not just, hey. My user has a hard time processing invoices and needs to get paid. It's no. It's this type of person that fits this mold, that does this every day, that feels like this, has this problem to solve. And if they don't solve it, this is the impact, and this is what we're we're gonna drive value to.

Right? And there's a lot of AI is not just in the solution. Right? It's not just putting out a product that leverages it, but, like, time is money. Right?

And and people are money. And there's a lot of tools and solutions out there. But if you don't if you don't have a a a strong problem, you have conviction around that's been validated quickly by the market, the customers, the people you're trying to you you're trying to bring it to, then then stop there. Right? Like, find something else.

Because you're gonna waste a lot of money, and then you're gonna realize that you had the wrong problem statement because either you went with gut versus data or you didn't validate with with real people and you made assumptions. And to your point earlier, you know, the 4th rocket doesn't fly. There goes the company.

Todd Gagne

We're done. We're done.

Brian trable

Yeah. Well,

Todd Gagne

good. Well, Brian, this has been fun, man. I, I I enjoyed working with you. I miss you, brother. It's it was, we had some good adventures, trips around the globe, doing some product management work together.

And, it's good to connect. So thanks for taking the time and and sharing your expertise.

Brian trable

Always, man. Anything for you. I look forward to the next time I get out there to Rapid City, and, we can we can hang out in person again.

Todd Gagne

Yeah. That's good. Okay. Well, thank you.

Brian trable

Awesome. Thanks, Todd.

Todd Gagne

If you enjoyed the podcast today, please just take a moment to, like, rate it or comment on it for us. This feedback really helps us, and it helps us get the word out to, like, helping other entrepreneurs and founders. Thank you.