2025 Pricing Trends

Hello, everyone. So we are now live. And I'm hoping actually sort of getting through to everyone. Thank you for hanging in there for a minute or two. I think we were confused about sort of which studio to join here. So my name is Ulrich. And I think you'll find also Rob and John from Pricing SAS here. And so we're going to have sort of 40, 45 minutes together. on 2024 in SaaS pricing. So here's how it's going to go down. I'm going to speak for around, I think, 20 minutes or so. I have about four points, so major takeaways from how I see SaaS pricing played out last year. And then we're going to have 20 minutes for open Q &A. So I think we have a chat over here. which should enable you to write comments and questions and so forth. So I'll try and check into that one. And you're welcome to write a question as we go along, or you can wait until the end. I'm going to stay around for 10, 20 minutes and then try to exhaust the questions. So if there more questions than we have time, I'll just try and stay along. If it goes completely crazy, then that's where we have the community, which I think is currently still at school.com slash SAS pricing. And I think we might be thinking about replanforming at some point, but I think right now it's still over there. So you can catch up on both the slides. They're not that valuable, but they will be there and the recording of this webinar over at the school community. But that's where we're at. So let me just start us off here. So I have sort of four main takeaways from 2024. And what am I basing this on? Well, it's mainly sort of observation. So I haven't sort of crunched a ton of data to sort of get to these. And also I would think that a lot of other people are crunching tons of data to get to these. So I thought that I'd rather try to...

take a step back and look at how I see the SaaS market play out and a little bit more sort of holistically say, what do I see as sort of the main drivers that are changing things when it comes to how SaaS businesses price and monetize their solutions? So the four sort of areas here are AI. So that was and remains sort of a big thing in 2024. There are some specific elements I want to dive into. Then there's pricing tech billing system so forth. There is a concept that I call commercial debt, which I'll touch and then finally pricing professionalization. So we'll just get started with the first one and then at the end of course we have the queue. So AI pricing dynamics. Basically I think. During 2024 three things happened. With AR and I think if you all think back to when. 2024 started. There was a certain sort of token cost and people were declaring that SaaS was dead because SaaS implies a 100 % margin on the delivery of the solution. And now we have AI and AI can also build the entire solution layer. So essentially you're going to have a cost plus based pricing where whatever value layer you have on top of the AI model is just going to be very superficial. So you can basically charge whatever the token costs plus a little bit. That was sort of the conversation on especially LinkedIn, but also elsewhere instead of Q1 in 2024. And then what happened actually during the year was that the token cost dropped about 97%. So that was of course was radical. And you could see that I think, for example, OpenAI they made sort of two major sort of pricing changes in 2024. And I think they started at around like $5 per million tokens of input. And I think they ended up around 15 cents with the, with the 4-0 model. So, so that was sort of a major change, but then also what happens was that the models got better. So you could actually ask them like more important questions. You can ask them to sort of do more complex and comprehensive tasks, which actually meant that you were now asking them more complex and more comprehensive tasks. So.

Token cost went down, but now the token requirement for the new things that it could do actually went up. And then inside the architecture of a lot of these AI models, you also started to see, let's say, compartmentalized models that could run more or less specific use cases, sort of on device, which is why I call Edge AI. So suddenly instead of having access to like a big cloud hosted LLM, suddenly you could have on-device models which started to make AIs relevant inside of robotics and in certain manufacturing and so forth. And I think these dynamics where token costs go down, but model complexity go up and also model deployment actually starts to vary, creates this scenario where people seem to be disoriented about where is this going to go? How should I price my product in this scenario? And I speak to a lot of high level products, sort of CPOs and also CFOs in a lot of B2B SaaS. And especially at the first half of 2024, we saw that people were very worried that they would lose money. So 2024 was when a lot of companies started to launch AI for real into their products, AI functionality, AI features and so forth. It had already begun in 23, but sort of the big bulk of sort of everyone doing it, I think was last year. And everybody was focused on wealth. We don't know how customers are going to consume this product. And if we just give them like an all you can eat buffet, we're going to go bankrupt because they then control our cost structure. And if we give them sort of just a license price, let's say $10 per user, then suddenly what we're going to see is that they're just going to spend more than $10 in token costs. And then we are losing money. So there was this, let's say defensive play in rolling out a lot of AI products because of the cost structure. But on the other hand, you also saw the product teams were trying to use these better models to solve more problems and deploy them in more different sort of instances and use cases. And I think this is where Yvon's paradox comes in. classic economic model where we say when the price of something goes down, the demand for consuming it goes up. So you think, well, if the price of gas goes down, people drive more miles.

meaning that suddenly they actually spend in total more money on gas. And I think that's what we're seeing with AI is that the more the token cost drops, the more the total spend on tokens go up because suddenly more use cases are now economical. And I think inside of this Yvon's paradox curve, there are a lot of sort of nuance that we saw play out in 2024 and that we're continuing to see playing out in 2025, which is that people don't know exactly how the So if you're currently running a piece of software and it has AI functionality and has a user facing, let's say, endpoints, so users actually control the consumption, you as a product owner or as a CFO have no idea about what the cost is going to be. But I think more importantly, which we also see with a lot of AI products that was sort of spit out in 2024, is that nobody really also understands the value that they're creating yet. partly because we haven't seen users actually adopt these products. So I had a around 100 million ARR US West Coast customer. had this conversation with them Thursday last week. And basically they own, they have large blue chip corporate customers. They have about a million in ACV on each customer, but the total budget for their category is around 40 million inside of these customers. So think US government spends. If some department spends 40 million on this, they are one million of that spent. And they're now rolling out AI. And I think the defensive play is to say, well, we're going to sell the customers some credits. And then they spend the credits and then we'll see what the price is once the costs drop at, you know, in the next cycle, next year, next quarter, whatever it is. And I think you can't win in this AI game if you're playing Yvon's paradox defensively. I think you need to play it offensively because what you need to do in order to win this game is you need to learn exactly how users are going to use AI if there isn't a paywall, if there's not a cost barrier to free usage. Because that is currently the question that nobody really has dared ask because it's going to place all the risk of consumption on you as the vendor where the user gets it out. So I think my solve for this right now

is to say to customers like the one I talked about in the US to say, well, you have 500 customers. How about you take five of them and maybe only divisions inside of these customers, but take a subset of customers and give them an all you can eat model. Just let them have as much AI as they can consume for a fixed price and see what happens. Because they're going to be basically one or two outcomes. Either they actually don't use it that much, which is maybe a product problem, and then you know it's a product problem and not a pricing problem. Or you can say, well, we now know exactly how much we should charge for a fixed fee model. So you de-risk it and then you can spread it out to all of your customer base. So that's basically scenario one is low consumption or you have extremely high consumption, which should also mean that you now either have a product problem because the consumption isn't generating value or you actually have an excellent value use case because all the consumption is creating value above and beyond the cost, which now gives you five of the best use cases in AI you could wish for. Because I think the game right now is to take a 40 million corporate budget, shrink it to 20, but then own 10 of it, as opposed to right now owning 1 million out of 40. And I think this is what AI allows you to do. but only if you play Yvon's paradox aggressively by figuring out exactly how adoption looks if you take pricing out as a parameter. So I think that's what we're gonna see a lot of in 2024 based off of the experience of lower costs and a broader use case base and let's say deployment on what I call H-A-I, which we saw in 2020. All right, that was observation number one. Observation number two, maybe a little bit less meat on the bone here, but I think it's interesting. So this is sort of what I call pricing tech. So this is billing systems and billing systems has been a thing for a while, right? You have Swaror, you have ChargeBee, you have sort of essentially sort of payments versions like Stripe, and these have been around for like more than a decade, a lot them. But I think I've seen them accelerate, especially last year.

where more and more of them, especially to me, are reaching out. And I also see that the way that they're reaching out is not just, let's say, 23 people just said, hey, we have a startup and we want to do users-based billing. And you have a lot of these, and even a lot of these exited already. What I see now is that they say, well, we want to do users-based billing, but we also want to integrate it. into a wider ecosystem of use cases and functions across the organization. for example, we don't just want to do billing. We also want to integrate with the sales functions, with the wider finance function, with rev ops and so forth. And we want to do more things. We want to go all the way into the configure price quote process that sales handles. We want to deal with contracting. We want to do billing and payments and ring conciliation. So we want to own more and more of the value chain that sort of is centered around pricing. And I think that is new and we haven't really seen anyone succeed with this yet, but I'm seeing a lot of startup territory, pouring money and sort of talent and brain power into solving a larger, sort of more comprehensive tech stack inside of pricing tech instead of just like point solutions around billing. So I think That's definitely something that we're going to see more of in 2025 and 26. And I'm going to expect at least a few of them succeeding, right? Which is going to disrupt to say how we are currently stitching this together with a lot of different sort of point solutions. All right, point number three, commercial debt. I think this is a little bit more like the waves of history. So we had a lot of like zero interest rate economics for quite a while. And we saw that with a venture bubble, I'd say in 2021, 2022, especially at the high points. And then there was like a hard stop, like interest rates started to go up.

And we saw that venture market sort of crashed sort of, then also IPO. So we suddenly got pulled back and we had some of the worst years ever in think late 22 on that. And then there was this shift to profitable growth. Like now people actually needed to make money. They couldn't just like funded on, let's say, external capital. They actually needed money from their customers. So, so let's say 2023, I saw a lot of that like, Hey, we actually need to like have profitable growth. But I think in 2024, it shifted a little bit because I think a lot of people then had actually already achieved profitable growth. And a lot of these had done it in let's say in a scramble, like they had pulled together whatever resources they had and become breakeven and become profitable. And a lot of companies now see for and say, so we now are profitable, but also we might not be growing as fast as we want. And we might not have the exit opportunity in the capital markets that we want. Partially because we have a lot of what I call commercial debt. So that means the way that we structure the customer portfolio is with a ton of different discounts, a ton of different pricing models, a ton of different product configurations, contracts and so forth. And it's a little bit a mess, right? So people are now shifting to cleaning that up. And I would especially say that this is true of let's say 50 million and plus AR companies looking at sort of if we need to have some sort of a real exit in the market through an IPO or through a trade sale, then we need to have a much, let's say more scalable customer base, which means that we can't spend all the time managing each individual account like a special snowflake. They need to sort of come together as, let's say a scalable unit of customers that we can then grow off of. And I think, so I've done a lot of these pricing projects in 2024 where we've talked about contract frameworks and cleaning up commercial debt and repricing legacy customers and so forth. And I would say across the board, there is a willingness to let go of some customers, to have some churn in order to have what is left be a little bit more clean and structured, more ICP, and then to be able to grow off of that. And I would say that on the other side, on the customer side, I also see that enterprise procurement and CIOs

also trimming down their tech stack. Like a lot of people bought tech that they really didn't need in 2021, 22, because it was just a little bit crazy. And now they're sort of taking a hard look at what they really need and then opting out of it. So I think actually on both sides, you say that you have a cleanup either on the customer side or you have a cleanup on the vendor side. essentially it's just 2024 was like a year where the market sort of like agree to part ways amicably on some instances, or really sort of double down on what worked and really make that scalable. So at least that, and I'm still seeing a lot of this going on actually inside the 2025, but I'm also more and more now seeing the projects that the customers that come to me say, oh, because we want to IPO in two years, because we have this sort of specific exit target in mind. which wasn't really the case when we talked a year ago, then it was just, we need to clean up because that's the right thing to do. Now it's with a specific end in mind. Last one, pricing professionalism. So this is a little bit longer term trend, but I wanted to mention it here. Maybe for like semi selfish reasons, I don't know. But one of the things that I also saw in my projects over the last year was that more people were asking for internal pricing functions, right? So they're saying, hey, we want to have new pricing and we want to clean up all the commercial debt and we want to do all these things. We would also really very much like to be able to do it ourselves once you, the consultant leaves, right? So we've actually done this where we say, so we then need to sort of hire someone and we need to have tooling in place and processes in place and governance. And we need to sort of figure out how pricing decisions are made inside of the organization. And that has become much more a, let's say an area of focus in a lot of the organizations that I work with. I would say that it is,

most definitely an area of focus once we hit around 100 million AR. I've helped at least three clients hire their first pricing team at around this sort of stage, 100 million AR in the last year. And people are more and more also realizing that pricing isn't just a numbers game. You actually need a more, let's see. often more senior roles here where there are more, let's say, change management focus. They can actually like speak to executives. They can actually, let's say, create alliances inside the organization and move sort of projects or change forward instead of just providing breakdowns and slides and analysis, which might also be necessary, but isn't enough to get the work done. actually very often the way we set up pricing teams and organizations is we hire both, let's say, a senior guy. that can do the change management part and then support it by an analyst that can do the analysis and the number crunching. And I would say that the pay range have also gone up for these to the point where especially the really senior guys that can do something are paid on par with top account reps, like top salespeople in most of the organizations. So it is being valued quite well in most of these organizations. And then one thing that I'm looking for, but I haven't really been able to sort of see it yet, at least not a scale is that the way that people are looking at pricing is still mostly a quant function, meaning that they are asking pricing to solve relatively narrow problems and they haven't yet, maybe they never will, but let's see, given the pricing function is a broader mandate to work across the entire value proposition and to also let's say get get quite involved maybe in product roadmap and sort of product strategy and all these things and and and support in that way. So we haven't seen that yet and that's still creating some friction inside of organizations. I'd say it's a little bit like in for a penny in for a pound type of issue. But but organizations are starting to take it more seriously and realizing it's sort of it's it's a part of the

function stack that you need inside of a healthy organization. So with that, I think those were sort of the four takeaways that I had and we're pretty much on time. So let's see if we can get the chat to work and we can have some questions.

Okay. So the way I'm to do this, I'm just going to start from the top and then see what we get. All right. So by the way, if you open up the chat, it's a little sort of speech bubble to your right. You can see that John actually links to the school community. So you can go there and you should be able to see a recording of this and send it to all your friends and all that. And even if you don't go there to see the recording, You can go there and see, I think, 500 plus other pricing people sort of nerd out on pricing. So that's an easy win. And PricingSAS also has their latest benchmark report from 2024, which they come out, I think, with monthly reports, John, Rob. So that's a really good way also to just like stay updated. All right. So Robin asks, which company I think is the best bet for the 2024 new billing wave? That's unfair because I invested in one. I think definitely Hello Proper is gonna win that game. They're small and they don't have a lot of venture funding, but great team and I love them. So I think that would be one. I think the incumbents, especially someone like Charge B, has a real, if they disrupt themselves a little bit. I think they could go from a product direction over there and win quite a lot. They could defend and take more market share and just own more of the wallet of their customers. So I don't think that the legacy players are out if they choose to play. I think that's sort of the key here. But I don't really want to point to a winner yet. think a lot of these are really early. They're like maximum series A for when I see them having this kind of product vision. But they make up for it in volume. There's quite a lot of them. So meaning that I spoke to at least 10 the last year, which is a little, which is, yeah, it's distinct.

Enrique asks, when introducing a new sort of disruptive product to market in your category, how much time would you say is necessary to prove that your pricing and parenting is effective? What metrics would you look at to assess effectiveness of your pricing and parenting? So if you're making a ton of money, it's probably good, right? I think you don't need to sort of overcomplicate this. And I would say the pricing and parenting is a major component of product market fit, right? So if you have product market fit and you're just selling like hotcakes, just keep doing that, right? Don't overthink it. If you're not in product market fit, you want to work on the product obviously, but you also want to work on the fit part, which is basically where the pricing and pay comes in. I wouldn't try to overthink it early. Like I usually say that between zero and the first two or three million of ARR, do whatever pricing, just like test out five different models, like sell, whatever you can to whoever you can in whatever way you can, because the purpose here is to learn, right? So the idea is to actually test out a lot of different things. And then as soon as one of them hits to where it just feels easier for customers to understand, it immediately makes sense, it's fair. And a lot of these are actually quality tip, Where you say, okay, the sales process just works better. This way of narrating the value of the product just works better. And that's what customers really connect to. That's when you then double down and try to do more of that, right? And then at around 3 million, can sort of like hone in on one model and then try to take that to 10, 15, at which point you probably need to clean it up again. Alright, HC asks here in the chat, the commercial debt drive, presume, is particularly driven by PEO and or near exit companies. For others playing the infinite game, is this an opportunity to win more business? Yeah, so I would say that.

A lot of commercial debt actually is the reason why B2B SaaS companies stalled out around 15 million. So the example I just used with Enrique was, hey, up to 3 million, do whatever you want, which basically is like, I'm giving you permission to have as much commercial debt as you like, right? That's okay. We'll solve it later. And then you need to not get too much commercial debt on your way to 15 million. Because if you get to 15 million ARR and you just have like a massive hodgepodge of different contracts and models and so forth, really that takes a ton of drive out of your sales team, out of your CS team, and also a lot of commercial debt actually tends to create more technical debt. So that just means that the organization just starts to move slower and slower and slower until it stops, right? So. What you need to do at that point is to really clean it up. Like it usually takes to be honest, like one or two years, because you need to go through an entire contract cycle with all of the existing contracts, renew them in a very disciplined way until you're out and out of debt. And you can grow in the meantime, like that's fine. But it's really only when you have that sort of reset of the entire commercial debt that you can really start to grow. And so, so the, the, uh, I had a US based company that I worked with and they grew from around 2, 3, 085 in, I think, four years. So massive growth, and then they completely stalled out. And then it took them another five years to get to 100. And that was basically because at the end of those five years, suddenly all the three-year renewal contracts caught up with them. And suddenly they had like two massive renegotiations with procurement teams every week. And it just completely like sucked all the air out of the sales room. So that's really what you want to look for. It's like, How much time do we have to sort of respend to just maintain this revenue base that we have? Or can we actually spend our energy growing net new business? And that's really what commercial debt is a lot about.

Alright, Jose asks to which department are these pricing teams reporting to? So I would say that the way that I usually structure these is that you have a pricing function that is the team that does the job. That is where you hire the pricing people and they actually work on a day to day basis and they report not to a specific department but to a committee and the committee consists of at least a product person, senior, usually the CPO. and the head of the commercial part. If you're marketing driven, head of marketing, if you're sales driven, head of sales. If you're a little bit of both, have both people in there. And then, especially if you're below, let's say, 50 million, the CEO needs to be in there too, right? And the idea is that the key component is to create alignment between these functions, right? Because you actually want a, like a straight line of understanding all the way from product through sales. through marketing so that everybody knows what we're selling. Everybody knows the story we're telling to the market and so on. So I say that this committee is a refereed by the CEO. He, she needs to make sure that everybody plays well together and the decisions are committed to and everybody takes action and executes. And then the input to those decisions like running the analysis, running the experiments, like coming up all with the ideas and obsessing over pricing, that's what the people in pricing function do on a day-to-day basis. So that's usually how we're structured. And then HG asked again regarding pricing professionalism. Do you see a similar move on the buying procurement side?

I don't know if I can speak to that, to be honest. So I think...

I don't know if I can speak to that. I would say that I deal with a lot of procurement teams. Am I seeing that they're getting more savvy?

some of them are. But I would say that. The really good procurement teams are actually the ones that understand the long-term value creation and don't see it as a fixed pie game, right? And that I haven't seen yet. Like I think the really smart executives, the ones that are good at buying software, they understand that it isn't a fixed pie game and that actually there could be a win-win scenario when negotiating with vendors. And I haven't seen that on the procurement side yet. And that's the way procurement is structured and so forth and a lot of software is sold. yeah, so I think there's still opportunity to make procurement more professional as well. So Greg asks, regarding pricing professionalism, where does top talent come from? Within the org, sales analytics, finance are coming in-house from consultancies. They come from a lot of places and actually, so all of the places that you see are great. I'd say product is actually quite a lot of times where people sort of do their first pricing and then if they really like it and they have sort of a quantitative bend, they can move in that direction. Other than that is mostly like MBAs, like fresh out of school, they join sort of pricing divisions in more established industries like fast moving consumer goods. There's a lot of industries that have long pricing traditions, and then they tend to transfer over into software. And that's what you usually see. These people, however, only rarely are the change management type people. It can happen, but they mostly are more analytical. So I would almost say that you have this issue where the change management type, like the senior pricing people and the analysts, actually don't have the same career paths.

which means that it's also quite hard sometimes for the analysts to upgrade or level up into the change management type roles, where the change management type roles often come sideways from other executive functions, really like pricing and then do that, which is actually, that's also how I hire and willingness to pay. I usually hire like former CEOs that I've worked with or like high level executives and then recruit them into pricing once they figure out that this is actually where value is created. And then So, but it's very rare that I just hire quant people and then upgrade them to the change management type roles. So I think, yeah, I see that in pricing in-house as well. Jose asks, what do you mean by narrow problems? Can you share a couple? I guess you mean price changes on a regular basis. So where do you talk about narrow problems?

Yeah, I would say like write out some more context to me and I'll try to answer because I think I I sort of lost the thread on where I mentioned narrow problems. So if you me a solid and repeat the question, I'll give it a go. Enrique asks for AI intensive products, what are some observations on companies that are pricing these products well versus not? So. I would actually say that a lot of the legacy players like Google with Gemini and so forth are doing pretty well. They're just like, hey, we are going to make a long bet and just include it at scale at a relatively fixed cost and roll it out. Copilot I think is a good one. So for a lot of the coding as well. So I think a lot of people actually are sort of moving in the right direction. And if I have to mention someone as well, think this was maybe like not a 2024 for the 2025 one, where Sam Albin did the whatever it was like the 200 pro subscription and then like launched it and basically whatever four, six weeks later say, oh, it's way too cheap because we're losing money on this. I think actually speaks a little bit to the right approach. Like he might have gotten the price point wrong, but the approach. made him learn that much faster. So he basically pushed something out at a flat price, lost money on it, but learned a ton. They could then, within weeks, turn around to better decisions. So I think a lot of people win who are willing to... So what he's doing is he's playing Yvon's paradox aggressively. He's putting out his own P &L at risk, losing money, but offering sort of...

An invitation for adoption on the user side that will enable him to learn faster than a lot of other players that are more defensive and don't want to. They don't want to learn with the risk of losing money along the way, right? And I think. The problem here is that. The winner in this is going to be one of the ones that play this aggressively. But also, of course, you have the risk of going bankrupt if you play it aggressively, right? So you have to sort of do it aggressive smart, which is sort of. what I'm trying to get my clients to do, which is pick subsetments of customers and test on those, where Altman can just test on the entire user base because he has unlimited capital. So I think you need to play the hand that you've been dealt, but I think everybody can play aggressive in pockets, like in smaller domains, in order to learn. And I think that's what most people are missing out on, is that they actually have that ability within their existing customer base.

I want to speak a little bit more on that point, actually. So, because I think another thing that I didn't sort of speak too much on in sort of the first part of the webinar here was what kind of pricing models we see with AI. sort of my analysis is that you basically have, I see this as a stack of two layers. So you have what I call a few layer, those are all the tokens. So there's a cost to producing like to consuming those. And then you have what I call a solution layer where you then take the tokens and create some kind of valuable output of it. And I think you can basically sort of divide a lot of the AI pricing models into whether they are like fuel pricing or whether they are solution pricing. So the fuel pricing wants basically wants to say, well, in order to like process the chat, We need a token per word and then we just price per token so that the more words you do, the more tokens you consume, the more you pay. So they price on a fuel basis, right? Which is great because it just covers all the cost. So you never lose money and customers understand it and so forth. So that's fine. But on the other hand, you have people that say, well, what is actually the purpose of our solution? Like what is the value that we're creating? So you can say the tokens or an input in a value creation process or value chain that leads to some output, which is what I'm trying to get done. So, well, why are we doing these chats? Well, we're doing these chats to handle customer support tickets. Okay, so how about we price them per support ticket instead and not per input of token word? And I think that's simple transition like the intercom is doing it and others are basically trying to...

figure out what kind of value chain they're trying to create for their customers and then say, okay, tokens is an input, what's at the other end? That's usually like a thing that is specific to them, like whatever their solution does. And let's say, hey, if we can make this tangible to the customer, we can count it and it's something they want, why don't we price by that unit? And then of course, what you get is that you get a loser connection. between the inputs, the number of tokens and the outputs, let's say the number of resolved tickets. But if it is strong enough, then whatever the right, like you're going to have a chat, which is like a hundred thousand words, and then you're to have a lot of chats with like one word, right? So, but that will sort of wash out. That's mostly noise, but you are going to have individual output units that you lose money on if you charge per output. But it is the ability to sort of do that and have that risk taking mentality. that really gets you to the value pricing of the solution pricing of what we're doing. And I think a lot of companies are trying to move in that direction. A lot of companies are doing it well. I think the problem is that the new use cases are popping up so fast that it is much easier, more generic to just have a token input model because it just fits into everything. So you just say, hey, we're pricing for token and then you can use our AI for these like 40 different things in our startup product. have added, right? So that is a perfect learning opportunity. But as soon as you lock in a really valuable value chain that you're supporting, you should basically move pricing from inputs, tokens to outputs of whatever it is that you create there. And I think if you have that framework for when you're analyzing AI pricing, you're going to try to see it a lot differently on how people operate. All right.

Aeon asks, to what extent do you see a shift toward outcome based pricing? How would you recommend that AI companies can measure and get aligned around the outcomes? I think Aeon, that was basically what I just answered. Like, think of it as, like literally think of it as a value chain. Like I do X, I input tokens into it and then something happens, something happens and then an outcome happens that I like, right? And across this value chain, you can basically measure different things like, we input whatever number of tokens and then we have number of chats and then we have number of total tickets, number of resolved tickets and then we have number of resolved tickets without human intervention, whatever it was, right? And then this price based on that. So the more, the closer you can come to the outcome that someone wants, the better. Is that an AI thing? Not at all. Like that was there before AI is going to be there, whatever, when AI is gone. I think, I think that is, that is more generally what good value-based pricing is and has always been about. I think it's just being more clear because a lot of people are trying to disrupt business models as well because of AI than they were, let's say, three years ago. So I don't think it's anything sort of net new that is due to AI. It's mostly just that people are starting to take pricing more seriously, partly because AI is now making a lot more things. possible than it were just a few years ago. But I think outcome-based pricing, especially for any kind of vertical software, is the way to go. And if you can own the horizontal infrastructure and domain, don't need to do outcome-based pricing. So Asia doesn't have outcome-based pricing. Czech DBT doesn't have outcome-based pricing. So they do very well because they're horizontal players, and they're basically providing the infrastructure to an entire market. And that's fine, but it is quite competitive. And you are leaving a lot of value on the table, which then the vertical players mop up with your horizontal solution. Enrique says, thanks. You're welcome. You like the idea of picking subsets of customers to learn from with them perfectly. So yeah, that's always been the case for product development, for pricing models, and so forth. Just pick a few. There's no rule, generally, that

you need to have the same pricing for everyone. That's basically what commercial debt is. So you might as well use it to your advantage, right? And then pick out a few customers and run specific experiments on them to learn. So that is a much more valuable way to deploy commercial debt rather than to just like have random discounting and sell some customers. So that's it. Katie says, great session. Appreciate you Katie. And Robin says, can you share an anecdote of a client who had terrible pricing? Yeah, I don't think I can share the name, But I can share some things here. So...

I'll give it to you. I met a guy, he was sort of the front row of a Kino and I did a few years back. And he was a bootstrap founder. He did about 10 million AR. So he was pretty happy. But the way that he did 10 million AR was that he basically had, let's say, let's just for the sake of marketing, he had a thousand customers that were paying him 10,000 each. So he had a flat fee. And he was doing some sort of like a signup software for events, like something like that. He was selling it to like different types of customers, like very different, like some were super large enterprisey, like large government contracts and all those were like smaller customers. But he was, was trusting everybody the same price. It wasn't 10K, but just for the sake of margin, the same price. And it's just. It's just a mathematical impossibility that that is the optimal price or way to charge it because you're going to say, well, look at in this case, it was signups. So it's just like, okay, so if you take and you rank all your customers, you can't rank them by price because everybody's paying the same. But if you rank them by signups, you're probably going to have, say, let's say that the, say they have 10 million signups, right? And then you're going to, and they all sort of pay like 10 K, but you're going to have someone that has. 500,000 signups, few super heavy user customers, and then you're going to have others that have 500 signups. So you're just going to see this discrepancy between how many signups they have and how much they pay. And since we know, because they are, that they all are willing to pay 10K, we can probably infer that the ones that are using it for half a million signups a year are willing to pay more than that. So we can basically take sub-segments of his customers You just say, hey, now instead of 10K, you're paying 50K. And if we just do that with, let's say, 25 % of his accounts, like the heaviest users, suddenly we're not doing 10 million of ARR. We're doing 22 and 1 million. And because we are targeting these specific customers, we can do it with an extremely low churn risk. So it's just one of these things where you see sometimes people have so much success with a model that

The opportunity cost of not looking at it in this way is basically costing them like 12 million a year, right? But it's sort of a mental block. Like they would much rather just do 10k a year for everybody because it's simple and it feels fair and like it's almost like a democratic approach to software. But it's just leaving a ton of money on the table. So we have that. Then the exact opposite of that is I had a client a few like this is five years back maybe and they had maybe let's say 30 million of AR and they had maybe let's say around 300 enterprise contracts on these sort of 30 million of AR. like relatively like mid-sized high ACV and we had done some work and we had concluded that we could raise prices around 80-85 % with very little trend risk. So they decided to not do that. And the reason they didn't do that was because all of these 300 different accounts were sold on 300 different contracts. Like literally their salespeople over the course of the 10 years that they have sold them had just written up 300 different contracts. A lot of the contracts that they didn't have anymore, like because they had been on a local drive or the sales rep had just left with his laptop. So we didn't have the contract. And so it was just like a prime example of commercial debt. Right? So just like everything was different, different pricing models, different everything. So this company, instead of repricing and sort of doubling revenue to, let's say, 50 plus million on these 300 accounts, which would have taken two to three years, just decided to sell instead. So they sold to a large trade buyer who then did all the work of the cleanup and unlocked a lot of value. But it was one of these examples where you have a success with a startup, but you don't get sort of the full value out of the product and the value you created for the market. as a founder or during an exit because your pricing model was just so much all over the place. So on one extreme, you have the flat fee founder that just religiously believes in this super simple model for everyone. And on the other hand, you have this totally gunslinger cowboy style of pricing that also will get you into trouble. And I think these two extremes basically are bad in each their own way. They're both not

good pricing because that basically requires you to have a little bit like a structured approach, but where you still get to price different customers differently, but without incurring all this commercial debt. All right. Let's wrap that up. Casey, you're going to get the last word here. You say, thanks, Oleg, as always, very insightful. I appreciate it. Robin, appreciate you, all of you. The recording will be over on the school community. So school with a K, skol.com slash about slash SAS pricing. That's where you get a lot of love from Rob and John from pricing SAS. And yeah, so see you over there. And I think actually the idea is that every time that Rob and John publish a new report, I'm going to do a webinar here. So this is going to hopefully be like a monthly thing. So I look forward to that. I hope you do too, right? So enjoy. Thank you, everyone, for tuning in.