Episode 68: Joseph Hanna of Pivoting Model Partners 
on Product, Pivoting, and AI as Agents

Episode Transcript  

 Shiv: Alright Joseph, welcome to the show. How's it going?

Joseph: It’s going fantastic. We're just talking about the weather. So yeah, come down to Miami, dude.

Shiv: Yeah, definitely plan to. And for the audience, why don't we start with your background and where you're coming from and we'll take it from there.

Joseph: I appreciate that, Shiv, and thank you for having me on the show. I know I was scheduled this almost a couple months ago, and I've been looking forward to it. And you know what? It's actually the topic that we're going to talk about today. It's how amazing things change and how quickly things happen in our world. Two months, likely, we would have talked about something completely different if we talked eight weeks ago. So it's interesting that way. So I've been in the software industry, technology industry for about 30 years or so. My background is software and robotics. I started my career with Oracle and some other Fortune 500 companies. And then from there moved to PE and VC backed companies for the past 20 some years where I helped lead product growth, ended up finding my own company, founding my own company in the mid 2000s and we took that through successful fundraising and then exit to Carlisle and EQT. After that, I spent four years heading strategy, IT, innovation, product, tech, all kinds of interesting things backed by those large PE firms. And then last year, I started my own consulting firm where we help companies with pivoting. We help companies think about innovation. We help companies think about how AI is going to impact their path forward. So all very timely topics.

Shiv: Yeah. Yeah. And why don't we start with the pivoting side because you've written a book on this topic called Pivoting as a Way of Life. So as you're helping these companies think about it, like you have slightly different take than let's say the lean startup model or a lot of the nomenclature or traditional thinking around you pivot till you find a good business model and then you kind of scale it. And I think you have a different perspective than that. So why don't you expand there?

Joseph: Yeah, no, and I appreciate you starting with this because it's very applicable to the topic of AI and other innovation topics. And you described it really well, right? That the lean startup models or other models out there usually are people follow them in a fantastic way in the beginning stages of building the company, right? So you are discovering, you're innovating, you're looking for that product market fit, you're looking for the right price, the right product, the right marketing channel. And then the thinking is you get to the point where you achieve product market fit and companies start to move from a pre PMF to post PMF model. So now we're scaling. Now we are changing our marketing from product marketing, which we focus on a lot early on to brand marketing. We're investing more in our sales and less in our innovation. We're not treating, we're not innovating. The founder starts to spend a lot of their time in things that are not customer-oriented, not product-oriented. And what I'm advocating is we have seen what that model does. You know, kind of the statistics talk for themselves, right? Nine out of 10 startups do not make it at all to be still to any sort of successful exit. Eight out of 10 startups that raise a seed round do not make it to raise an A round. So that's pretty pathetic. Then Pendo, the product tracking firm, issued a very depressing, maybe is the best way to describe it, statistic that a couple years ago, and now it's even worse, they showed that over 80% of the features that we're investing in, people are either rarely using or never using. And they translated that to close to $30 billion of investment in SaaS features that our customers do not care about. They just do not care about. And I like to use this analogy and this other data point. The Institute for Cancer Research in the US, which is the leading research fund coming from the government, their annual budget is about $7 billion.

So we're spending more than four times of what the leading cancer institution is spending on cancer research on things that are collecting what I call digital dust. And what I'm advocating is we don't stop at product market fit. We actually try to remove the term product market fit from our thinking completely because it's impossible to measure. It's usually happened in the past. By the time you realize that you have it, you're already ready for a pivot and change. And we move from that to a continuous innovation model. And to do that, you need a way to do it. You need an operating model. So we introduce what we call pivoting life operating model, or PLOM. We introduce a way to allocating your capital and how to measure how much you're copying and how much you're actually innovating and how much you're spending on keeping the lights on and what we learned from companies that we worked with and some others in the public sector on what works and what doesn't work. So the idea is Lean doesn't stop at PMF. Take Lean, we developed an executable model of Lean and continue to do that literally and definitely. And I'll just finish with this. We know the very popular ender-seeing code, the product-market fit is the one thing that matter, right? 2007 kind of changed the way a lot of investors and lot of founders start to look at businesses and kind of led to a lot of literature and lot of, basically a whole generation of investors and founders following that. And we challenge that and we change it to the only thing that matter is really your ability to take an idea to cash, repeat, and how fast you're doing that. So when we help PE firms, for example, assess acquisition targets of their own portfolio companies, we look at the results. But the one thing that we look at is how quickly are they iterating, how adept are they in changing, and how to measure that. So the one thing that matters is your ability to change, your ability to innovate, how quickly you're doing that, looking backwards to see if you achieve product market fit or otherwise is a waste of time.

Shiv: So I want to touch on something and there's a lot of threads you opened up that I do want to touch on. But the first is just this idea that a ton of features are being developed that are unused. I completely agree with that. And just to use a parallel from marketing role, like, you know, it's a common marketing saying that 50% of marketing spend is wasted. Just people don't know which 50%. And I think product development is kind of similar in that, like, yes, you're developing a lot of features that people don't use, but when you're developing it, you're unsure of which features will actually lead to users to be more sticky or come back. And there's tons of these stories, right? Where companies like Slack thought it was a, it was a gaming company. And then suddenly the chat module that they developed for the game turned out to be the product, right? So was that wasted development or not? I don't think it is wasted development if it gets you to Slack eventually, right? But then at the same time, there must be like hundreds of features inside Slack that aren't used. So talk a little bit about that is just the focus inside products. I'll give you, I guess maybe the flip side of it just to set this up is I totally agree that every product, once you have some idea of what the customer cares about and what they value, there's a very limited and narrow feature set that really is the deciding factor between whether they buy, whether they stay with you. And then everything else is on the periphery that you think is going to drive purchasing decisions, but it's really not. Or not even purchasing, but just usage even. Usage and stickiness and sharing or working with other people or social components and things like that. So that's kind of all my notes on it, but I'm curious what you think.

Joseph: Yeah, it's again something that we can talk about for hours, right? I wrote 400 pages on it and we can write another three volumes on the topic. I would say there are two things that we advocate that companies use to be more deterministic and have higher probability of getting the traction, getting that scale, getting that feature adoption rate that we advocate as one of the most important metrics that you should be tracking all the time. The first is a relatively simple but adapted metric, the two by two metric of the differentiation of your feature set versus how much customer perceived value in that feature set. And we expand the definition of feature, Shiv, to not just a feature of the product, but it's a feature of how you're delivering the product, a feature of how your team is structured around the product, your pricing is a feature, your marketing strategy is a feature, where you position yourself in the competitive state is a feature. All those are features. We define 20 of them that you should be always looking at as features. And then you look at where you're differentiated and how much customers care about that. And you start with what you have today. You start with what your competitors are. You start with where you think and perceive that the customers are seeing the value in your product. And you usually going to end up with few features, and again, go back to our definition of feature, and very highly differentiated and very highly valued by a customer. And that's your, we call it the kudos quadrant. That's where your innovation, your differentiation, your moat around your product is going to come from. And then you go around the quadrants and then there is what we call high IQ copycat, where it's a commodity feature. A lot of people have it, but customer places high value on that. And those are important. You need those. Those will allow you to win deals. These will allow you to retain the customers that you have today and continue to defend yourself against competitors. 

Shiv: And maybe even prevent you from losing deals.

Joseph: Yeah, exactly. Exactly. And then what we call the oops quadrant or the low IQ copycat. And that's where it's not differentiated and the customer doesn't care about it. And unfortunately, a lot of companies, a lot of investments as per the statistic we just mentioned, a lot of features ended up in that. And then the last quadrant, which is actually a very important quadrant and applies to that Slack story, is what we call the ivory tower quadrant. And that's a very differentiated feature, but your current customer, your current personas that you're targeting, did not place a high value on that, right? So that's a quadrant. When you invested, you created something that's really amazing, really differentiated, really defendable, but the customers that you're working with today do not care about. And that quadrant has the blue ocean strategy. We all know that strategy. That's your blue ocean quadrants. And you start to look at everything that you have on those quadrants, not just for where you are today or where your competitors are, but you start to do that quarterly. And every investment that you're going to make, you proactively put it on that little, you know, it's a simple tool, but it's very, very, it's not as easy to implement as people may think. You start to put your roadmap investments on that and see how it's going to change your center of gravity, if you would. 

Shiv: You're saying the blue illustration strategy is in an area where it's defensible, you have a moat. I'm very familiar with the book, we talk about it a lot here, but you're saying it might be features that your customers don't care about or aren't aware about. Like help me, help me understand that piece.

Joseph: Yeah, that's a good point. So that's your current target market. Your current personas that you're going after may not care about. It's a very, it's an amazing innovation, but your current target market doesn't care about. And what you can do is again, borrowing a chapter from, Blue Ocean Strategy, start to find other markets who may be more interested in this, who this may create more value for you in other markets. And I'll tell you just a quick story about, about this and my startup, we were in the, in the recruiting AI space. We, you know, we used AI before it was AI, right? 10 years ago, completely different for what we have today, but we created a feature set that was very, very advanced in, in analytics and supply chain, supply and demand analytics and understanding competitive intelligence around, around recruiting and so forth. And as we went to market with this, we discovered that our target market, original target market, which was recruiters, didn't really care about that. That was a different target market. So we didn't throw that away, but we changed who we target with those features to be the right target market for it. So that's where some of those ivory tower quadrant features will lead you.

Shiv: Yeah. What about innovation for your existing ICP where you already have some sort of a moat and differentiated advantage and you know they care about certain features or capabilities that you can build or you haven't built yet.

Joseph: Yeah, so the way we help companies kind of think through that is when you are fully operational, you have customers and you have revenue regardless of what size that revenue is. You have a certain, and we define mature as you have a level of customer retention that shows that customers care actually about your product. Your product reached a point where there is actual retention and you can point to that. When you get to that point, we advocate that a third of your R&D spend, a third of your messaging, a third of your investments should go in innovations or pivoting. And pivoting for us is continuous innovation. It's not a bad, hairy situation because we're doing something wrong. It's an ongoing and continuous innovation. A third go to that. A third go to what we call the copycat metrics. And that's basically things that customers see that you need for retention or you need for winning new deals that if you don't have that feature, you're not going to win these new deals. So we don't throw away copycat. Copycat is very, sometimes it's just as important as innovative features, and that's a third. And then a third go to, again, that's later stage. A third go to technical debt and keeping the lights on. And that's just as important as the other two. And what we observed, and there is bunch of examples that we mentioned in the book, if your investment in the copycat is over 50% for an extended period of time, you're going to be a commodity product, a commodity company, you're going to hit that cliff very quickly. So yeah, we don't say it's bad, but we say it's deterministic ratio of our investment need to go there and we need to be methodical and intentional about it.

Shiv: Yeah, yeah. No, that's a great point. think one of the mistakes that I've seen companies make, and we talk about it a lot all the time too, like we're a management consultancy. So we have private equity clients and we're doing work for them either post-close or doing diligence on the marketing side. And we often get pinged and asked like, do you do agency work? And we say no to that all the time because there are thousands of marketing agencies and it's a red ocean and it's totally commoditized and the work that we do is super strategic and there's no one like us in our space. And so we have complete blue ocean and complete differentiation and our clients love us there. So we always say no to that kind of business. And if every time we try to or think about going down the path of starting an agency, I kind of kind of play out the investments that we'll need, the processes, the people, the systems, the tools. And I can easily see that destroying our core business, even though it feels like a great revenue opportunity for us. We talk about in the, terms of like software, the same principles apply. Like if you're, let's say you're a proposal management software or a contract management system, you might need some CRM components, but you don't need to become Salesforce. There are thousands of CRM tools there and a few behemoths in the room. So you try to go down the route of having every CRM capability inside your platform. You're very quickly going to look like those platforms but not have all the advantages of those platforms. And so very quickly you will kind of go into decline. So I see a lot of companies kind of take that route and make those mistakes and they kind of have to reverse and kind of draw back to some of the things that they've invested in, often losing a lot of money in the process. .

Joseph: Yeah, real quick, I'd say that the example we use for to teach that or to bring it home is the Ford versus Ferrari movie, if you familiar with that movie that took place. True story took place back in the 60s. And it was Ford attempted to copy a chapter from Ferrari's playbook, which is, you know, when long term endurance races. And they literally said, we need to think about it. We need to think about our market the way Ferrari thinks about our market. And that's bad copycat. Ferrari is actually the most valuable European car company right now in spite of selling only 14,000 cars last year, that's that. Ford is something completely different. So what works for your competitor may or may not work for you. And I say, just your example, if you're in marketing technology and doing something there, you're going to get a lot of requests for CRM type features. If you're in talent acquisition, you're going to get a lot of applicant tracking system, ATS type feature requests. If you were 100 years ago in the early days of automotive industry, you're going to get a lot of features from horse and wagon type features. So understanding that, again, there's some of them that are really smart. The horse and wagons had wheels. We needed wheels, right? So that's a smart copycat. But then there is also the commodity copycats that you shouldn't go there. And we call it the death trap. Once you start to go from your core differentiations to the high IQ copycat, and then you run out of ideas, you go to the low IQ copycat, and then you're definitely going towards the commoditization.

Shiv: Yeah. And then you're competing on price and costs and your margins erode and over time your business declines. So all of that makes total sense. I want to transition this conversation to the areas where you need to invest more and have a blue ocean strategy. And we talked about AI and that being a component and a lot of conversations about agents and agents replacing SaaS over time. It's actually kind of connected to all of this because in a future where agents are replacing SaaS, we don't have features. You literally have an interface with a chat bot or something that you're interacting with, and it just gives you an answer. And you never have to interface and click around and look for a report or set something up because the agent just kind of does it for you. And that means far less development on the feature set side. And then also, your platform needs to be focusing on a very core set of customers. So I want you to kind of talk about that and along with that, just a secondary idea is there's a book that I'm a huge fan of, it's called Profit from the Core. We reference it all the time and it talks about this idea of identifying your core business, strengthening that business and only then expanding and expanding to adjacencies that make your core even stronger more than a random adjacency, let's say. So how do you balance that with the idea of doubling down on core customers versus new innovations like AI and kind of just both of those questions. I'm curious what you think there.

Joseph: Yeah, again, another great point and an area that we can spend a lot of time on. That concept is not that different from in 2010, 2011, the concepts of Nail It Then Scale It, right? It was a great book and it's a methodology that is unfortunately kind of re-treated the importance of product market fit and then investing in your strengths. Investing in your strengths, strengths, strengths instead of trying to overcome your weaknesses that the market is perceiving that. And we, what we advocate is balancing the two. So that's why we say 30% need to be innovation, need to be pivoting, need to be looking forward to what's next. 30% need to be strengthening where you are today. And then 30% need to make sure that you actually stay, stay agile by handling technical debt and so forth. And, and, and the innovation doesn't mean that you go away from your core customers, right? You may be serving them in a different way. You may be approaching the problem or your own operations in ways that those customers have not even imagined on their own, right? So that's where the innovation and continuous pivoting doesn't mean, you know, relieving your core. It actually means strengthening, hopefully strengthening, where you are today and creating even a stronger moat around yourself. And it's one of those things where if you don't innovate, you're going to be innovated on, right? If you don't transform, others will transform on you and customers may or may not be your customers tomorrow. The idea of where to go with AI, for example, is exactly that. So AI is transformational technology. I cringe when I hear people saying this is just like cloud or this is just like what happened to mobile or the web. And we go through those big attractions that happen every 10, 15 years or so in our industry and say this is just another one of those. It is not. To me, this is as transformational as moving from analog to digital or what happened to Kodak, right?

That's the kind of transformation we're talking about here. We're not just talking about developing mobile applications and meeting your customers where they are living every day. And that's the problem that many companies that we work with see, you know, suffer from today. They are thinking of AI as another IT implementation or another tool that they're going to adopt and experiment with that will improve their operational efficiency or improve their customer acquisition costs and so forth. Smart companies, companies that are really going to shine in the next decade are companies that are going to really take advantage of AI as a way to transform what they do. Transform the way we operate, transform the way we deliver, transform the way our customers see us, transform the way we develop. I say this, if you are a sub 10 million ARR, SaaS company today, that's your golden opportunity to go transform into where AI can help you. If you are sub 50, then there are segments of what you do that you need to invest in very heavily. If you are 100 or more, then you're starting internally and then turning your products, right? But if you don't do it, native AI competitors are going to eat your lunch tomorrow. It's interesting, talk about big companies and how they are starting to think about work differently with AI. Moderna just changed their chief human resource officer role to be the chief human resources and digital. So they are starting to think of digital as part of the resources, as part of the workforce that's going to be shown to work every day to do tasks and functions and roles like we do it. So if you think of AI as a tool or AI as a co-pilot, you're likely in the very early stages of maturity of understanding what that is. If you think about it as an agent, you're just hitting it in the right time. If you start to really think about it as a colleague, as someone who's going to show up to work next to Shiv, next to Jo, to be able to deliver pieces of work, then you're ahead of a curve today as it stands in 2025.

Shiv: And so help me understand that. So I think the part of the question on the innovation around AI and how it's going to change the feature sets that these companies need to develop, because in an agent-driven world, you don't need as many features. So help me understand that a bit more.

Joseph: Yeah. So it's interesting that the term B2A started to show up in the business literature a little bit more and more and more people are talking about it. that's going from building a business to consume B2C or B2B, right? I'm now building a feature set or products that are B2A, that they are built for agents to consume my product or to buy my product or to redistribute my products, right? So that's how you take pieces of what you do today that are intended for human usage and change them to an agent target customer is going to be very interesting transformation in the next few years here. And you really need to start thinking about this today. So agents are going to become your customers or, depending on what you do, they're gonna they're gonna be part of part of you selling to they're going to represent procurement departments and enlarge and large companies They're going to represent you and me on on ecommerce websites. So building B2A pieces of what you do becomes very important and then internally, how you're developing and how you're building your own workforce? Going to change as well because now you have those agents who can go do a whole bunch of work, not just analytics, not just crunching numbers, not just showing you charts, but autonomous pieces of work where they can make decisions on your behalf, is going to be a big deal. So when you start looking at your workforce, you're to say, OK, I have this job description, but that job description now or role definition needs to be divided into what's going to be done by a human and what's going to be done by an agent. Jensen Huang from Nvidia had a great quote recently that I really like. “IT departments are going to become HR departments for agents.” They're going to recruit them, onboard them, promote them, measure their performance, offboard them, and put them in a team, and they will become part of what we do every day.

Practically speaking, you have to look at every division or every category of what you do today, whether it's a product feature or something that you actually perform internally, and start thinking about what of that, how can that be improved, how can that be changed, how can my customers get more value with me introducing AI agents and not necessarily cutting costs. We actually see the most successful companies are not cutting costs because of AI, they are freeing up resources to expand, to do something completely different, which is a really cool way of thinking about it.

Shiv: What about, and that's great. And I guess one question I have is like, I get using agents and AI to make your internal processes more efficient and customers are going to do that and companies are going to do that as well. What about, and I'll give you an example of something that we're working on. So, and I'll follow that up with this question is, so a bulk of our work is services based to the private equity investor, but they're high priced engagements. Our minimal offering is a due diligence offering and that starts at 25,000. And one of the things that we thought of is there's a large market of private equity firms that needs marketing advice before they've signed an LOI or they're just doing initial analysis on a portfolio company. And so we are developing our own AI platform that allows them to self-serve and get an initial diligence report or an initial analysis on their portfolio company with our own proprietary training data, which is locked and unavailable on other sources, plus our own proprietary algorithm. So everything is unavailable on platforms like ChatGPT and Gemini, et cetera. So where they would come in and answer a handful of questions and they would get a very detailed diligence readout on any company that they want to look at and for a fraction of the cost. it's almost like we're competing with ourselves and cannibalizing some of our core business…

Joseph: I love that. Yep.

Shiv: The logic is that it's not gonna be to the level of depth that let's say if our full services team comes in there, but that becomes also a lead source for us. We can upsell our other services. And then for the PE firm, they're able to get access to analysis even if they can't necessarily afford a full scale engagement or they're not sure if the deal will close. So my question here is a use case like that. This is an innovation that we've made a big bet on. We've spent hundreds of thousands of dollars into this. It's coming out next month. I think more companies need to be thinking about things like this because this technology is now possible. It wasn't possible five years ago. So talk about the role of agents there and even in the world of software, how adjusting the product offering and leveraging agents as a way to expand markets, bring down costs, increase the value you're providing, make things more accessible and then things like that.

Joseph: Yeah. you hit it on the head there, Shiv, because you clearly are introducing a new product that is AI native product, right? And you're freeing up your high IQ human capacity to go do more strategic work or interpret what's going on, what the AI has generated, or help companies implement those recommendations, for example. So that's a great example of an innovation that expands your market, that expands who has access to the amazing IQ that's sitting within your four walls, right? And they get it at a lower rate and they can do it on their own, whether they are on a plane or somewhere they don't need to meet you, they don't need to provide data, they don't need to do all of that. That's a great example of you likely putting a bunch of companies out of business who do that lower end work, right, and they do it using human capital today, or do it using publicly available ChatGPT-like or open AI type LLMs. I introduced to the world the term that I called Large Product Model LPM, right? Which is the LLM idea, but you adapt that to your specific product, you train it with innovations, you train it with data, you train it with input that only you have access to, and AI now knows your product really well and can grow from there. And that's the same idea here. There is literally dozens of companies being formed every day that are just using public domain data. And that ship has sailed… 

Shiv: Yes.

Joseph: Because your customer can't likely do the same work on their own with a lower level analyst using ChatGPT and good prompt engineering. 

Shiv: Yeah, we talk about this all the time. It's like, we talk about this all the time. It's like you have ChatGPT Gemini, Claude, all these tools, DeepSeek, et cetera. All these tools have access to publicly available training data. This is what's in Yelp or YouTube or Reddit or Quora. And yeah, there'll be a future where Reddit charges a fee to use their training data. So I can see something like that happening, but in general, publicly available training data is already being consumed by LLMs and they're able to generate outputs on that. That is understandable. But, there are, let's say you're a neurologist and you need to do a surgery in a better way. None of those tools can help you do a better surgery because those tools don't have the requisite training data. Now imagine a company does have access to the history of all the neurosurgeries inside hospitals and the outcomes and the patient data and all of that stuff. And then they put an LLM on top of that. Then that tool would be able to help the neurologist and that tool is impossible for something like ChatGPT to duplicate. So the moat is not the LLM. The LLM to me is the commodity. The moat is the training data and every company has a unique advantage, and this connects to the feature conversation is like your customers are buying something from you for a particular reason and it's not the copycat features. It's something that is unique to what you offer and so you have a unique mode around that offering and so if you build a training data set around that and you have a proprietary process which is basically an algorithm executed by humans, you can get the AI to duplicate that. Now you have something that nobody else can duplicate and you can almost create a new product for whatever your market is and find a way to penetrate at a cheaper cost without really any viable competitor because the way other people are doing it today is either with commoditized tools or in a much more expensive way with humans.

Joseph: I could not agree more. And that's why kind of the next iteration of AI is the vertical applications. So the foundation, that is done. And anyone investing in more foundational work right now or training AI and machine learning models from the ground up, they better have a really, really good reason why they want to invest money in that today. That's back to the cloud computing analogy, right? No one is building their own cloud anymore. You're using one of the publicly available clouds and you're comparing them and dealing with security and dealing with all of that. The same thing with AI foundational models right now. They will continue to improve. are billions of dollars being thrown into that. If you see yourself or you get a request from your team to invest into something like this, stop. Please stop. And then you get to the vertical knowledge and the vertical applications and building those agents, training them based on your proprietary data and your own product and your own expertise and your own customer data as well. So that becomes kind of the next level, right? So you have the foundation, you have your own data, and then you have the data that your customer brings in to make that fairly unique to their experience, to what they are getting out of the product. And that's the next iteration, and that's where everyone need to be focused right now. So if you are a large company looking to invest into new products around AI, that's what you need to focus. If you're a startup today, that's a hundred percent where you need to be investing your capital.

Shiv: How do you think, we started this conversation with product management and AI is one of those things where in order to do some of the things that we're talking about, there's like a, I don't want to say distraction risk, but there's like this division of resources where you have business as usual and you have this new area of business that needs funding and that needs to come out of your core business that is profitable or that has the resources and this new area hasn't started generating anything yet. So you kind of have to divide up your resources between these areas and pivot again, connected to what you've been talking about to a new version of the company. And you're putting, basically making a bet. And I can say just from firsthand, we experienced this all the time, like with this new platform that we're developing, it's called Inquisio. And as we are investing more in it, like we have the core business that's super profitable and that's able to fund it, but we have this distraction risk of the core business needing more resources or people, whereas a new platform needs development and its own, it's its own beast that needs to be fed, right? So how do companies balance priorities like that against each other to make sure that they're sustainably growing and have a chance at seeing both of those areas of the business thrive?

Joseph: Yeah, and that's not different from what we have been doing for decades as product managers, right? Or product leaders or entrepreneurs. Everyone has limited resources. I don't care how big you are. I don't care how much funding you have. There is limited set of resources that is available to you, whether that's compute, whether that's human capital, whether that's data, right? And time at the end of the day is the ultimate resource. And you make decisions every day. Some of them are bets, some of them are based on guts, some of them are based on actual data to try to balance that, right? And there are two things I would say here that we use or we introduce that are somewhat new but have deep roots in resource allocation. We introduced an engineering term called the ideality equation, and that comes from TRIZ methodology that's pretty big in manufacturing and process engineering, but hasn't been adopted in the digital world very much. And the ideality equation is basically, it looks at every decision you make as a contradiction. You are making a contradiction. And the equation is, here's the benefits of the decision you're making, divided by the potential cost, how much it's cost you, hard dollars, resources, time to do it, plus the harms. And the harms can be things like, other products or other areas of your business not getting that resource allocation. Or by developing this, you're cannibalizing. You said that yourself, right? That you're taking part of your businesses and basically cannibalizing that, so you're going to lose this. Every single decision we make in business today, it should be measured by ideality and the goal is to optimize or maximize ideality. So this is no different. The big difference is how you think about those AI investment decisions. And if you think about them in a traditional implementation, here is the potential benefit, here's how much I'm going to save, here is how much my people will be more efficient in doing certain tasks. And you stop there, then you're limiting how much advantage you can take from that transformation. But if you think about them as a market expanding opportunity, as a differentiation, as a leadership position, as ushering in the new era of computing, then that's where things start to be very interesting. Again, the resource allocation is as old as business. And balancing the two, the difference is thinking of this as true transformation, not just an incremental evolution to what you have today.

Shiv: Yeah, I really like this point that you made is that's basically what product management has been forever. When it was mobile or cloud or any of these things that have come up in the past, like, or you just the internet, you have to make this decision that there is a wave coming and for business continuity purposes or to shift into the new world, you have to make a major investment and transform the business. So I think that's a great point. And that's a good place to kind of bring this to a close, before we do just what are your final thoughts for people that are kind of considering this that are on the cusp? I know, even today I got a question from a friend saying like, Hey, when did you decide to make this big bet on AI and kind of build your own platform? Like, and I kind of walked him through my logic, but founders and entrepreneurs and people just in general that are investing and thinking about making this kind of a bet. What would you say to kind of encourage them on that path?

Joseph: Yeah, great question. So I will go back to something that we mentioned earlier, and that's the third, third, third, right? So you're not betting the farm. If you're an ongoing business, you're not building something from scratch that's AI native and I'm starting a new business. That's a different equation. But if you're a going concern and you have a business, you have revenue, you have customers, you have investments, you have capital, you're going to go back to that third, third, third, whereas a third is keeping the lights on, staying agile, making sure you handle your technical debt. A third is to introduce features for your current customers and acquiring customers they're looking for. And the third is an innovation. And that third of an innovation, what we're advocating right now is majority of it should go into AI-related projects or you're to fall behind. So that's where the balance is going to come. So that's one, you allocate your capital. 

Second is where you start, where you start in an AI transformation, because there is always that question of, waiting a strategy? I'm waiting until the dust settles. I'm waiting until we have clarity on who's going to be the winner in the foundation model. So we can go there. Waiting when it comes to what we're dealing with right now is not a strategy. If you wait, you're really going to miss the train. 2025 is the year to start for a number of reasons. One of them is we're not going to have a huge increment. We're not going to have another chatGPT moment in the next 12 to 18 months. So it's your opportunity to catch up, if you would. So where to start? You're going to look at the various areas of your company. So you have your finance, your operations, and that's you should be piloting some of those tools, but don't commit just yet. You have HR, compliance, safety, let's put those in kind of one bucket. That's an area where you should start to pilot and make decisions on areas where you're actually going to use that. Then sales and marketing. That's an area where you have to be very, very cautious because it's very easy to acquire a chatbot or a customer services bot and train it on your data and assume that that's going to, you know, gonna help you, it's gonna reduce your cost, but you need to be very cognizant of the impact to your brand, the impact to your customer experience, very similar to what we did when we started to outsource on offshore a lot of our customer services, right? So think about that when you're thinking about a sales and marketing type approach. Product management and strategy and discovery is what I advocate to be the place where you should start this year. It's an area where there is a ton of data, your, I can assure you, regardless of how good do you think you're doing in that area, you are underserved in the resources they are allocating to that and what you're getting out of it. There are data points and feedback and customer input and tickets and market research that you're not getting to. So there is an opportunity there. It changes pretty fast. And then it's fairly complex and it's also less reliant on sensitive data. So you're not dealing with data about people, data about financials that can get you in trouble with an AI model that's not trained very well. And the last thing maybe that's maybe soft but really important, people who are in that field, product management, product strategy, are likely the most change-hungry people that you have in your organization today. And that's a big deal. You need to focus on area where people are going to not be afraid of change, but actually welcome the change and go for it. So product management, strategy, discovery, research is an area that we, I believe all companies should be investing pretty heavily in and start this year.

Shiv: That's awesome. Joseph, thanks for coming on and sharing all your insights. And before we jump, like if people want to learn more about you or get in touch, what's the best way to get a hold of you?

Joseph: Find me on LinkedIn. That's an easy way. I'm going to put a plug for my book here. Here's what it looks like. It's got two pretty high praises and a top 50 bestseller in its category on Amazon now for the few months that have been out. So we'd love to hear from you guys. We'd love to continue those debates and discussions. It's a very exciting time.

Shiv: That's great. And we'll be sure to include that, the link to the book and your profile and everything else in the show notes. And with that said, Joseph, thanks for coming on and sharing your expertise. I think, you know, normally we don't talk as much about product and AI, and I think this is a really illuminating conversation for a lot of folks. So appreciate you coming on and sharing your wisdom.

Joseph: Of course, I appreciate you having me, Shiv.