Steven Forth is Ibbaka’s Co-Founder, CEO, and Partner. Ibbaka is a strategic pricing advisory firm. He was CEO of LeveragePoint Innovations Inc., a SaaS business.
In this episode, Steven explains how AI can significantly enhance software customization and pricing, enabling more precise value delivery for each customer. He emphasizes that while AI introduces new capabilities, it does not fundamentally change the core principles of pricing.
Podcast: Play in new window | Download
Why you have to check out today’s podcast:
- Dive deep into the comparison between AI and the transformative impact of SaaS, providing valuable insights into the future of software pricing and delivery
- Understand value models and how AI can optimize pricing strategies
- Explore how AI can enable highly customized software configurations and pricing, making it easier to deliver and capture value tailored to individual customers
“Before they invest a lot of time and effort in developing or pricing their own AI, find out what their customers are already doing and how they’re thinking about the problems.”
– Steven Forth
Topics Covered:
00:47 – Skepticism as well as optimism regarding AI and addressing the hype surrounding AI
04:15 – Exploring how AI might impact software pricing compared to the changes brought about by cloud-based SaaS.
07:05 – The way AI can optimize software configurations and pricing based on individual value drivers for each customer
10:44 – Discussing how SaaS fundamentally changed software pricing compared to AI
14:11 – How AI, while it enhances capabilities and user experiences, does not fundamentally change the underlying principles of pricing
16:31 – What is a value model?
18:15 – Pricing being tied more closely to value with the rise of AI
23:27 – Steven’s best pricing advice
Key Takeaways:
“The underlying ways that we think about pricing have not changed because of generative AI and will not change, but some of them we’ll be able to execute on much, much more effectively.” – Steven Forth
“I think of pricing not as being something that is data-driven, but something that is model-driven.” – Steven Forth
“AI is changing how we build, deliver, and experience the functionality that software and data can bring. But as it does that, it also opens new ways for us to think about the pricing of the applications and the data.” – Steven Forth
People/Resources Mentioned:
- Pros: https://pros.com
- Microsoft: https://www.microsoft.com/en-ph/Don Norman: https://www.nngroup.com/people/don-norman/
- Jacob Nielsen: https://www.nngroup.com/people/don-norman/
- Totogi: https://www.totogi.com
- McKinsey: https://www.mckinsey.com
- Tom Nagle: https://www.linkedin.com/in/thomas-tom-nagle-2bb2288/
- Zilliant: https://zilliant.com
- Pricefx: https://www.pricefx.com
- Vendavo: https://www.vendavo.com
- Tomasz Tunguz: https://www.linkedin.com/in/tomasztunguz/
Connect with Steven Forth:
- LinkedIn: https://www.linkedin.com/in/stevenforth/
- Email: [email protected]
Connect with Mark Stiving:
- LinkedIn: https://www.linkedin.com/in/stiving/
- Email: [email protected]
Full Interview Transcript
(Note: This transcript was created with an AI transcription service. Please forgive any transcription or grammatical errors. We probably sounded better in real life.)
Steven Forth
Before they invest a lot of time and effort in developing or pricing their own AI, find out what their customers are already doing and how they’re thinking about the problems.
[Intro / Ad]
Mark Stiving
Welcome to Impact Pricing, the podcast where we discuss pricing, value, and the generative relationship between them. I’m Mark Stiving, and our guest today is the one and only Steven Forth. He’s been on many, many times before. so we all know Steven. Welcome, Steven.
Steven Forth
Delighted to be here, Mark.
Mark Stiving
So many of you probably heard Steven and I on a webinar. We recorded a little over a week ago. By the time this comes out, it’ll be a couple weeks probably. Hopefully you enjoyed it as much as I did. but one of the topics that Steve, well, actually before I give you the topic, I want to just address something, Steven. I am not a skeptic of AI. I am more skeptical than you are, but I am not a skeptic.
Steven Forth
Given that you’ve gone to the effort of training an AI with your own writing, I’d have to guess you’re probably not a complete skeptic. Well, you don’t dismiss AI. One can be skeptical without being dismissive.
Mark Stiving
I think that’s a fair statement. And I would say my skepticism is I’m not all in the way a lot of people are. In all honesty, I remember sitting in a room. The presentation was like, they call it a million cups of coffee or something like that. And somebody in the room said that this was several years ago. He said blockchain is going to take over the world. And I’m like, yeah. And that’s not going to happen. And so I kind of feel that way with AI, that there’s so much hype that it’s got to be overhyped. I guess it’s possible it’s not, but it’s got to be.
Steven Forth
So, like all of these things, right? The short term impacts take longer to play out. but the long-term impacts are larger than we imagine.
Mark Stiving
I think that’s a great statement. That really nails it.
Steven Forth
So, is AI overhyped right now? Yeah, I mean, since it couldn’t be hyped much more than it is, it must be over hyped. but if you think of the hype cycle where in a conventional hype cycle the peak of the hype cycle is higher than the plateau. I think with AI, the plateau is going to be much higher than the peak of the hype cycle. You know, you go through that, it goes up and then goes down and then comes back up. so we’re probably near the peak of the hype cycle for generative AI. but on the other hand, though, the progress in generative AI and AI generally has just been accelerating. So the hype cycle is actually having trouble keeping up with the real progress that’s being made,
Mark Stiving
Which almost says that we have a hard time imagining what we are going to be able to do with AI.
Steven Forth
Yeah. And how deep the changes could be.
Mark Stiving
So, let’s talk about AI and the topic that Steven and I are going to talk about today is AI versus cloud-based SaaS. So cloud-based SaaS came out and everybody said it was going to change the world. And Okay, Steven, to be honest, I was a skeptic and you know, I would hear phrases like, all software is going to the cloud, or it’s all going to be subscription-based software. And I got to say, every company I know of in software that’s not cloud-based is trying to figure out how to get there. So, yeah, that was absolutely a true statement.
Steven Forth
It did happen. It just took 20 years.
Mark Stiving
Yeah. So the question that I’d like us to chat about is, SaaS undoubtedly changed pricing, but is AI going to change pricing that way? Because I have a really hard time seeing that.
Steven Forth
So, I think Mark, the answer is yes. So let’s take a step back and just sort of think about what cloud-based software and SaaS did, because I think that there are three key things that happened. One is that with cloud-based pricing, the rollout of new features is orders of magnitude faster. So I’m not a particularly good chess player but I have a rule for chess where I’m pretty sure I can beat you. And that’s, I get to make three moves for each move you make. And that’s kind of what the cloud software vendors did to the conventional on-premise software because they controlled the entire environment, they could innovate faster, and it took time.
Mark Stiving
Yeah, along those lines, they stopped customizing software and they started saying, Hey, this is what we sell. Do you want it or not? I mean, they’ll configure it, but they’re not customize it.
Steven Forth
and that sort of connects us to, I think the really profound change thatAI is going to bring to software is that all software will be, in a sense, customized, because think of the example that we gave in the webinar, right? Totogi, which is the OSS that can be configured in minutes rather than months. And if you can configure it in just a few minutes, that changes all sorts of things. One, is, you can explore lots of different configurations, so you can get much better configurations because you have more time to explore and try different things. But the other thing is, if you can configure the software in minutes, you better be able to price it in minutes. otherwise the commercial things don’t connect.
I think you’re right, that SaaS forced people to stop doing customizations and rely on a fairly narrow set of configurations, which is why we get things like good, better, best software, right? Because you want to package it up into a simple set of configurations that are easy for people to understand and easy to buy. But what happens if with generative AI, we can rapidly configure based on a conversation that you have with the configuration system. The software really helps you find the right configuration for you. So rather than driving it down into a very narrow range of packages, and I think as pricing coaches, we would probably tell people not to have more than five or seven or 10 different packages. But that now all goes away. So, the whole part of pricing that has to do with packaging could be dissolved by the solvent of AI. and that changes pricing tremendously.
Mark Stiving
Let’s go down this just a little bit. so I always recommend that you start with good, better, best, right? We can’t always implement it, but I get it. But I always recommend we start that way. And the reason is it’s an easy decision for a buyer to make. Right? So, if you give me too many choices, I can’t, it’s the paradox of choice. I can’t make a choice. So I get that completely. Now, if you give me a set of infinite choices, but I’m not having to make the choice. I’ve got an AI that is essentially making those decisions to optimize my value delivery. That’s so far amazing. We step back to the pricing piece of this, and now what we have to do is what you and I have talked about in the past, and that’s truly understand what those value drivers are and how we’re going to be pricing based on what that configuration is going to drive for that customer.
Steven Forth
Yeah. So I think that the job of the pricing function, because I think the pricing function will be fulfilled by software as well as for any specific customer at any specific time, find the configuration that delivers the best value to the customer while at the same time optimizing price for the seller. So there’s a tension there, right? Because the software is going to have to be good at resolving that tension. And some companies will lean more towards configuring this to give me the best possible price. and other companies will lean towards configuring this to get the most value to the user, and then we’ll have more value that we can capture in price. But this is going to be a function that is played by the software and it has to be as capable. The pricing part of this has to be as capable as the configuration part.
Mark Stiving
Yeah. I’m picturing in my mind, almost a slider scale where the user is saying, how much value do I want? And how much am I going to pay for it? And the AI could continue that.
Steven Forth
Yeah. The tension between buyer and seller does not magically disappear. It doesn’t.
Mark Stiving
No, absolutely not. and in the world of good, better, best, buyers are doing that exact thing, right? They’re on that slider scale. They just happen to only have three choices. Not an infinite set of places where they could move their slider to.
Steven Forth
Yeah. But as you say, I think, if you expect the buyer to have to manipulate all these sliders and do the configuration, that’s too much work. And one of the great things about AI is that they help people make decisions, their decision making engines. And by removing that load from the buyer, the buyer is actually able to explore more options with the help of the AI than they would be willing to do on their own.
Mark Stiving
Okay. So, I’m going to say what you and I just said, and that is that in the world where there’s a large implementation and we have to choose a specific configuration out of an almost infinite set of different configurations we could choose, then AI will help us find the optimal configuration. And AI will help us price that optimal configuration. And so I could see that being a true statement, right? I mean, I could see that being the future, but I don’t see that being as big as SaaS is to the software industry, right? I mean, I think that’s true for, I’m going to buy Model N, I’m going to buy Pros, I’m going to buy some highly configurable software. Absolutely, that’s a true statement. but it’s not true when I buy Microsoft Word.
Steven Forth
Yeah. Well, it remains to be seen whether things like Microsoft Word have a future. The other big thing that’s going to be happening here, and this does not necessarily, well, maybe it does, the other big thing that’s going to be happening is the user interface assumptions are changing. So, Don Norman, I think it was Don Norman, it might have been Jacob Nielsen. Anyway, one of those two guys has a model for how the user interaction has changed over the years from command lines which you and I are probably old enough to remember.
Mark Stiving
Sadly.
Steven Forth
…to Windows, to the current way that mobile UIs work, the way that we interact with intelligent agents is going to be completely different from how we interact with our mobile phones today, and how we interact with our laptops today. The whole Windows-based paradigm is likely to melt away. Now, what that implies for pricing, I have no idea. But when the user experience changes profoundly, I can’t help but think that pricing is going to follow on. Mark, I actually think that this is a bigger change than the shift to cloud and the shift to SaaS.
Mark Stiving
The user experience or the whole AI piece.
Steven Forth
The whole combination of things, the change to user experience, the change to configuration, the change to how we develop, the change to how the costs operate. Because the cost of compute is going to be much higher for AI-based applications than the cost of running a SQL query is for a conventional application. So when you consider all these things together, changes to user experience, change to dynamic configuration, changes to the underlying cost structure, the impacts are going to be huge. And I don’t think we know what they are, and I think we won’t know what they are for another two or three years. but it’s going to be a period of really dramatic change.
Mark Stiving
Okay, it’s going to be fun to watch. But here’s my argument. When I see what SaaS did to pricing, what SaaS really did, in my simple view of the world, they took the fact that I got you to pay for my software once, even though I’m delivering value to you month after month after month. And I got to change that to now you pay me every month for my software. And so now it extends the lifetime value. It makes me build better products. It gives me the ability to upgrade you over time. I mean, it gives you tons of capability, tons of options in that whole pricing world. When I see AI, I’ll buy the configuration argument because that really feels different. It really feels like a change. But most of the other things feel to me like a new capability, right? So, back when you and I had a command line interface, by the way, you got to admit the first time you typed in a command and got a response, that was cool.
Steven Forth
It was incredible. And just as an aside, I’ve been studying the large climate models recently. and guess what language they’ve written in?
Mark Stiving
No idea what.
Steven Forth
Fortran.
Mark Stiving
Oh, I should have guessed that. Fortran, PL/1, see something old and ancient.
Steven Forth
Yeah. Because there’s layers and layers and layers of code there. and they’ve been writing for a long time. Yeah. But I digress.
Mark Stiving
That’s good. It just feels to me like, like when we went from command line, if you thought about pricing and you went from command line interface to Windows, that doesn’t really change pricing. It changes the capability of what you could do. It changes the speed and the amount of things you can do. So we can deliver more value, and maybe we could charge a higher price. I mean, we know with Moore’s law, prices actually went down, but theoretically we could have charged a higher price. And so it feels like AI user experience is exactly that. Now I’m going to have a conversation with my computer instead of me typing something in. And so, great. I can do something faster, better. It’s smarter, but it didn’t change the fact that it’s delivering value to me. Am I missing something there?
Steven Forth
No, and I think you said something in the webinar that I think is worth repeating, which is that none of this changes the underlying methodology and frameworks of pricing. So the underlying ways that we think about pricing have not changed because of generative AI and will not change, but some of them will be able to execute on much, much more effectively. So let’s just take the example of value-based pricing. Value-based pricing is most effective when you have a value model. I can remember a gentleman at a large company saying to me, he said, I understand everything you guys are saying about value models, but for my business, I would need over a thousand different value models. Now, I can’t afford to hire the people to build a thousand value models. I can’t afford to hire you to build a thousand value models. And even if I could, your company would not be able to deliver. and nor would McKinsey for that matter. So, for me, this is an academic exercise except for my most strategic products. But if all of a sudden you can build value models, just like you can do dynamic configurations that completely change the playing ground. Now everyone can have a value model, and that value model can easily be configured for a specific customer situation and specific competitive situation.
Mark Stiving
Okay. I think that is a very fair statement. Before I say what I want to say, would you, for all the listeners, define a value model for us?
Steven Forth
So the way that we use value model is we’re using Tom Nagle’s economic value estimation framework. and basically a value model is a system of equations that describe the economic impact that your solution has on a customer relative to their competitive alternative. And when you actually look at these equations, most of these equations are pretty simple equations. They’re not fancy, fluid dynamics things. They’re pretty simple pieces of algebra, but they can interact in pretty complex ways. So a value model is just a system of equations describing the economic impact that you have on your customer.
Mark Stiving
Yeah. And for any value model, when you go talk to a customer, all the underlying assumptions have changed, or at least all the underlying numbers have changed. So you’ve essentially got to customize it for every potential buyer. Some might be relevant, some may not be relevant.
Steven Forth
Yeah. I mean, when we’re doing this right now, you switch value drivers on and off depending on the customer and the competitor. and you dial the variables up and down, again, depending on the customer and the competitor, and we divide the variables into different groups. There’s customer variables, there’s solution variables, there’s external variables. So for example, things like interest rates or energy prices or labor rates, those are external to the customer and the solution, but they can still impact the value. So, I think of pricing not as being something that is data-driven, but something that is model-driven, but I really make those models much easier to build and to work with and to evolve.
Mark Stiving
Okay. Now, I think it makes a lot of sense to say we can use AI to make better pricing decisions. Now, in a lot of ways that’s different from what I think you and I have been talking about for a long time. So, for example I think that Pros have been using machine learning and AI to do pricing for a long time. And so yes, we can use AI to help us make better pricing decisions, and we should, right? I’m not taking anything from that, right? We absolutely should be. but I thought a lot of what we’re talking about is how do we price other applications of AI and is it changing the way we think about pricing, right? I’m a hundred percent in favor of what you just said.
Steven Forth
So, I think these two things are in a dance together, though. So, clearly AI is changing how we build, deliver, and experience the functionality that software and data can bring. But as it does that, it also opens new ways for us to think about the pricing of the applications and the data. So, there’s two partners dancing together.
Mark Stiving
Okay. I’m going to try to repeat that using SaaS as an example. So before we came up with SaaS, you were using my software and you were always getting periodic value from my software. The only difference was I couldn’t charge for that. I could only charge for the lifetime subscription and then the maintenance fees or whatever that happened to be. That’s the way the world was set up. And once we came up with cloud-based software, I could monitor your usage. I could monitor the number of logins, I could monitor a lot of different things.And now I have this ability to capture the value that I was always delivering. And with AI, I’m going to be delivering value, by the way, I think I’m going to be delivering value in new ways, right? So, I’ll buy that, but we’ll also have new ways of capturing what that value is that I’m delivering to clients.
Steven Forth
Yeah. I think there’s probably two big vectors here. One is, I don’t like the term dynamic pricing in this context, because dynamic pricing means something very specific in the pricing world, right? It means what Pros and Zilliant and Pricefx and Vendavo and those guys do. So they have dynamic pricing engines that are really for revenue optimization. So maybe we need a different word here, such as adaptive pricing or generative pricing, to describe this new reality where price is much more tightly coupled with value than it has been in the past. because you can configure more easily and more fluidly and then those configurations are being driven by the value being provided, and then you can price to that more dynamic configuration.
Mark Stiving
Yeah. I have to tell you, that infinite slider that we talked about earlier just resonates in my head. I could see how that really plays a role. And if we can get buyers to understand value that way, then they could choose where they wanted to be on that slider.
Steven Forth
Yeah. And I think we said earlier that this is going to allow us to make better pricing decisions, but I think far more importantly, it’s going to allow buyers to make better buying decisions. And, if we’re being customer centric, that’s actually more interesting because, right now buyers often struggle to understand what they should buy and how they should configure it. and as a result, they choose not to buy because they don’t have the bandwidth to think through all these things. And it’s why we have a good, better, best, right? It makes that decision much easier.
Mark Stiving
Good. Quick story for you. A good friend of mine owns a bike shop. And so, 10 years ago, maybe 15 years ago now, I wanted to buy a titanium bike. I wanted him to sell me a titanium bike. And so it is from a company called Seven. And when I ordered the bike, they sent me this, I don’t know, I’ll say a seven-page questionnaire of questions of how do you ride, right? So, how’s your cornering and how much do you sprint and blah, blah, blah. And I’m just like, I can’t answer these questions because if I answer these wrong, my bike’s not going to be right. And, so the decision, it’s like, I need AI to answer the questions. I don’t want the vendor to answer the questions. Does that make sense?
Steven Forth
Yeah.
Mark Stiving
I need an AI buying solution, not an AI as part of the vendor solution.
Steven Forth
It’d be interesting to see how AI, we haven’t talked about this, but, how is AI going to impact procurement? How will AI change the job of procurement? And, I don’t have a good answer to that question, but I think that it’s part of the puzzle.
Mark Stiving
Yeah. Steven, as always, you’ve got my brain thinking way too hard. But we’re going to have to start to wrap this up. I’m going to ask you the question I always ask anyway, you because you always come up with a good answer. What’s the one piece of pricing advice you’d give our listeners that you think could have a big impact on their business?
Steven Forth
So just keeping us in the context of the AI thread that we’re on. I think the very first thing that people need to be doing is going out and finding out how their customers are using AI before they invest a lot of time and effort in developing or pricing their own AI, find out what their customers are already doing and how they’re thinking about the problems.
Mark Stiving
Yeah, I think that’s great advice. And can I say, Steven, thank you. This is my personal thank you for a second. I find AI interesting, but I would not be thinking about it as much if I weren’t talking to you on a regular basis. So you keep pushing my boundaries on what AI is and does. So, I highly appreciate that. Thank you.
Steven Forth
Well, thank you, Mark. I learned an enormous amount from these conversations and I think that one of the metaphors for the AI shift is that it’s becoming more conversational. And let’s face it, humans learn from each other through conversation.
Mark Stiving
Yep. Yeah. I have to say my favorite use of chatGPT is when I turn the voice on and talk to it and just ask you questions. I love that.
Steven Forth
Yeah, Tomasz Tunguz at Theory Ventures, I think had a post about that recently after GPT 4.0 came out. This is again, going to be transformative. Although I am getting tired of people creating an avatar of themselves and then having a conversation with their avatar, that seems a bit narcissistic to me.
Mark Stiving
Well, I won’t do that. I promise. Steven, thank you so much for your time today. If anybody wants to contact you, how can they do that?
Steven Forth
The best way is by email, Steven with a V at Ibbaka, I-B-B-A-K-A dot com. And I am very easy to find on LinkedIn, and I’m reasonably responsive.
Mark Stiving
And reasonably prolific.
Steven Forth
Probably, accessibly prolific.
Mark Stiving
To our listeners, thank you for your time today. If you enjoyed this, would you please leave us a rating and a review? And if you have any questions about this podcast or pricing, feel free to email me, [email protected]. Now, go make an impact!
[Ad / Outro]
