Clay: Everyone wants to price like Clay. Should they?

A decade ago every founder wanted to price like HubSpot. Now they want to price like Clay, the poster child for credit-based pricing in the AI era. In March 2026 Clay split its pricing into data credits, spent enriching contacts, and actions, spent running the workflows built on that data. With enrichment data getting cheaper everywhere, the money has moved to the work, and actions keep growing as an account gets more sophisticated even when its data stays the same.

The episode also takes the pricing page apart: plan gates that hold back the very features that drive consumption, two credit pools where one might do, and the agency tier Clay has not built yet. Then credit policy, from why every credit model ends up offering rollovers to why a partial rollover is barely better than none.

Pricing Page unPacked takes a real company's pricing page and breaks down the decisions behind it, the trade offs, and what it says about how the company actually wants to grow.

Hosted by Rob Litterst and Ulrik Lehrskov-Schmidt

For more information, catch us on https://www.willingnesstopay.com/

Transcript

Copy Transcript Copied

SPEAKER_00 0:00

One of the ways in which Clay won historically was by just being very generous. And so like they're kind of an exciting company from that standpoint, despite the fact that they have like a very boring business.

SPEAKER_01 0:09

So the CFO is just gonna look at that and say, why are we allowing Robux? Like that's the easy.

SPEAKER_00 0:17

Welcome back to Pricing Page Unpacked. I'm Rob Litters and I'm joined by Ulrich Perkoff Schmidt, the CEO of Willingness to Pay. Each week we take a real company's pricing page and break it down. The decisions behind it, the trade-offs, and what it tells us about how the company actually wants to grow. No slides, no scripts, just a real conversation between two friends who live and breathe pricing. Let's dive in. Ulrich, today we are talking about a really fun company that I think has become kind of like the favorite poster child for credit-based pricing uh in this AI era. We're gonna be talking about clay. Do you have any uh any thoughts that you want to get off your chest about clay before we dive into the background?

Why everybody wants to price like Clay

SPEAKER_01 0:56

No, well, two things. So, so when I started to advise on pricing like a decade ago, people would come to me and say, Oh, I want to price like HopSpot. And then, you know, that's that's sometimes a good idea, sometimes not. And then we have that conversation. And then about three, four years ago, it shifted, and then now people want to price like clay. And I was like, oh, I want to price like clay. And okay, so that's sometimes a good idea, not always. So, but we're gonna get into that. The other thing I'll say is like it's a bit of a caveat, is that I've actually spoken a lot to the to the team inside of Clay, uh, also ahead of their latest pricing release, uh, especially around some of the problems that we're trying to solve uh with this recent change. So I'll I'll try and sort of um be a gentleman and not sort of um you know sharing a secrets that I can't share. Um, but uh we'll go through.

SPEAKER_00 1:43

Yeah, I think um it I if you're talking about Zona, I've I've caught up with Zona too, and she's she's amazing. She's one of my favorite people in the in the pricing community. Um and we will definitely get to that recent change because it's very, very interesting. Some quick background on Clay. So they have kind of like a um a really interesting origin story that's kind of becoming legendary. So they they were founded in Brooklyn in 2017. Um, the CEO is Kareem Amin, who is a very, very cool character. Like, if you ever heard him on a podcast, he's like a really interesting dude. He's very cerebral, he's very different from the conventional tech CEO. I think like there's kind of like a prototype for like the Silicon Valley, like high growth CEO, and Kareem bucks a lot of those conventions. They were unprofitable and like really didn't make any money, I think, for like five or six years. Like it took them a long time to get to their first million. I think part of the reason is they kind of didn't really know exactly what they were building at first. And then they really stumbled into this kind of like beautiful um use case in go to market. And so Clay is an AI-powered go-to-market development platform. Um, you can think of it as kind of like a spreadsheet style workspace for finding, enriching, and acting on customer data. Um, they have a few core products. We won't go crazy here, but there's clay tables, which I just referenced. There's clayGent, which is an AI research agent. Uh, they have a tool called Sculptor, which is a natural language co-pilot that will build tables, workflows, and analysis for you. They have uh signals, which show you job changes, funding, like all sorts of different um intel triggers that would make it worthwhile to reach out to somebody. Um, they have the sequencer, which is a native email campaign engine with inbox warming and account rotation, and audiences. So they have a unified segmentation layer across CRM signals and enrichments, which syncs back to your CRM and ad platforms. Right now, the company has about 14,000 customers, or at least that was the case in January 26th, the the kind of last note that we've gotten. So I'm sure it's much larger than that. Um, they serve solo founders through G enterprise GTM teams, and their enterprise net revenue retention is over 200%. Um, a big thing with these guys is they literally spawned this whole idea of a GTM um engineer, um, which is something that just like never existed before Clay. Um, and like if you look around LinkedIn, you'll see like a lot of salespeople now just call themselves GTM, um, which I think is also kind of like a trickle-down effect of this entire thing. Um, a few more details and then we can get right into it. From a revenue standpoint, they are, I think in May 2020, 2026, Sacra estimated they're at 150 million in ARR. I might even take the over on that just based on kind of some some of the things that I've seen in uh on LinkedIn. Their revenue has grown three and a half X year over year. Cash flow positive for parts of 2025. They are still a private company, valued at 5 billion in January 2026. And they have about 800 employees, and that is where we will stop before we jump into pricing and and everything they're doing there.

SPEAKER_01 4:50

I think they're

Data is commoditising, so what is Clay selling

SPEAKER_01 4:51

they're they're arriving in a market at a certain time when outbound and a lot of these sort of go-to-market motions start to become a lot more like efficient and optimized and sort of technical, right? So so for a while in marketing, you had like, okay, we're gonna do like advertising. And then it was like, okay, for outbound sales, we're gonna have, we're gonna buy lists from companies and then we're gonna call everyone on the list. And then and then it turns into sort of okay, we're gonna not only sort of buy the lists, we're also gonna chase them with like email sequences, target ads, or we're gonna, we're gonna, we're gonna find out where their where their socials are, like on LinkedIn and so forth, and then we're gonna DM them and and so forth. And so this notion of how outbound would work. And as as one thing started working, like the market would floctuate and then it would stop working again because now that sort of way of getting customers was like saturated, and then it moved to okay, what else do we know about customers and how to get to them and so forth. So they're like clear sort of built on this idea of like getting leads, getting to the needs in a way, and sort of building the infrastructure of that. Like so, so before the before 2020 and just after in the in the six years that passed, I think that whole thing from a technical standpoint become like vastly more sophisticated. Uh like like data enrichment, this idea of of like the scale of outbound emails you send and the and the sophistication and the personalization that you do is sort of a a way that you try to sort of overcome like the kind of volume that everybody is is exposed to in marketing, right? So, so it's also a category overall that that is beginning to see some commoditization. So I think like Clay and others is like, hey, Apollo, like a lot of these other data management companies, like it's making a lot of money, it's six sense, it's making a lot of money on like, oh, we can like we can give you the secret data about your leads. As that gets around, somebody's just gonna start to undercut you. And I think we're gonna see that. So I think Clay has to decide are they selling the data or are they selling sort of the functionality and the workflow and the way of working around the data? Um, the data itself, I think, is going to be commoditized, like at least that's a thread. And the workflow is instead of like maybe that is being maybe attacked by by the LMs, like by Claude, that's just like, hey, we're just gonna create a workflow for you. So I think from Clay's perspective, like they're in that everything is still fine, but I think these are the threats on the horizon for play.

SPEAKER_00 7:19

I'm a former salesperson, so like I like thinking through kind of like pre-clay, how you would do enrichment and kind of like the key companies in that regard. It was typically like Apollo or ZoomInfo, right? Like you would use one of those two, you would enrich your CRM and Salesforce, and like that's how you would get your emails from people, and then you would reach out. Um, I think what Clay originally did is like they had this like waterfall method where they would have like data from all of those companies. So ZoomInfo, Apollo, like all these other kind of like data brokers, and you would give them a list of people that you wanted information on, and Clay would like with their waterfall basically run through the entire list. And if like Apollo didn't have somebody's email, they would check Zoom info and then they would check the next one, then they would check the next one. And so they had this kind of like aggregator effect of like being able to reference all of these different data providers to try to enrich whatever it is that you have. They realized that they wanted to build something on top of that. They realized that the data was kind of commoditizing. Um, and I think like if you follow Adam Robinson on LinkedIn, I think he has like a new startup called like Molt Set, where he's basically trying to like give away contact data for like 30 bucks a month or something like that, for like, you know, thousands of contacts. And so validates this idea that like data is commoditizing. They've kind of put this new metric on top of their pricing page, which we'll which we'll get into called actions. And it's all about kind of like the orchestration on top of the actual data. So it's like you have like this data layer, and then you have like all the campaigns that you're running on top of the data layer. They launched an ad product, so it's not just outbound, it's it's also like running ads and stuff like that. So super, super interesting. And you you brought up a good point. Like they they basically are like a place to like generate leads and enrich your leads, uh, which isn't like the most exciting business, but they have really good marketing and really good brand. And so like they're kind

Two pricing metrics, data credits and actions

SPEAKER_00 9:07

of an exciting company from that standpoint, despite the fact that they have like a very boring business. Back in the day, people would always talk about HubSpot's pricing. Clay has become kind of like the go-to pricing model that people get excited about in this age of AI. Clay has four plans. They have a free plan, a launch plan, growth plan, an enterprise plan. And for each of these plans, they have kind of like two pricing metrics. They have actions and then they have data credits. And so the data credits get consumed when you enrich contacts, like we were talking about earlier. The actions credits get consumed when you build orchestration layers on top of them, run campaigns on top of them. From my conversation with Zona, it sounded like what was happening is like they were building out this functionality for orchestration, but they didn't have a pricing metric to account for orchestration. So they started charging more for data credits, and then customers started looking at the prices of their data credits and then seeing what they could pay elsewhere for data and realizing that there was just this really big premium with clay. And so I think when they made this latest pricing change, they basically got rid of the premium on their data credits. I think they sell it at like a you know 20% margin or something like that. It's it's almost at cost. And they have this kind of like additional layer on top of it where I don't think all of their customers are really there yet with orchestration. So I think like they actually cut down their margins in the short term for this long-term play. But I think like their bet is that their customers are going to start building way more workflows on top of clay, and that actions metric is going to allow them to monetize better long term.

SPEAKER_01 10:50

I I think they might still have it. Um, but they actually used to have like a bring your own data. I think they had to do it because some enterprise customers already had existing data contracts in place. So it's like, okay, I'm a big customer, I want to use clay, but and like we'll get this data, but we have this big other data that we also want to use in Syclade. So basically basically like let them in, and then they got the full value of the whole automation orchestration layer, but without actually paying for the underlying data. So I think this is just a really good example of like a new pricing mechanism, like the the price for actions. It does two things. Like it's it's sort of like a real pain point, like you, in the sense like you have customers sort of getting away with murder, like they're getting value out of a product without paying for it. And you sort of, and so in in a weird way with there's like an edge case that just grows and grows, and now it becomes like an actual use case for customers, and you haven't found a way to monetize. You want to plug the hole, you want to put something in place so that you actually make money where there's value. And then the other thing is that oh, this actually like starts a whole new dimension of value creation across all customers, because it's not only the one that brings their own that have actions, it's actually everyone. I think there is there is some sort of like other repeatability in actions that there isn't necessarily in data. So if I have a database and you know, it's let's say 50,000 people, and then I want to enrich it, then I go in and enrich it, and then I have a lot of like activity right then, but then it is sort of enriched, and then I'm only looking to update it on a like whatever, every week or whatever my cadence is. And that means that suddenly my my consumption on the data layers is way less, unless I take another 50,000, like new lead-based. But if I'm working like relatively on the same leads, I just want to keep them updated, but the data is now kind of mine, so they can't sell it to them twice, like the actual set. They can update it and so forth. And then the actions is different. If I want to put up like an like an automation or something that actually writes all of my 50,000 leads, let's say every day, then that happens on a daily basis. So that actually accumulates in a different way. And if they didn't want to like, oh, I want to, I want to create a little bit of like a like a better sophistication, I want to segment them before I write them, and then I want to segment them, and then the ones that do X, I want to start up another loop, and now I re-retarget them with this other thing. So the actions actually, as the as the account grows more sophisticated, their marketing becomes better and more like nuanced in the way they work with the needs, actually grows and grows and grows. So you can have two customers, and this is also the I think the fundamental monetization issue that Clay was trying to solve is that you can have two customers that have like the same amount of data, but one customer is using all of the workflows and automation in a way more sophisticated way, hence getting more value out of that data. So by measuring the action layer, you're actually measuring sort of the throughput of the system in a different way. And and so and that's I think where the real value is. So I think it kind of started a little bit as a sense that this was going on. Then they had this sort of pain point of customers getting away with like getting bringing their own data. But then I think it turned into a much better monetization engine overall. I would even say that a an aggressive move for someone like Clay could be to just offer it for free. Offer the data for free. I don't know the economics of that decision, but yeah, if you could somehow gate it so that it doesn't get out of the system and say, hey, you can have it for free, but you have to use it in here. Yeah. Like you have to use our like tool set to to do your marketing and then just make all the all the money on the actions, I think that would actually that would that would that would create more value than it would uh I think you're totally right.

SPEAKER_00 14:32

I mean, I I think like if somebody was going to try to challenge clay, um, like I I I feel like clay probably has too much existing customer revenue to totally commoditize data. It would be a very bold move if they did that now. But like if they if you're a new competitor in this space, I feel like that's the play, right? Go free with data and just come after clay.

SPEAKER_01 14:56

Yeah, but you you kind of need you just went through the like the list of products they have. That product suite or portfolio is sort of the like the workbench where you can perform all the actions, right? Right. If you don't have that, you don't have a monetization mechanism. So so the thing, unless you want to catch up on the whatever nine years of product roadmap that clay has built, and and they're good at building this kind of product. Totally. You you you're you could give the data away for free, but customers might still prefer the actions since the data is already cheap. So I think this is more like a defensive, like screw you moment for clay if data starts to be commoditized. So whenever something starts to be commoditized, my question always to my clients is can we just make it free? Like instead of like we're gonna have we're gonna have a race to the bottom and we can't make money here anyways.

SPEAKER_00 15:46

It's funny because like Clay was the first company that I saw that kind of like had this type, this sort of a dynamic where you know they would like route you through all these different data providers. But now I think now I see it a lot with like the inference clouds. Like I think like base 10, um fireworks, together, these inference clouds that are basically like helping companies um pick the right model for whatever tasks their agent is doing to lower costs and you know increase performance. Like I feel like it's kind of like the same dynamic, right? It's like that they're helping you manage these underlying sources, which are actually like models, and kind of route and choose which ones to actually use. Um, so it's it's interesting in the age of AI how much like this kind of value has become, has become important. Like what we're seeing some really key players who are who are kind of like embracing this dynamic. Another thing that I wanted to touch on here that I think is really interesting.

The plans, the usage tiers and what they gate

SPEAKER_00 16:43

Um they actually have a free plan, which like a lot of companies I feel like are starting to get rid of their free plan or make it practically unusable. Clay actually has a pretty solid free plan. Like you can get in there and play around a little bit, especially if you're a startup. They have like a credit model. They traditionally have had a credit model, they do not charge per user. They allow unlimited users, and that's very intentional. They want to get as many people using clay as possible because the more people are using clay, the more credits they're gonna be consuming, the more actions they're gonna be consuming, the higher their price is gonna go, the more they're gonna get out of the product. The thing that they do that I think is really interesting is within their launch plan and their growth plan, they have like different tiers of usage. Um they have like different tiers of actions um at different price points, and then they have uh different levels of data at different price points. Um, and so like within each tier, there are kind of like these specified cliffs based on the number of actions and data. Had you seen this before? And what do you think about this kind of like level of price customization by plan?

SPEAKER_01 17:45

So they have four plans, free launch, growth, enterprise. And then in each of them, I get little two two small sort of drop-down menus. So the launch one, it starts at like 100k actions per year and 30k data creators per year. And then I get to sort of like click on it and then I get to sort of choose like a larger volume of these, right? So I can actually have more data in launch than like the the lowest amount I can have in growth, like the higher plan. So it's actually I can have more volume in a lower plan. I think this is like this kind of pricing design, it's not super bad, but it's also not like how I would normally design it. If you then sort of look at the sort of the feature set that you have, like in each of the plans. So each of the plans will then allow you sort of more features. So for example, uh enrich phone numbers, that's the first paid plan, launch plan. And then you can run a limited search and trap job changes and like some with their signals. And then you go into growth where you then have auto-sync and enrich CRM. Um, you have uh integrate with any API, automate things with WebO. This seems like they're saying, okay, so you have to buy a larger plan if you're like a more sophisticated marketer. Like if you if you're a more sophisticated marketer, you have to buy a higher like like that's that sort of tracks in a way. However, some of these things that are gets to growth are also the things that are going to actually either consume data credits or actions. We're making people pay for things that would allow them to consume like the usage that they then pay for. I'm thinking that the way that they have structured it is actually like they're almost like they're gating the things that would that would make them adopt the product and consume more. Yeah. So I think there is there is if I were looking at it, I would say, okay, are there things that we can that we can put in as features um in some of these plans that are not directly tied to like limiting the number of data creators or actions that customers can take. And then everything that allows me to consume actions or data, I would just make as available like very, very early in the in the in the product stack. So that's sort of one element here where I think they're they kind of want to have these different tiers, and then they want to sort of like they sort of run this sort of sophistication track. The next is I don't think that the difference between the launch and the growth tier is worth anything. So it's like the launch tier starts at 167 bucks a month, and the growth tier is like three times that is like 450. I would want to see for like a plane shift at least a 5x in pricing, if not a 10x, because the difference is like a few hundred bucks. So I think it's not enough to really sort of truly get into like another group of customers. Another point here is that it looks like I'm consuming data and actions separately. So I think I think there's gonna be an evolution here where I would sort of challenge them, and they're probably gonna challenge themselves if they haven't already, of like, why don't we make this like one fungible pool? I buy a hundred thousand and then I can consume it through either actions or credits. And so I can have a scenario where I run out of actions, but I have a lot of data credits left over, or vice versa, right? So especially at the enterprise level, you're gonna have these scenarios where they're probably gonna feel forced to say, okay, we you can take like some of these, like this pool that you haven't depleted, and you can spend it over here. We're gonna like credit you or we're gonna do some things because fundamentally there's just a friction. Because right now, I I have to estimate two numbers, and I'm very unlikely to like run out of those two numbers at the exactly the same time. So if I'm out of sync, I consume one faster than the other compared to my estimate of how fast I was gonna consume it. You're gonna have this drive by customers to want to like dip into the other pool to close the deficit they had in the first pool. This is fundamentally what fungibility is in these sort of credit systems. And they've just they they now have two. And then the last thing I'm noticing here is that some of the things that they're putting in as features are actually uh interesting, especially from from a from a credit management perspective. So, for example, they have role based access control. So they have a few things where uh where I think that what Clay is doing, and I think it's the right thing to do, is like, yeah, it's a workbook level credit budgets and viewer role. So they're basically saying it's an enterprise feature for you to be able to. To whatever, buy a million credits and then not let everyone have access to it. So you can say, hey, like Sarah gets like half a million and Mike gets 100k credits. And that's how it's going to be used. So everybody can't consume everything. Those kinds of features as it actually pertains to your pricing model and mechanism are going to be hugely valuable. And I think that that plays really smart for being this upfront with it and actually putting it as an enterprise feature because those kinds of like cost control and like account control mechanisms are something that that only a very few people have and customers really, really want.

SPEAKER_00 22:41

Yeah, I think that's

How much do I need, and selling against an estimate

SPEAKER_00 22:42

a really strong differentiator for enterprise. I have a I I love, I have a bunch of thoughts on what you just said. I I really like um your point around some of the, it seems like specifically like the growth plan has some features that could really drive usage, like auto-sync and enrich CRM. Like that just seems like something that like maybe should be in launch. Like I kind of see what they're doing here. It's like they want to have a plan for kind of like startups, and then they want to have a plan for like established businesses, but everybody uses CRMs, and I think it makes tools a lot stickier and probably leads to more usage. Like I would just be very surprised if moving that down to launch actually cannibalized a lot of revenue, but you know, maybe that's the case. Well, one thing that I think is interesting about like actions and credits, I think it's really hard for a customer to estimate how many data credits they'll need and how many actions they need, which I think lends itself really well to kind of like a consultative sales process. When I was at HubSpot selling HubSpot, we had these two metrics in our pricing strategy. We had the number of contacts and then the number of emails that you could send, uh, which we, which was a multiple of 10x your contact limit um per month. So if you have a thousand contacts in your HubSpot database, you could send 10,000 emails through HubSpot per month. And so, like part of our consultative sales process is like figuring out not only how big a prospect's database actually was, but also like how many emails they were sending, like how many campaigns they had going out. Because if you have a thousand contacts, but you send 30,000 emails per month, then we would need to lock you in at the 3,000 contact level, right? And so it lends itself really, really well to um like a consultative sales process where I think like Clay already has this kind of like evangelical educational vibe with their company where they're kind of like teaching people to become GTM engineers. So I think like all of that kind of like works really well together. I would be very curious like what the relationship is between actions and data credits and like how that scales. Because I think like that, that's like a that's a very interesting piece of their pricing strategy. Um, like I wonder if there is like a very clear kind of correlation between how many data credits you have and how many actions you're taking. There probably is.

SPEAKER_01 24:54

Like the number one question in like most consumption and users-based like sales scenarios is like, well, how much do I need? So like how many actions do I need? How many data credits do I need? How many API calls do I need? How many, whatever it, whatever it is that you're selling, like how many do I need? And sometimes, like for some use cases, the answer is really clear, right? So if you run a shopping mall and I'm selling you based on footfall, you have a pretty good idea about the footfall. Like you're like, okay, we do like 4.2 million a year, like give or take, but you like you know this number. But if you're building some sort of like whatever, agentic AI solution and I'm counting API calls, like I have no idea how many times my agent is going to call the API. Like, like it's like who who knows, right? The problem then becomes if the sales process requires you to make an estimate, especially if the pricing model punishes you for guessing wrong. For example, if I buy a bunch of credits and I don't use them and they don't roll over, I get punished. Like I misestimated, I bought too much, I now lose it, right? Or if they do roll over but only partially, or I have some sort of commitment to it and so forth. I think with if you have if you have something where customers need to estimate something, you need to sort of like think through how they could regret being wrong and then sort of take the risk out of the scenario by allowing rollovers, by you know, by having sort of discounts that scale like in a in a simple way. So even if they had so so there's no like say overage penalty. That could be another way. I could if buy too little and I need to buy more, do I have to do it like at an increased price? Or or is it just like whatever I would have paid, even if I had known up front, right? So these kinds of scenarios are are very critical to getting customers out of this analysis paralysis where they feel like there's this impossible, unpredictable piece of math they have to solve. And like between one and a trillion, that's the answer. And you know, like there's only one that is correct. And if I do any other thing, I'm gonna lose money and I'm gonna feel stupid. Because what it does is it increases your sales cycle and it likely also decreases your conversion rate, especially with these sort of hard-to-predict consumption scenarios. You need to build in a lot of generosity in the way the pricing model deals would be

Generous or overcorrecting, and the missing agency tier

SPEAKER_01 27:11

wrong on these estimation scenarios. Let's look at rollovers and some of the credit policies they have. I think, I think my impression so far of their pricing is it's still good, like it has a lot of good in it, but it's also, I think, a little bit less generous than it used to. Like in in some like so in in ways where both like they're charging for more things, but also they're they're trying to like prevent certain kinds of like gaming scenarios where they're like they're playing defense on some use cases or edge cases where they're saying, yeah, you're not gonna get away with that anymore. There's a balancing act here, like you want to do that to some extent, but not too much. I think one of the one of the ways in which clay won historically was by just being very generous. Like there was just a ton of value. I could do a lot of things where, like, oh, is this free? That's like, oh, great. Like now I'm I'm people leaned in. This obviously like reveals a lot of like scenarios where people get a ton of value without paying for it. And so, so, so the question is like, are you are you adjusting or are you overcorrecting? Especially like how they have designed the credit system will tell us, like, will give us another clue to answer that question. I think.

SPEAKER_00 28:22

I think like one interesting thing about gaming the system and about Clay specifically is you know, they they came up with this idea of GTM engineer. And when they did, there weren't a lot of actual GTM engineers out there, right? They were like the first company coining this like name and role. And so a lot of their earliest power users were actually like marketing agency owners and like lead generation agency owners that would use clay on behalf of other companies and like really ramp up their usage and kind of like become power users. There's like this whole ecosystem of like marketing agencies that use clay and kind of like teach their customers how to use clay. And I would imagine that you get this agency that's like one or two people and they're working with like a bunch of clients, they have like a ton of usage. They probably don't want to pay a lot. Like, I would think some of that dynamic is probably to try to solve for that and like try to figure out like how they can right size their agency users against you know their their bigger company users or something like that.

SPEAKER_01 29:23

I agree. And I think actually now that you mentioned it, I think that is like sometimes you you you miss what's not there, right? And so why not create like an agency tier? Right. Like a true agency plan for like, hey, we we know that like half you guys are agencies. So this is how it's gonna work. And so there's like build it into build it into a business model where we essentially say, we're gonna build the infrastructure for you to manage all your accounts because you need that. So it's like not only role-based access, but like a bunch of stuff around managing accounts, like deploying credits to each of them, controlling budgets in these customer accounts, um, making sure that you're maybe serving like two apparel stores and like not like Mike doesn't have access to both because like you've signed like competitive clause things, so forth, right? So all these things like that that matter to agencies, and then making sure that the pricing is really scalable. So essentially like cutting agencies in to say, hey, if you can generate like whatever, a bunch of business on your customers, you get to like buy it for like 50 cents and sell it for a dollar. Like the discount slope and the pricing on both the data and the actions would work in that way, and then essentially have allow agencies a method of essentially taking the infrastructure bill and then putting it onto the customer, right? So right now, if I'm an agency and I buy clay, I'm doing a bunch of things with clay for my clients. And maybe I have 50 clients, and now I have to take my one invoice from clay and I have to like break it into like 50 sub things, and I have to like put it on each of my accounts or like find another way to monetize it through my business. So I think they're they've sort of missed a little bit that to support the business model of the agency, where the thing that is powerful there is that the buyer and the user are two separate entities. Like the agency is the user. Well, they're also the buyer, but the payer is the end customer, right? If you can create this kind of distinction and rework a distinction like that, you're gonna have tremendous pricing power because the agency doesn't care. Well, they care to some extent, but they care less if you charge their customer more. So if the agency can say, hey, yeah, we have this spend of like, you know, the go-to-market platform that I told you about that we're using, you know, that's gonna be like 7K this month. But like we processed your million leads and we did all these campaigns and all that, so that's that's worth it. And you kind of have to pay that as a flow through cost. And then the brand they're working for is gonna be like, sure. And then it's taken care of. So the pricing power actually is somewhere else than with the agency, right? The agency just becomes like a flow-through between clay and the end customer. But the agency is the one that selects clay because they like to work with them, they like like the solution and all these things. Some brands are really good at supporting this kind of agency business model, and others just miss it. And actually, I think I think Clay would have it as a they would have it as a primary driver because it's such it's such an agency driven domain.

SPEAKER_00 32:14

Right. So they have like a whole solutions partner page where you can kind of like dig into that and become uh become an agency partner. It's funny, they're uh it looks like they like modeled their partner program after HubSpot. They have like four levels artisan, advanced artisan, studio, elite studio. And it looks like you get 20% of like pass-through revenue as an agency. I'm sure they have other incentives for partners, but like that definitely reduces the value for partners. So curious to see how how Clay continues to uh to evolve that program.

SPEAKER_01 32:48

We have that with another project that I'm working on right now. So around like similar size, like revenue as clay. We're building a clean system for them that makes a ton of sense. Um and uh and they've they've actually had like a pretty good partner product, which they call it. Um, and we can just see from from all the partners that we're talking to, like building out the partner program and so forth, is just like running this kind of infrastructure for them that both works commercially, but also from a from a sort of product point of view, is just hugely important. It will take care of like the low end of the market. So usually it's not like the enterprise customers that are coming in through agencies, sometimes it is. They have a good range of different customers, and then a very, very like large chunk of your long tail or smallest customers, maybe customers you would actually never have are brought in by agencies. So it becomes like these sort of aggregator customers, like an agent is like an aggregator for like 20 underlying accounts. Yes. And if that's sort of part of your go-to-market model, like this is how you get customers, um, you kind of give need to give the agencies something else than just, hey, you get a discount, right?

SPEAKER_00 33:52

Exactly. And and it's important because to your point, like these smaller accounts churn more. So if you have an agency that's working with them and doing all the work for them, it makes it a lot stickier than it usually would be. Um, because there are all these kind of like services and strategy around how they're actually using the product. They're not just like pure product users, it's usually part of kind of like a bigger retainer or a bigger, bigger contract. Um, so to your point, it's it's a way to not only aggregate customers, but also increase their attention of traditionally fickle small businesses.

SPEAKER_01 34:22

I actually sort of now that I miss it, like usually when I design these kind of partner programs, I usually reserve like for for sometimes the higher tier, sometimes for the too higher tier, like if you go and do like silver, gold, platinum, diamond, whatever, I actually put in like uh maximum churn levels. So like you cannot be a platinum partner unless your churn is like below whatever 10%. It ensures that partners start to worry about the quality of their service and their customer relationships like, hey, we're actually gonna give you like another 5 or 10% revenue share if you get your, you know, partner friends like your shit together and like stop losing customers and actually deliver good value to them throughout product. So it becomes a way of sort of incentivizing lower churn and like stickiness and better customer service through your partner program.

SPEAKER_00 35:06

I was just gonna say the uh you said incentivizing. I was just gonna drop the Charlie Munger quote. What is it? Show me the incentive and I'll show you the outcome or something like that. It's uh it's a slapper, it it always resonates. Um, I I love that. I love that uh that churn requirement.

Why every credit model ends up offering rollovers

SPEAKER_00 35:21

That makes a ton of sense.

SPEAKER_01 35:25

Let's look at the credit model they have.

SPEAKER_00 35:27

Let's do it. If you're not watching this, um, Clay has this like command K um like shortcut on their site where you can search basically anything. So I just searched rollover, I'm gonna do it again. Um, and you can see if you you just click and it'll take you right to where they talk about actions and data credits and rollovers and all that stuff. If we're looking at their credit model, like the the the biggest thing I think is like when I when I talked to, so I talked to Zona. Zona runs pricing for Clay. Um I've talked to her a couple of times, but the first time I talked to her was a few years ago. And it was right after they made a bunch of changes to kind of like the operations around their credit model. So they added the ability to do top-ups and they added the ability to roll over some credits. And she said the big reason a big reason was because customers had kind of like lumpy usage for Clay. It wasn't the type of thing where like every month it was the exact same credit burn, the exact same number of emails that they were sending. They would run different campaigns in different months, and that would drastically fluctuate the amount of credits that they were using and the amount of data they needed from month to month. And so customers pushed back a lot when the credits didn't roll over because they felt like it just like wasn't fair. Like they weren't actually able to use all of the credits that they were supposed to have in um in the course of their annual contract. And so they changed that. Uh, they have a pretty generous rollover policy now. Um, actions do not roll over, but data credits roll over up to two times your monthly limit. So that accounts for this kind of like lumpiness, but only to a degree. Like you can't run all of your actions or you know, consume all of your credits in one month. You have to have like some level of consistent usage. One of the things that John and I noticed when we were digging into credit models is the longer you've been offering credits, the more likely you are to offer rollovers to some extent. And if you talk to anybody who's offered credits for a while, it seems like it's just kind of like an inevitable progression where your customers are gonna push back. Like at some point, unless you have like a per a platform where the usage is the exact same month over month, which it literally never is, um, your customers are pretty much always going to push back and try to get a credit rollover. We'll be curious to see what you've seen in that space. It definitely seems like a pattern um based on what we've observed.

SPEAKER_01 37:46

So I think two things happen when you when you launch a new credit system. So so one of them is that it's probably like a new thing for you. So you're like, oh, we're gonna launch a credit system and you know it feels it's new, so it feels risky. So I'm gonna, I'm gonna, I'm gonna be a little defensive, so I'm not gonna allow the rollovers. And then when you launch the credit system, you find out, oh, it actually does a lot of things well. Um, you know, I can I can plan for it differently, whatever. Like you you get you you get used to the version that you launch. The reason that people allow rollovers after a while is that one, getting to know your own pricing model actually de-risks it. So now you understand it better, so you can actually be a little bit more generous. It's it's not that dangerous to allow it. And two, customers want it. So they're gonna push you, right? So you have this, like there, like these sort of two motions that that do it. Rollovers, they they don't look good in a spreadsheet because, like, okay, so customers didn't use the credits, so now we're gonna like just let them consume it in the next period. Like, so spreadsheet says that that's less money than if they just bought a new batch, right? So the CFO is just gonna look at that and say, why are we allowing rollovers? Like that seems to. And then so what you fail to quantify is that the sales conversation that happens at the beginning of the relationship is a lot easier if you allow the rollovers. So you say, hey, if we allow the rollovers, uh, a few things is gonna happen. We're gonna have a faster sales cycle, we're gonna have a higher conversion rate, and we're gonna have people buy more credits. So if if they're not gonna roll over, I am now heavily incentivized to not buy too many. If they roll over, I'm gonna have this sort of dreamy conversation with my sales rep around what I'm gonna use the part for. Right? So so if the rollovers are there, suddenly my ACV up front is gonna be a lot higher. So what's gonna happen is that, well, then I'm going to probably spend more credits when I have them. So, so in almost all of these sort of credit scenarios, and we see it in in consumer scenarios, we see in like small businesses, we see large businesses, that if I have credits in my account, I'm more likely to spend them as opposed to if I don't have them and I have to sort of like buy them, pay as you go, or like go and ask like for budget to buy them and so forth. But it doesn't show up in any of the initial modeling when I'm launching the credit system. It usually does after a while, once you you get the hang of it and you learn it, it's like, oh, okay, so if we just let them roll over, they're gonna buy more, they're gonna back by faster, more gonna say yes, and they're gonna spend more. Okay, so no, that's a lot of good things happening all at once, will allow it, right? So what I was wondering with with Clay's rollover plan is that so they say actions don't roll over. Okay, so I I think rollover policies, and they're not differentiating it here as far as I can see. Maybe you can find it on the pages that they need to be thought of differently between monthly and annual plans. So if I'm on an annual plan and I only get like two months of rollover, what's that? That's not even 20%. Yeah. It's like 16, 70% of credits. Okay, so it's it almost feels like a it feels like I lost all my credits, but you're kind of giving me a handout. I would be then even be more pissed off than getting like a 16% rollover than just getting none at all. If it's a monthly plan, I think two months of rollover is the right amount. Because that means that I could buy them in January and then I have to spend them by March. And if I don't, like, okay, then now it's on you. Like you you you've had like three months to consume what you thought you'd do in one month. Usually the rollover I suggest is that for monthly plans, you let it rollover twice. Yeah. I think I even had this conversation with Sonos, maybe I can say Chris. I don't know. But but other than that, yeah, I think that's fine. And then um, and then for annual plans, I'll let it roll over once. Because that allows you to have reasonably close like revenue recognition. Um, you can still plan for it. Um, it it solves nearly all of the like the sales psychology, like people are fine with two years, um, without giving you sort of any like issues on the on the in the spreadsheets and the in the in the rev rec. So that's what I would normally do.

SPEAKER_00 41:45

So you would allow a customer on an annual plan to roll over all of their like basically a full year of credits?

SPEAKER_01 41:53

Yeah. So usually when we talk about rollovers, we have a few mechanisms that we can work with. So one is the amount of credits. So all of them, none of them, some of them, right? So do we do partial rollovers, do we do fullovers, do we do any rollovers? My argument here is that if you're gonna if you're gonna have credits roll over, let all of them roll over. Because if you only let, let's say, half rollover, you're gonna get you're gonna get almost all of the negative psychology. Like people are gonna hesitate and not buy too much and so forth, almost to the same extent as if you let nothing roll over. So you might as well just let everything roll over. Because as soon as you go below 100%, go into partial rollover, you get almost 100% of the negative psychology. So you get none of the benefits. So it's like, okay, so for me, it's like it's a binary, like either full rollover or no rollover. Like the partial is is like it should work in theory, it just doesn't like because of the way the human mind works.

SPEAKER_00 42:45

Totally. It's one of those things where there's still, to your point, you want to reduce all the friction possible to like get somebody to use your credits and use your product. And and with a partial, there's still that kind of like unanswered lack of clarity of like whether or not you're in the clear. And yeah, I think psychologically that stuff um can be a total blocker.

SPEAKER_01 43:04

The other mechanism you have is that you can make you can make the full rollover contingent on something. Let's say a very simple thing is you keep being a customer. Like you have to like still have an account with us that you pay money into like real dollars into. You can always you can also do something like you keep buying credits at a certain volume. And then you you can say, well, we're gonna let that look well, how many credits do you need to buy? It could be like a minimum amount, you just pick a number, like 5,000. It can be whatever you consumed last year, that that's what I would call transpose consumption of credits. So I call it transpost credit rollover. It's a little bit technical, but it actually, like if you want that, if you want accounts that don't build up too many credits over time and they smooth out and you still want to have the full rollover, having them commit to buy at least as much as they used is a really, really good mechanism.

SPEAKER_00 43:56

Yeah.

SPEAKER_01 43:56

You kind of want to have it mostly like a sales led enterprise. Context because that's like you need to explain. But the other thing you can do is just to say, hey, you need to buy at least as much as you bought last year or like 80% of it. This is what Dunn and Bradstreet did in the 90s, like pre-internet stuff. They sold market data on a credit model and say, hey, we're gonna sell you $100,000 worth of credits and we're gonna let them roll over if you keep to buy $100,000 worth of credits. That's the deal. And that they had a like a billion dollar business on that.

SPEAKER_00 44:25

That is a badass name drop right there, Ulrich. That's awesome. That I love when people can pull examples from outside of like modern

Top-ups, overages and multi-year contracts

SPEAKER_00 44:34

PLG SAS done in Bradstreet. That's an awesome, awesome drop. Okay, so we talked about rollovers. Anything else you want to add there? Because I think the other side of that equation is top-ups, right? It's like what happens when I run out of my monthly credits or my annual credits and I don't want to upgrade to the next tier, especially with monthly plans. I feel like top-ups are really important because people don't want to feel we've already talked about the lumpy nature of Klay's product and how every month isn't gonna be exactly the same. And there are gonna be times when you're, you know, sending more emails or, you know, reaching out to more people, running more campaigns, whatever it is, and you don't want to have to upgrade and then downgrade like every month. Like that's just super annoying. And so Zona had mentioned this was also another concern. Um, one thing that I think is interesting here that they call out in their FAQ, which Clay has a great FAQ. Um, they scoped the launch and growth plan so that 90% of customers never hit a usage limit. And I know Zona had mentioned that with actions as well. They were very intentional. They looked at um, you know, the histogram of usage and and and really believe that they set the limits on actions to a point where it wasn't gonna trip most of their users up. Um, they've done the same thing with credits. A year and a half or maybe two years ago, they introduced top-ups um to allow people to buy more credits, I think at a slight premium rather than having to upgrade to a new plan. How do you feel about top-ups, Ulrich? And how do you usually think about that um when you're working with customers?

SPEAKER_01 46:02

Let's just like separate the the options you have. So, okay, a customer buys a certain volume, whatever, 100, and then they need more than that 100. And you kind of have like the two options of either you allow them like an like an ad hoc non-recurring purchase of like another hundred, but just for that year. Or you commit them to a new higher level. So you're like, okay, now your subscription is for 200, and then like next year I'm gonna build you another 200. So those are basically like the options you have, right? So some companies actually like punish you for both scenarios. So if you come like mid-year and you say, hey, yeah, I actually needed 200, they're like, Yeah, you should have known that in January. Now I'm gonna like, I'm gonna not only like upgrade you to a higher subscription, but I'm not gonna give you the discount that you would have gotten if you had like just told me straight away. Customers obviously don't like this, but uh, but but there you are. I would roll out the red carpet, like if you want to commit to giving me more money in the future, I will like let whatever discount I would have given you on day one also apply now to just like do it. What happens is actually that sales incentives are usually built on their upfront commit, especially if like if you let the salespeople like design the model, they will sometimes be like, hey, hey, like we want them to be punished if they come later and need more, because I wanna, I wanna like, I wanna make money now, right? I want I want them to commit up front. And then, you know, cash today is better than cash tomorrow, so you can get your CFO in on that plan or like so. There's just a lot of like, there's a lot of people that can agree that maybe that's a good idea, but it really is, right? So you you kind of need to, okay, this is the discount. If they want to like commit to more, we're gonna let what we're not gonna punish them for it. Tob ups can be a little bit different because the top-up is a customer saying, you know, I said I needed 100 and I actually need 200, but uh I'm not gonna promise that in the future. I think that's okay to punish them for that, because you could you could you can then say, hey, I'm gonna commit, um, I'll give you the 200, but that let's now agree that you need 200. So you're also gonna get 200 next year, and then that's just a new normal. And with that, I'm just gonna like let you like pass freely through our like discount system so you get the best possible price. But if you're telling me that that this is sort of uh like not what you want to commit to, and you just want the extra capacity like out of the blue, and so I can't plan for whether I need that next year, you just want it now, then that comes as a premium, right? And I would use the uh the friction between those two to get them to commit to a higher subscription plan, right? So I think like normally overages in that sense is like you know 20%, something like that. Sometimes I have a CBD like twice double. Yeah, you whatever makes sense for you. Um, but I think having those sort of like ad hoc top-ups that don't roll into recurring um at a premium, that's fine.

SPEAKER_00 48:45

Totally. Yeah, I I've seen some gnarly premiums there for sure with uh with top-ups, but I I think it's warranted because I I think top-ups, to your point, like there's a great there's a huge incentive there to shift from monthly to an annual contract, right? Because like that gives you your full pull pool of credits like at your disposal whenever you want them. Um so I'm sure the clay sales team uses that as one of the levers to try to push somebody to an annual contract. If you're still not willing to do that and you want to stay on monthly and then you consistently need more credits, yeah, I think you should pay a premium on uh on those credits that you're that you're adding.

SPEAKER_01 49:20

Nuance to that actually is multi-year contracts. So grenades and multi-year contracts. So I have a three-year contract, I'm buying whatever, a million credits for each year. Do you get all three million into your account on day one? Which is actually sort of I'm actually sort of, I can see little pros and cons to both. What most of like what we end up doing in most of the of the projects I work with is that we actually don't give them three million up front. We give them one million so that we have this opportunity for them to run out. So, like into let's say in like deep into year two, they suddenly hit like their limit. And now we use that as an early renewal motion to say, hey, you need you need more than you thought. And we can either upgrade you or like, hey, there's also these five other things I want to sell you. So we actually need you to sort of pull the contract forward and get an early renewal on it. That's that's a pretty strong argument, but but it doesn't work. Some people try to do it inside an annual contract and say, well, we we're gonna give them like each month, but that that takes a lot of the argument away from buying an annual contract. Why do I need to pay you up front if you're not giving me like all that stuff now? So for whatever reason, it seems to like not work in an annual context. You just give everything up front, that's the only thing that works. But in a multi-year context, most customers of FI were actually only getting like an annual allowance.

SPEAKER_00 50:32

Yeah, I mean, uh, like look at uh look at Uber. I I think Uber ran through their AI budget in like March this year. It seems like like I don't know exactly how they were using it. I would I would think Uber, if they're working with Anthropic or OpenAI or whatever, would be on an enterprise contract where the volume was agreed upon up front, but maybe they were just using the APIs and just letting it rip. But like that feels like a missed opportunity for OpenAI or Anthropic to kind of like go in there and right size their usage, right? It's like you just burned through your like you burn through your annual uh level three months into the year.

SPEAKER_01 51:05

I think we had like a period of choke and maxing there where no, so I think I think CFO

The verdict, buy hold or sell

SPEAKER_01 51:11

sort of entered re-entered the picture at some point.

SPEAKER_00 51:16

At the end of every episode, we do a buy, sell, hold, classic Wall Street. And what we're going for here is just like a buy is love their pricing strategy. A sell, don't love their pricing strategy. A hold is, you know, it's good. It's okay. I think they're buy, they're not a strong buy.

SPEAKER_01 51:34

And I think they're I think there are things that they could still work on. And I think so. When they have launched this, they will realize that they have introduced new friction they didn't have before. And they will they will hit their heads on it and they will sort of figure out how to solve that. But that doesn't take away from the fact that the fundamental structure and what they have actions, data credits, and a reasonably good like separation of functionality across tiers is really strong.

SPEAKER_00 51:58

So it is about agreed. Agreed. I I would buy it as well, and for the reasons that you just said, I think the underlying structure, which we always talk about is by far the most important thing, is very strong. I think there are definitely some kind of nuances that could make it better. But um, in general, I think it's there's a good reason why people look up to Claire right now. And it there, there's a good reason it's a there's a good reason why why they've kind of become like um the go-to example for so many of these companies.

SPEAKER_01 52:26

Absolutely. Yeah, let's let's uh I think I think this is again, we say that every time, right? But I think this is definitely one of the companies where we should like check in like at least once a year and sort of see how they're doing. Um, 100%. I'm gonna be curious if if they're gonna maybe make some adjustments to this based on some of the friction that they're seeing. But I think a lot of what they lost in their latest update is gonna be like a like a fundamental sort of driver in their business model going forward. So good for them. That is it for this episode of Pricing Page Unpacked.

SPEAKER_00 52:54

If it was useful, subscribe to the podcast on your podcast provider of choice or find us on LinkedIn, the Pricing SaaS community, or on YouTube.

SPEAKER_01 53:01

And remember, your pricing page is not just a page, it is a strategy statement. See you on the next pricing page on packed.