The Pricing Mistakes That Kill Even Great Startups with Madhavan Ramanujam

20% of what people build drives 80% of willingness to pay. Price is a signal of value. I mean, a $200 wine, most people in a blind test would say it's actually great, even though it was a $2 wine. And actually worked on the product for 2 years without no idea of whether someone would actually pay for this. I ended up working with over 300 companies, more than 30+ unicorns. Price before product, period.

Most founders think pricing is an afterthought. That mistake has killed billion-dollar companies. Today we're sitting down with the man who's advised 30+ unicorns, helped build entire pricing playbooks inside Silicon Valley, and has seen more startups fail from pricing than product. This is the pricing conversation founders should have had on day one. Well everyone, very excited to have Madhavan here for the Library of Minds podcast. He is a pricing guru and an expert and investor in Delphi, author in two books. One of them here, he just released Scaling Innovation. And I'm super excited to have you here to talk about pricing.

Super excited to be here, Dara.

So how does one get into the expertise of software and tech pricing?

Maybe I can talk about my journey, how I got into it. Yeah. So when I was graduating from Stanford, this was probably 17 years ago, and in classic fashion, I wanted to found a startup and We had a team and I was this guy who was in charge of the marketing plan and everything else. And we went and pitched to a VC and I remember getting the question, how do you know you'll actually make money on this innovation? And I pulled up my spreadsheet with all kinds of assumptions and I showed it to him and he said, you labeled them right, those are assumptions. How do you truly know? I was like, I don't know. So I got into pricing by just guessing and hoping for the best. You know, build some products, slap on a price and hope to monetize. That's probably how most most founders approach pricing. That same week I got a call from a company called Simon-Kutcher, which is regarded as the largest management consulting company focused on monetization. And I was like, that's interesting. That's exactly what I wanted to learn. So I thought I will go there and learn the arts and tricks and go and start a company. One thing led to another. It was such a great ride. I stayed there for over 15 years. I ended up working with over 300 companies, more than 30+ unicorns. Josh, who's my co-GP now at 49 Palms, and myself, we started the tech practice and scaled it. It was a great ride. My journey into pricing was by accident, but what I really put out in these books is what is the science behind monetization? Because most founders approach it as an art, and it has to be a good combination of art and science. My whole motivation for writing the book was to give back a bit of what I knew so that founders don't have to like start the journey that I did, which is discover it by chance.

Yeah, and I started reading a bit of your first book as well where you talk about how some products fail not because of engineering but because of pricing. And I'm curious if you have an example of maybe two companies that had near identical products, but it was really the pricing that completely change the trajectory of one over the other?

Yeah, I can think of several examples because when you talk about pricing, it has different layers. Is it a pricing model? Is it a price level? You know, what does it actually mean? So maybe take a couple of examples. The classic example is, you know, Blockbuster versus Netflix, same product. It was a DVD that you can put on into your, you know, DVD players back in the day and actually watch a movie. Blockbuster used to charge based on, different titles, late fees, you need to get to the store, and it was a per movie basis. Very painful for those of you who have done it in the past. Then Netflix came up with a new pricing model that was revolutionary saying, just subscribe and it's a flat fee per month and you can get to keep the DVDs however long you want, no late fees, etc., which is a different pricing model. Rest is history. Blockbuster did not adopt to that model and Netflix really crushed it with the how to charge question. Which is often way more important than how much to charge. Another example that comes to mind is Xing versus LinkedIn. Both of them actually started at the same time, identical products in terms of being able to like network and, you know, form a professional network or a personal network kind of products. But Xing started with, rushed in with actually subscription plans and put a lot of features behind the paywall. LinkedIn actually purposefully kept the product to be free and thoughtful. The story of LinkedIn is actually well captured in Monetizing Innovation, where I wrote a chapter on LinkedIn's journey. But even from the very early days when we had worked with them, they were very thoughtful about how to land and expand, what is the overall strategy, what is that critical mass of getting consumer adoption before they actually start switching monetization. Their entire business model was different because they started monetizing on the talent side, then they started monetizing on the sales side, then they started monetizing on consumer side with like premium profiles, et cetera. But all of that was thought through even in the pre-IPO days when we worked with them, which was a kind of identical product. But LinkedIn crushed it with their pricing and pricing models and Zing didn't.

Wanna go deeper? Talk to Madhavan's digital mind on delphi.ai, link in the description. That's a good segue to another question I have. LinkedIn, like you said, I met the CTO of LinkedIn. He talks about how they were optimizing for profiles and revenue really wasn't a target early on. And so when does it make sense to delay pricing and monetization based on you're hoping to get network effects, you're hoping to get data that you can monetize later versus trying to generate revenue right now? Because LinkedIn could have done that. They could have made revenue a target in the early days, but they didn't.

For sure. And even back in the day, We actually tested that there was a willingness to pay on the consumer side, but consciously that was a delayed decision. We knew it was there, so we knew they were building the right feature set. For instance, even things like InMail, that was a classic debate. Should it be free and give it to everyone to connect with everyone on LinkedIn, or should it be a premium feature? There were also considerations like an InMail is actually a premium thing because you don't want everyone in the network to call everyone. It is a premium thing to actually connect. But there are certain things that need to be free. So it was very thought through but already tested. So in the book Scaling Innovation, I talk about how to architect towards profitable growth. And that is the key. Not profits, not growth, but profitable growth. What does that mean? That actually means being able to dominate both market share and wallet share. And the mistake that founders often make is, you know, they focus on either one at the sole exclusion of the other. And then get it all wrong. So if you focus on market share from the get-go and say that you're disrupting the market, but you're not thoughtful about how you would monetize or get the wallet share, most likely you have given the farm away in your entry-level products, trained your customers to expect more for less, and then you would struggle to actually monetize, right? If you focus too much on the wallet share, you're probably missing out markets that you can actually go capture. So the key here is having equal attention to both market share and wallet share at all points in time in the company journey. But that does not mean equal effort because at different points in your journey, you might actually want to push one lever versus the other. So naturally, if you have things like, you know, network effects or you're building, you know, core data that can be monetized, you need to delay monetization, but be thoughtful about how you do it. And even test that early, that that's actually— that delayed decision will pay off at the end. Not hope for it. If you just build on hope, that's not a strategy.

Sure. And how do you think about testing monetization strategies and new categories where maybe not everyone really knows what the pricing model will be, both consumers and businesses?

Sure. So that's kind of why we wrote Monetizing Innovation, because, you know, I used to work with a lot of products which were, you know, new category products. I used to get a call saying, "Hey, we need help with monetization and pricing." I'm like, "How long do I have?" It's like, "I needed it yesterday." I'm like, "How long did you take to build the product?" It was like 2 years. I'm like, "So you worked on the product for 2 years without no idea of whether someone would actually pay for this? That seems so wasted." The idea here is that in Monetizing Innovation, what we talk about is to do a product market price test. Which is important because many founders, when you talk about product-market fit, they would say, yeah, we tested the product for making sure there's a fit in the market and we are developing the right product. I mean, someone comes and asks me, do you like this sparkling water? I like it. Do you like it at $10? The whole conversation is different. So if you don't put pricing as part of your product-market test, you're often hearing what you want to hear. So even in new categories, what we talk about is, let's say, 6 months before you launch the product, You know, put wireframes, blueprints, you know, just the same kind of product or even a prototype in front of people. Let them struggle with it and have that same sales and marketing conversation that you would have, that you'd have after launching the product, but try to do it before. And ask the simple questions around, would you pay for it? And if someone says no, the most important question to ask is why? And then you start hearing all of these jobs to be done, unmet needs that you can actually productize. And that's the key, right? So the idea is to understand that broad brush strokes, is there a range of willingness to pay? You can get precise and optimize as you actually build out the product. Often what happens is if you just build the product without even knowing whether there's a willingness to pay, you get into all kinds of issues that you just built a product, slapped on a price, and you hope to monetize. Sure. So testing for pricing is highly possible in the early stages. And maybe the other way to think about this is as an entrepreneur or founder, You don't have a choice whether you'll have a pricing conversation with your customer. You're going to have it one way or the other. They're going to build a product, slap on a price, and have that conversation. Or do what we kind of practice, which is have that conversation early and test and learn. And if someone actually says no in that same— after you pitch the value, after you had the same sales and marketing conversation, someone says no, chances are 5 months after you build the product, they're going to say no again. At the end of the day, you're selling it to the same buyers. You're testing and learning. And there are various ways to understand What is the willingness to pay that actually makes sense based on, let's say, economic value that's unlocked for the buyers? And how do you actually come up with the right price anchors even in new categories? So if there's one chapter in Monetizing Innovation that I'll encourage readers to read, it's chapter 4. It's called How to Have the Willingness to Pay Conversation. And we really demystified that whole thing so that founders can go and put it in action. I mean, Rahul Vohra, for instance, read the book and set a pricing for Superhuman, which was an established category, but the alternative was free Gmail. How do you price Superhuman? So he followed the principles in the book and even talked about it in an A16C podcast. That's how I got to know him.

Is there a strategy in free and then making people pay versus paid and then trying to catch up with everyone else who maybe had free?

Yeah, I think the, like, There's always been this lure of free for distribution, right? I mean, that's the quintessential founder's dream. I'll give it up for free and everyone will want my product, right? I mean, so I think everyone thinks about that. The key is even when you think about free, there are various versions. Does it need to be a free trial? Is it a freemium? Is it an always free product? So in the book Scaling Innovation, we talk about what considerations will lead you to which pathways and should you even do free in the first place? And not every category needs to, especially if there are, you know, costs behind the scenes. If your product is such that you can offer a free variety, then you need to be thoughtful about landing and expanding. Most companies only land, they're not expanding. So you need to have a pathway. Here's the thing that I've learned, which is a counterintuitive truth across all the companies that I work with. You know, 20% of what people build drives 80% of willingness to pay. And that 20%, ironically, is the easiest thing to build. So what founders typically do is build that 20%, put it on the market for free, and then train the customers to expect that that's the free product. And then they're chasing their tails to build 80% stuff that's only driving 20% willingness to pay. So you kind of already gave the farm away. But if you do these kind of price experiments and understand what drives willingness to pay, you can be thoughtful about the free experience in such a way that you might gate it based on either features or maybe even in terms of certain usage amounts and then have a free-to-paid conversion that is actually thought through in a meaningful fashion. If you can afford that in an AI business, that's a great business to have. But that's by itself not sufficient because you also need to build switching costs in the business to make sure that, okay, if someone else comes and offers a free product, are people just going to go to that? Or what actually makes them sticky. Just giving something for free also signals different things in the sense that was it really valuable?

Sure. It's a psychological thing.

Oh, totally. I mean, sometimes the price is a signal of value. I mean, a $200 wine, most people in a blind test would say it's actually great even though it was a $2 wine.

Yeah. I remember a couple times there have been customers who have said, "You're massively underpricing. Like, I would pay $100,000 for this." And every year AI gets cheaper and cheaper. And so we could say, yeah, we'll charge $100,000 for this. But at a certain point, the real value of these things are going to become clear as more and more people can build them. When you think about pricing in the age of AI, should you be pricing from a, let's just change our pricing model every year to match the market or do what we did, which is start with the end in mind and price for where the world is going to be 5 years from now?

Definitely the latter, which is keep the end in mind and, you know, price for what that is. But I would also encourage people to adopt a value-based pricing mindset with AI, because what has really changed with AI is two things: increased autonomy and increased attribution, which means you have more pricing power. So avoiding the cost-plus mindset If you're in a cost-plus mindset, naturally if the costs are going down, you're going to keep changing price to reflect the cost. But if you're in a more value-based mindset, then you're basically, when the costs go down, that's when you'll actually get margin and you can actually then say, okay, it's not just revenue, but it's durable revenue. I mean, you can reach $100 million in ARR, but if you're negative 15% margin, you need a pathway to actually get profitable too, right? So you can't expect that the price will keep going down. So I give an example like how you price, for instance, FinAI prices based on number of tickets resolved based on an AI agent, right? If an AI agent resolves that ticket, that's when they charge. If a human intervention is required, they don't charge. That's a value-based metric. If they just base it on like how much it costs them to resolve the ticket, that's a slippery slope. Yeah. Right? So being in that value-based mindset is key.

Now let's say you have product market fit and it works in a subset. Should you spend more time dialing in that pricing model or when does it make sense to start investing in ads or B2B if you're starting as a consumer company or any other company?

Yeah, I think, I mean, I often say that how you charge is way more important than how much you charge. Pricing model is probably one of the most important decisions that companies need to grapple with from the early days. And that keeps evolving as you go on, right? Especially for AI companies. I mean, should you be on seat-based? Should you have a hybrid? Should you be on usage? Should you be on outcomes? Those are all kind of like day one questions because you're, you know, starting to train your customers on how to actually engage. And we actually developed a pretty cool 2x2 framework in Scaling Innovation. We can talk about that if you want as to when to pick which model. But I— Okay, cool. So when you think about AI companies specifically, what has happened? Autonomy and attribution. So think about a 2x2 where autonomy on the y-axis from low to high and attribution on the x-axis from low to high, right? I love 2x2, so I'm going to persist with that. If you take the bottom left quadrant, it is where autonomy is low and attribution is low. When I mean autonomy, that means, you know, you're running it in a copilot model. You still need humans in the loop. The AI cannot be fully autonomous. When I say low attribution, it means that your products kind of help with the copilot motion, but they don't necessarily translate into key business outcomes that your customers are tracking. If you're in that quadrant, most likely you have to be in a seat-based model because you're a copilot and you can't attribute too much value. So for instance, you know, take Slack, which was a great company in the previous vintage. You can say my productivity went up when I used Slack. You cannot measure it. You cannot monitor it. You cannot charge based on it. The attribution is weak. But it's a copilot and it's an interesting product, so it's a seed-based model. If you go to the bottom right where attribution is higher but autonomy is still lower, you get into hybrid models. These are companies like Clay or Cursor. They actually have a seat-based model, but they also layer in AI credits. So each plan comes with a certain amount of AI credits that you can use. If you use it, then you need to buy packs of AI credits. These are hybrid models because attribution is more clear because you use Cursor or Clay, then you know that, okay, there's, I could write my code much better with Cursor, for instance, and there's clearly attributable and you save time for, you know, people in writing code, et cetera. But it's still in a, you know, Copilot mode. So they're in the hybrid model. If you take the top left quadrant where autonomy is high but attribution is low, most likely you're a backend infrastructure play which is running on its own, doesn't need humans in the loop, but the attribution is still not fully aligned with business outcomes. Twilio is a great example. It's on a usage-based model in that quadrant. The golden quadrant to be in, and this is not for everyone, is if you're fully autonomous and also fully attributable, right? If you are in that quadrant, then you can be on an outcome-based pricing model because the AI does things on its own and it's clearly valuable. So then you can be on an outcome-based. So FinAI is a great example. They actually are in that quadrant where, you know, the agent runs itself, resolves the tickets. Their customers would say the ticket is resolved, the humans didn't even come in the loop. So they charge based on a, ticket resolved by AI agent. If humans are required, they don't charge for it. Or companies like, for instance, Chargeflow, which actually recover chargebacks, that's pure incremental revenue. Then you can say, I'm taking a percentage of that chargeback, for instance, as your monetization model, right? The key is to like pick the right archetype for your products based on where you are in the journey on autonomy and attribution. And then build some pathways to actually get to more attribution and autonomy over a period of time if you can, get to that, right? And that's the key. Like even if you take coding platforms, they start— GitHub started in the bottom left quadrant on a per user. Cursor and others have actually even GitHub Copilot now moved to like a hybrid model. But then they're already talking about, you know, fully autonomous QA agents that you can hire. And it's an outcome-based kind of model that's going to the top right. So I think picking the right pricing model from the get-go is key and then having a pathway to how it'll evolve. If you're in a B2C situation, that could take different forms. For instance, what do I get started with right now? Which side of the marketplace should I monetize? Is there a monetization on the other side of the marketplace? Can I layer in additional revenue streams through ads and others? Those all become decisions as you go along, but none of those will actually make sense till you build a critical mass of consumers, etc. That's also Why a pricing model conversation is from day one, but it evolves as you build your company.

Makes sense. And what is the biggest mistake you've seen a company that you've worked with make? And tell us the horror story.

Probably delaying monetization for too long and then training the customers to expect more for less. I mean, the classic example that comes to mind is Evernote, which probably did We almost had about 200-250 million users on the free product, very compelling free product. There was no necessity for someone to even take the paid product because the free product was really very compelling. They delayed the monetization for a long time. Of course, they got the volume of users, but they didn't monetize on that in time. Then the question is, companies like, for instance, Notion, probably started that in a different way where they were thoughtful about what is the free trial experience, what are the paid products, what are subscription products, and also having products for businesses, et cetera, and being thoughtful about that. So I think Evernote realized that and tried to pivot, but it was a bit too late.

A couple more questions I want to open up to the team. When you think about raising the ceiling and lowering the floor, getting more people into your product versus generating more revenue of the people that are already on the product. How do you— what's your decision tree for when it makes sense to do one or the other?

Yeah, you need to actually be thoughtful about both of them at the same time. I don't think it is a priority on one or the other, right? If you just prioritize, let's say, getting new customers, then you're actually going to see that you did not monetize on your existing customers. Indirectly, what that means is you did not train them on what value you're actually bringing to the table. So they are likely to churn. If you focus too much on your existing customers, then you probably were not, you know, focusing on acquiring more customers in the first place. So it's being thoughtful about both and having equal attention to both of those. The efforts might be different because initially it might be more about acquisition because you don't have too many customers to also have ongoing conversations. But at some point, the effort on existing customers might be needed more than acquisition. But paying equal attention is key. Yeah, absolutely. I mean, and I've seen too many companies get it wrong where, you know, they either focus on one or the other at the exclusion of the others or focus on neither. That's actually the worst where you're neither focused on market share or wallet share, but such a, just a core set of customers and you're kind of a community builder. You missed all the markets that you could have built for because you're obsessed with this customer base.

It was in your archetype in the book.

Yeah, exactly. And then you also trained your existing customers to expect more for less. So you don't you don't even build a community in terms of wallet share. So I think being thoughtful about both from the get-go, that's my answer.

What's one question you wish founders asked you more?

I would definitely want founders to ask me, when should I think about pricing? And my answer is actually encapsulated very well by the folks at First Round Capital. They summarize the book Monetizing Innovation in 4 words: price before product, period. Which means you're actually testing for your product, whether people need it, do they value it, is there a willingness to pay as you build it so that you can build towards successful launches as opposed to slapping on a price and hoping to monetize. So I would wish everyone would ask me when, and the easy answer is should have already been thinking about it. Right now, yesterday. Yesterday, exactly.

If you could have dinner with one person dead or alive, who would it be?

I would say, I mean, this is probably such a cliché answer, but I have to say it. I would say Steve Jobs. In Chapter 2 in Monetizing Innovation, my original title was You're Not Steve Jobs. And the editor thought it was too provocative. So we toned it down to like, Why Good People Get It Wrong.

That would have been so good.

That's exactly why you should do it. I know. I mean, but the rationale behind that is everyone Everyone talks about how Steve was a product genius, but what is probably often missed in literature is how he was also a pricing genius. When you talk about Apple, they have actually mastered things like coming up with a higher-priced product, skimming the market, building for different segments, having behavioral psychology in that. Like, I remember going to like an iPhone store and iPhone X was available for $1,000. I mean, I was like, what's a $1,000 phone going to look like? And I go there and I know all the pricing tricks. I go there, I'm like, no, I'm going to buy it. $799 phone without the Retina display, that looks pretty good enough. So I'm like, I fall for that because that's my segment. So if you look at what Apple did, they didn't price their products, they productized to different price points. And that's actually quite deep because there's a product for every budget or wallet that you can actually buy. So they were actually very thoughtful about their monetization. And even when they released the Apple Watch, everyone was talking about the $17,000 gold watch, which no one bought. But the $700 watch didn't look too expensive compared to the $17,000 watch that was actually talked about, right? So I think when you think about it, there actually been— and Steve is especially, I mean, my personal view is that he was also a monetization and pricing genius. If I could have, you know, go back in time and have dinner maybe with at least have a conversation with the Delphi AI for Steve, that would be fascinating.

Mateen.

I have two questions.

First question, pricing models, they constantly change, but something that doesn't change is human nature.

Are there any stories or resources that have taught you a lot about how humans just think about pricing in general? That's a good one. So there are, we call it like behavioral pricing, just like behavioral economics. If you take like B2B situations, at least till date, it's humans having a human conversation. You can set all the pricing you want, but you're sitting across the desk and negotiating pricing. Things like behavioral and how people think is way more important than the price sheets or whatever you actually use to come to that discussion. How you frame that actually matters. Maybe I'll tell you a story which was super impactful for a recent founder that we actually work with, AI company. This founder was actually— is a kick-ass product, great autonomy, a great attribution, and was able to demonstrate millions of savings for their customers, purely incremental because his AI was there, like in the tens of millions. But the going-in assumption was, I want to charge this on a $50K per year contract. And our question was like, why? You're unlocking many millions. Why $50K? Why not more? And he was like, I don't know if I have the courage to actually put a larger number. What happens if I lose the deal? I want to make sure I have logos. So we actually told him, okay, look, present two options on the table, right? We said, say that the first option is $50K fixed deal plus 10% of the outcome, or a $500K fixed deal. And that actually gave him the courage to put a $500K price point at the table because he had a $50K price point at the table. The entire conversation revolved around What is that 10%? How do you measure it? How do you monitor it? Where do you actually add value? That's a fantastic conversation to actually have because that's value-based selling. You're all focused on what is that 10%? How do you add value? No one talked about the 500, and that 500 got negotiated down to 400 because the buyer at the end of the day said, you know what, that's great. I understood how you add value. They did the computation. They're like, if you do this 10%, we are going to pay them way more than the $500K. So let's just go with the $500K option. I like that. And then negotiated him to $400K. So instead of the $50K getting negotiated to $40K, which was a default, he 10x'd his price. Now this is completely behavioral in some ways, behavioral for giving the founder courage to actually put a bigger price on the table, behavioral for the buyer to actually say, you know, how do I actually look at all of these and actually choose? Often I say that giving choice in these conversations is way more common than you think. People try to make pricing simple and then erroneously make it very simplistic. Even if you go to a pizza store, buy the pie or buy the slice are two different pricing models. You encounter that daily. There are behaviors that are ingrained in us where we can choose and self-select. In that same situation, another buyer actually took them up for 50K and 10%. That's a high-class problem because now you actually invented an entire pathway to make your product better and kick ass so that you can take part in that 10% of the value creation. These kind of human considerations are very important. Also, price signaling value. There's a Stanford experiment where a $2 wine branded as a $200 wine and a $200 wine branded as a $2 wine. When you do a blind test, everyone would say the $200 wine was better. It doesn't stop there. They actually did an MRI scan and figured out that the dopamine effects was way more higher when the $200 wine was actually at play, even though it was a $2 wine. Not only do we perceive value, we think it's valuable, we perceive it. Price is a signal, anchoring high, and all of these considerations are important. We devote entire chapters to this in both books. I think it's a great question. Were there instances where you, learned something and you were impressed by a founder that you remember? There's always a balance between what data tells you and where your conviction actually lies, right? As a practitioner of the field, I used to index more on this is the data. This is what you see is back in the day, right? I mean, when I started working with founders. But then conviction is a totally different thing. If you actually believe in something and the future, The future manifests itself to your beliefs. You start creatively visualizing the outcomes. That is something that I learned from this founder that was super inspiring, that even though I thought there was a really hard pathway to take, just by him repeating this to his teams and the market, it just happened. And it was probably one of the most fantastic products out there. You use it today. But the answer that we actually gave that founder was a tad bit different than what he actually did. What I realized is data, insights, and things like that are all useful. It goes a certain way, but then conviction is a whole different thing. Probably for the next 100 companies that I work with, I used to tell my teams that we are change agents. We need to enable change. How do we build conviction in our answers and manifest that to reality? I think that was probably my biggest lesson. Probably already take that example right away with the specifics of the situation, but I'll be giving too much credit to one founder when I work with a lot, so I don't want to be sounding too partial here.

Great. Well, thank you so much.

Awesome.

Appreciate your time.

Thank you so much.

Thanks for tuning in. For personalized advice from Madhavan, head to delphi.ai and ask his digital mind.