Credit Systems (& AI Pricing)

guys so you're a few minutes early so you get to do the sound test with me so let me know over in the chat if you can actually hear me

And yeah to all you thank you for. Giving this another shot. So this is sort of the redo of the of the failed attempt Thursday. So I'm actually sort of since we are we're only marketing to the people who we burned Thursday. I'm not expecting us to have. A ton of people in the live session today. It's just the nature of how those things work. And but we're going to wait a few minutes. And then and then we'll see. So.

We are at 2 o'clock Central European. I think that is 8 a.m. Eastern. So we are about to begin. I will start within a minute or so. So thank you for joining. If you're joining you will probably burned Thursday when we had a technical error. It was a technical error caused by a human error. So apologies for that. But thank you for making the time again to do this redo. So in the nature of things as we're doing a redo, I think we will obviously record the webinar and we will send it to everyone who were signed up originally. And and also because of the nature of the redo we probably have going to have a lot fewer people today than we had yesterday which means that the likelihood that we're going to have a ton of questions is like a little bit open. I don't know how many questions we're going to have a really like questions. So what I'm going to try and do actually is because we don't have a lot of people I will keep the live stream chat open. As a talk and then I will try whenever you have questions just fire away and I'll try to sort of catch it sort of mid presentation and just pause and have that conversation there. So I think that actually will work out fine as we're going to be less people than usual. OK. But with that said welcome to the webinar on credit systems and AI pricing.

I have an agenda for us which is to talk about what credit systems is what it is and how it works and how to sell and communicate credit systems and then term conditions and revenue recognition of credits. So what you will notice here is that it doesn't say specifically AI anywhere and it's because there's no rule that AI should be priced with the credit system or that credit system should only be AI. so it is not sort of perfect overlap. It's simply recognizing the trend that a lot of AI systems are being priced with credits. And so I'm going to let's say tilt a skew a lot of the commentary as I go through credit systems on how I think it sort of specifically applies to a lot of use cases of AI products and also have a lot of different sort of pricing pages that use credit systems and all of those are. With one notable exception from AI products. So so that's the idea. And for those of you who do not know me, my name is Ulrich. I'm the founder of willingness to pay.com. We are a niche specialized B2B SaaS pricing consultancy. We work globally with a lot of name brands and a lot of let's say scale ups. And we've done yeah several hundred projects in the b2b says pricing space. have been doing credit systems. At least for the past 7 years and I have been doing them in a lot of different verticals from maritime to marketing tech to financial tech.

To project management and so forth and I did it way before AI. I've done it in AI as well. And it's always been. It's sort of a little bit of darling when you're a pricing consultant because credit systems are fun. They're a little bit complicated. They can be at least they can provide a tremendous value unlock. And. So it's really it's really it is a very powerful method if used correctly and and I want to tell you a little story which was sort of the first time that I used the credit system. I think we're back in. 2018 early 2018 I was in at that point my firm was in a sort of co-working space with a lot of other. Tech companies. over the low tech companies and one of them was icon finder.com. So this is essentially a stock photos business just for icons. So if you want a little icon of a man or a bicycle or a car or doctor or something you like what most people do is that they will just search the web and they will say free icon doctor bicycle whatever and then the sale will spit out different sites and their site is one of the largest ones. If not the largest ones, think they have way over. This is back then seven years ago. I think they had way over three or four million different icons at that time. And they would they had a model where they said OK, we either have a pay as you go so you just like buy an icon and you get the right to use it a little bit like stock photo or you have you buy a license with a right to download a certain number of icons. So you buy the right to download five icons 10 icons or whatever. And the founder had a massive churn problem because people would go in, would, even if you could get them to subscribe, they would stay on for the 5, 10, 20 icons for a while and then they would be returned. So this was the first time where I...

gave anyone the advice to say, hey, instead of selling them the right to download five icons a month, why don't you sell them five credits a month? And then if they don't use them, then they roll over to the next period. And I had like, I think 10 or 12 other things that I thought I also thought that he should do to monetize better. But basically none of the other ones really mattered because as soon as he instituted the credit model to the icon business, His churn drop from something like camera what was but it was around 20 % a month to basically Stop 5 % a month. So we've literally caught the churn to 1 fifth of what is worth before and then that also meant that his customer lifetime value suddenly like 5x and he was actually able to avoid a dilution financing round that would essentially Making loose control of the business so so it was it was a it was a very big sort of easy almost too easy fix for a business to do something like this and I think since that I've always thought about especially if we have very if we have businesses that have a lot of the Ryan's in the user so I need a lot of icons I need a lot of X in January, but I need nothing in February and then I need a lot again in in March then that the Ryan's can of course be priced on a user's basis or a pay as you go basis but it's really hard to price with a license just a right to five icons and that's where credit systems can come in. So I think that's sort of the origin story of the first one and then we have multiple. When this is done you will receive an email in the email there will be this recording and there will be these slides and then also there will be a link to this community dot pricing says dot com so these is our partners over pricing says they have a community it is awesome I am part of the community you should be too. For no other reason except that you know if you want to know about pricing that's the place to go and any follow up discussion Q and a etc. Will be over there so the slides.

and the recording would also sort of permanently live in the community and whenever you have any questions around it just you know go into the community tag my name and I will get a notification and I'll probably go in and answer it so even if you're just watching the recording here and you still want to ask me a question go to community.pricing.sas.com and it will take care of it. Okay so with that let's get going. What is a credit system? So I thought about it for a while. sometimes actually I've been using the hashtag usage based pricing paid up front is not sort of the most eloquent hashtag but it's sort of it's it brutally explains what I think credit systems actually are is that they are a special hybrid almost between a license and a uses based modality. So. In my book there are four different pricing modalities or ways in which you can execute a pricing model and that is flat fees just your customer you pay a thousand bucks a month whatever it is. There's a flat fee then you have a license so we count something number of users and you pay for the right to use that is usually prepaid so it's paid at the beginning of the period and then use it is up to you as the as the customer. And that is opposed to usage based modalities where you get access to a product you then use it and then at the end of the period we look at how much you used and then you pay according to that usage. Credit models are hybrid because what happens in a credit model is that you start a subscription. And then upfront you pay for and get a number of credits. 10 credits for 10 icons whatever it is. Then you have a period that can be a month, a year, whatever it is, where you then spend some of the credits. There's a lot of edge cases. We'll get to them later. Then when the period ends, if you haven't spent all of your credits, if you only spent seven and you have 10, then the remaining credits get rolled over so they get credited to the next year. And then when the new period begins, and this is crucial, you pay for and get another 10 credits or whatever the number is.

So each of these steps is crucial because they do different things for the credit system. So first of all, Because it is a recurring purchase. I buy 10 credits every year, whatever it is, it is a subscription. And because it is contracted that there's a certain number that I buy, it is also annual recurring revenue. So it is not reoccurring in the sense that usage based might be reoccurring. It is actually contracted that I buy 10 credits a year or whatever the number is. So given some different nuances in the terms, which I'll go to later, I have had this approved in both GAP, Generally Accepted Accounting Principles, and IFRS. So internationally, I can't remember what the acronym is, but the non-GAP, the global sort of internationally accredited accounting system. As well so so they can be both subscriptions they can be recurring revenue they can be recognized as such and also from from a contractual basis that is exactly what they are. So this is sort of basically what it is and then there is a lot of different nuance here so I have broken this down to 12 questions so I call these the 12 necessary questions to answer when designing a credit system. And I think it's because...

Even if you lack one of these, you do not have a credit system yet. You have an outline of something that could turn into a credit system. But I think these are the necessary elements for you to answer that you should go through. So if you are in the process of designing a credit system, you should go through these 12 questions. On some of the questions, I think that there's a lot of different options. On other of the questions, I think there are some options, but I think a lot of them are wrong and I'm going to have a clear So I'll go through each of these 12 in order. I have ordered them a little bit into these categories where the first four basically talks about sort of the core dynamics of credit system in terms of what does the credit represent? What can I buy with it? What does it cost? Etc. And then how we communicate it and what then the terms conditions and revenue recognition aspects are of the credits. that's what we have there. first off, let's get started with the first question. So what is a credit system? So there are four of these questions. So the first is. Can credits by one thing or several things so I call this multi or single purchase systems. We could also call it multi or single metric system so as in pricing metric as in price per user price per store location whatever it is and the question basically is here. Is it one credit equal one other specific thing? Or is it one credit and then that gives me access to a lot of different? So I'm going to give you sort of two examples to explain here. So the first one is audible. So this is Amazon's audiobook business. So if I buy audible for the one here on the left, the premium plus it says for seven pounds 99.

Here a month I get one credit a month to buy an audio book from our entire collection yours to keep even if you cancel that last part is actually new it used to be that even if you cancel you would lose the credits so you had to sort of keep the subscription here, but apparently they found out or are testing that they might be able to recover you if you. you get to keep your credits. So the idea is every month I pay eight bucks basically I get a credit I can buy an audiobook with the credit. Can I buy anything else? No. The only thing that Audible sells is audiobooks and that's what you can buy with the credit. So this is what I would call a single purchase system. So they only have one metric which is audiobooks and there are no different kinds of audiobooks there are no add-ons there's not anything else that's what it is. If you then go to something like Eden AI. I have just like copy pasted in the top of their pricing list. I think actually they have something like 15 or 20 of these different roles. Where they say OK so the feature is an image and the sub features background removal and the model is sent aside and the price for doing this thing per 1K files is 75 cents I buy. credits that are dollars. I'll get to that in a second. So it's basically okay. So I put in a hundred bucks into my account and then I can buy all these different sort of actions or metrics through Eden AI. So they have Invoice parser object tracking a thing and they just have different prices as you can see the image also ML classification is profile priced at 0.0005 dollars so. 5 % of a cent and then they have different price.

So if you're selling access to, let's say, a platform or an ecosystem with a product that has a lot of different ways in which it can generate value, then often you run into multi-purchase credit systems. So let's go back here. So can credit by one thing or several things? I would say that the single purchase system is really good if you have one thing to sell but that one thing is just really variable over time in like it has seasonality in the volume so January is high February zero March is high April zero so if you have this sort of. Very yeah uneven or seasonal. Volume or you let's say you sell I've sold another credit system to marketing agencies that really needed them whenever they had a project So it wasn't even a specific time of year. It's just like whenever business was good They needed a lot of the product and whenever business were bad that needed none of the product, right? So we need to sort of scale up and down but we didn't want to use this base system because we wanted to know when We could get the cash flow and we wanted them to commit a little bit to actually using our system Multi purchase systems are good when you really have a lot of very very different use cases inside of your system. So for Eden AI for example, I presumably can do a lot of different things with this product. And you know depending on what I do most I then get to pay for that thing. So this is really where the multi purchase systems come in. That doesn't necessarily mean that I have. High or low seasonality or that I'm using it more in January than in February. I might be able to actually have very stable use. But what happens is that I have very different use cases inside of the product so you might have one user that you know needs to have a lot of invoice parts parsed and then you might have a another user that needs a lot of background removal right? They might not be the same user at all, but they might be perfectly good types of customers for Eden AI.

So we want to price them differently. And that's really the difference here. Okay. Then can you actually measure, track and invoice the stuff credits by? I don't really have a screenshot to explain this, but the idea is that whatever it is that we have, we need to have it so concrete that we can track it and we can track it just as if it was usage based pricing. And We actually need to track it to let's say a very high degree of. Precision and also in terms of time delay and so forth so it needs to be real time and it to be precise when you just be able to order it because we need to send invoices based off it. So it is more difficult to have credit systems if we have some sort of a say manual or human measurement or evaluation of whether something took place or not. And that can be difficult. So you can have say you can find a license systems that are a little bit more vague with regards to this. We're going to have whatever the right for 20 production lines and then because the production line actually only measured once a year when you have to sort of walk the factory floor and actually see where the system is plugged on that can work well for those types of businesses, but it generally doesn't work well for credit based systems is the measurement and billing of of credit system has a very high threshold that needs to be pretty tight. So which is also why we actually see credits often being very technically defined. So for something like this for even its background removal the model sensor side as a precise price per 1000 files right and then so on we go with audio books. have one audio book and. And I'm pretty sure that they're going to force anything even that.

Vaguely might or might not even be audiobooks that I can still get access to like cover pages or descriptions like they're just going to cram everything into this definition. So it's just a part of. Of what is let's say viable as an option for something to be priced as a credit. The base price of a credit is at $1 equals one credit as you see here for example audible does not do this it is let's eight pounds for one credit and this generally is OK if you have a single purchase system right so if you're only buying one thing then it's actually pretty simple that I know that an audio book is is is eight pounds like I can do that math pretty quickly because the only thing I can buy with this credit is an audio book. The problem is that here for Eden AI they actually do something else they say hey one credit equals $1 and then now I have the dollar price of everything in here but imagine that these were actually not dollars that they were just credits and then one credit was two and a half dollars. Then suddenly it would be way more complicated to understand what the price of the evening I was because I had to do sort of two pieces of math in my head. First I needed to have a dollar to credit conversion and then I needed to have a credit to metric conversion before I could do the math of well how much does it cost me to parse an invoice. So what we see is that generally whenever we have single purchase systems like audible, we can do whatever ratio between credits and dollars that we want. That's pretty easy. But whenever we have a multi-purchase system, actually want to have a very, very close relationship between dollars and credits. However, I do not.

And I repeat I do not advise you to make the credits the dollars. So what Eden has here is that they actually ask you to deposit dollars into the account. So they say hey if you give us $100 we will credit $100 not 100 credits but $100 to your account and you get to use it. And the reason for this is regulatory. So if you start to hold people's money in account, you effectively can be regulated as a bank. So you will now be asked by the SEC or whatever financial regulatory body that you have in your country to prove that you are holding your customers money, not their credits, but their money in a safe way that you are not being used for any anti-money laundering or money laundering purposes and so forth. Unless you actually have a banking license and you're highly regulated to begin with, I strongly recommend you to not have dollars as credits. You can definitely have the exchange rate be one to one. So just say, hey, $100 by 100 like unicorn credits or whatever it is. And then you're saying hey one unicorn credit gives you right to buy this other thing. That's fine. You just can't have the dollars being the account because of the regulatory issues. So there is a difference between actually having dollars and just having the dollar credit parity. And so yeah base price of credits is usually one to one you can also I think I have other versions of it. Let's say a little further down here. Here we have we have a similar like here. We have a one to ten ratio not true one to hundred ratio where something along the lines of 50 bucks gives me 5000 credits. This is Lindy AI or 300 bucks give me 30,000 credits.

And I think importantly, even though we go from pro to business here in this plan, they actually retain the ratio and the ratio is pretty easy to do in your head. I would probably make it easy. I would probably say 50 bucks or 300 bucks. They're trying to use the charm pricing of doing 99 base pricing. Just make it one cent cheaper. But actually I think they're shooting themselves in the foot here. because the exchange rate now becomes less clear. I just have to spend an extra second doing the math that it's basically $1 for 100 credits. So it's one credit a cent in this setup here. But generally if you have this $1 equals one credit or $1 equals 10 or 100 or something where it's very easy to do the math usually in factors of 10 then that's what you want to do with a multi credit system. And then you now start to add discounts. So we have audible here 8 pounds for credit and then you can buy 12 credits on a page you go model. can buy two credits a month or you can buy 24 credits for an annual subscription and if you do the math here, so you said well one credit was 8 bucks right and if you'd buy two credits a month you get it for 15. So you you know you save a pound. That's great. actually makes it a lot cheaper right so instead of. Let's say paying 180 pounds a year then suddenly if you go annual you just have to pay 110 so you actually save 70 pounds that's a very 35 40 % discount.

So what you can actually do is you can say, hey, you can make the credits cheaper as you buy more. That is a very common practice. I would say it works especially well when you sell something that is very equivalent to a dollar. So I have had a lot of credit system, which is my preferred way of solving this, where I one credit equals one dollar. And then in certain volumes, I get to say, hey, you're now buying the credit for 70 cent on the dollar, 60 cent on the dollar, whatever the discount is at that volume. Because actually it feels to the customer that is being sold like you're selling your discounting dollars, right? You're saying, hey, this is worth $8. We've established that as the base price, but I'm selling them to you for 50 cents. Wow. So you get to spend a dollar inside my store, but you only pay 50 cents for the dollar. So that is a perfectly fine way to handle the discounts. But the base price the initial entry level price needs to be one to one or one to ten one to. So this is a little bit also my let's say best practices that are rolled into this one here. and we have Pavel thank you Pavel for asking the question. I don't know how long you've been sitting on it. You say hell rig one of the reasons larger companies for example enterprises don't like pay as you go is lack of predictability of their spend the credit system is essentially a layer of abstraction from the underlying pay as you go and credit requirements also need to be sized somehow for example. How do I know how many credits to commit to what is your recommendation and approaching this? How do you enable your sellers and customers to size the deal appropriately so?

Absolutely, I so I have this Question number eight. How is it German? How many critics custom by the beginning? So I'll cover the last part of your question In a little while Pavel, but I would say that you're absolutely right that that from a psychology point of view There is a very clear dynamic where customers feel when they're buying credits that they're just like depositing money with you, but they're actually not losing them yet. Like they haven't spent it right. They can they can decide on the spent later. So it's actually easier for them to commit to something upfront compared to if you were selling a license model. So to say hey, why don't you deposit like 50,000 credits with us worth $50,000 and then you can you know you have time this year next year to really decide what to spend it on. that doesn't feel as hard of a sell usually as selling for 50k worth of licenses where they're committing to the usage. So that's good. And then importantly, when they now have the credits, depending on who can use them, what you often see is that customers spend a lot faster than they otherwise would even compared to usage based models. Because usage based models, even though you don't have a commit upfront, what actually happens is that people tend to say, well, You know, if I push this button, I do this thing, I know that there's some kind of micro payment. Like it feels like I'm paying for the thing whenever I'm engaging with it using it. But with credit system, then you say, hey, that's a sunk cough. I already have the 50,000 credits. Let's just burn them, right? Which means that people actually adopt faster in credit systems. So credit systems, bar none, have a very...

way of getting adoption because people almost use like it was free right so and they buy like it was usage so so that's a that's a really really strong sort of psychological cocktail and exactly as you say it feels predictable if I buy for 50,000 worth of credits I budget 50,000 right that's my spend and then I know that I should be able to get it done with this spend. If they then run out in September and they have to buy more for the rest of the year, you know, who knows? They probably got more value out of the product and then you can have that conversation with them then. I'll talk a little bit more about sort of initial volumes in a second. So how to sell and communicate credit systems. So this is how will you explain credit customers and how will users keep track of credits? It is... It is some of the I almost initially actually I had only 10 questions and I I put these two communication questions in because I just thought hey they are actually critical and my best sort of. Example is to say hey you want to be able to explain what it buys so that the purchase like you can buy audio books or you can buy all of these. Different sort of technical outcomes let's call them that. And that's really what it does. And then you need to make the terms that simple. So I'll cover that more in terms like how many do I need to buy upfront? What happens at renewal? What happens if I run out of credits? Can I get credits back? All of these things. So essentially what you would say terms conditions and recognition all the questions from seven and on is actually part of communicating the credit. And if you do one. create a very complex set of terms and conditions, it just becomes harder to communicate and sell. So I'm actually a proponent of whenever you start a credit system to make it as simple as possible. So and when you outline it and you show it to customers, whenever you have questions that take more than just a second to answer, you force cut the complexity.

until you make it dead simple, right? Even to the point of, we could charge for these things in our multi-purchase system, but let's cut it. I'm actually doing a project right now with a 40, 50 million ARR US based company. We're building a credit system and the initial list of things in the multi-purchase system we had was maybe 30, 40 line items. And you said, hey guys, we like. It's great, we might end up there in a couple of years, but when we launch this, let's get it down to 10. There's going to be 10 things that really drive the value of this product. And we don't want to over complicate it. We don't want to overwhelm our sales people, our product people, all the customers. Let's just make it simple. And then in terms of the terms, let's just make that simple too. Like there's a commit, it has a price, you need more, same price, it rolls over, same volume. So there is a real benefit, especially like in version one of your credit system to just make it like err on the side of simplicity. More simple than you think it need be you can always add the complexity later In terms of how users keep track of credits this can also be seen in terms of a timeline Are you launching it now or is it an existing system? I think whenever you launch something it is fine to just have it in a spreadsheet that finance runs as a test and Then basically when the customer asks well how many credits? How do I know how many credits I have left? Your response basically is, well, you give us a call and we'll tell you. We'll look it up and then that's it, right? And we'll keep track and have an auditable track so whenever you want to challenge the invoice, we'll look at it and we can sort of open books on it. You want to end up where it's part of the product.

What's actually built in and a customer dashboard like just just as you have an administration module where people see who has used its rights and who has login details. You basically want to say hey, this is where you can see how many credits you spent how many credits you have left and whether you are sort of above or below the trajectory of that year. So let's make the math simple. Let's say that you have you know. 120 credits for the year, the year starts in January 1st. Let's say that in April you have spent 60 credits. So in April you have spent half of the credits. So here the dashboard should say to the user, hey you're spending at a faster rate than we would expect if you were to spread out the credits evenly over the year. And then if the customer says well I'm in the whatever summer holiday business so I'm you know I'm going to spend all my credits now because I know that my seasonality is such that I don't need credits at the end of the year so it's fine and then that's okay. Or the customer might say hey. I'm actually like getting a lot of value out of the product. need to buy more and then they can reach out. So this is also can be a really good tool for your CS people or account executives, whoever it is that does the expansion to just call customers say hey take a look at your dashboard like you're spending like a lot of volume in this product. Is it because you don't know how to use it or is it because you're just getting a lot of value out of it? Let's have that conversation and maybe even sell you some more. Do you need more credits? Because the next level would unlock a discount or whatever it is. So this can be a very good reference point for you and the customer to keep track of things and also to power the upselling. And that's usually how we do it. Okay.

So terms conditions and revenue recognition. So this is sort of the last block of questions here. So will credits expire when how is it determined how many credits by the beginning? How is it determined how many credits customers by renewal? What happens if customers run out of credits during the year? Who in the customers organization can approve to buy more credits? And what happens if customers churn are credits paid back in dollars? So. I will share my opinion on these. This is where I have more opinion than just best practice and then I will show you and you'll also get what I consider to be the best practice like template for terms and conditions. So. But let's go through these will credits expire and when? Yes is the answer. They will expire. You need them to. If they don't expire you're in trouble because you essentially have an unlimited term liability on your balance sheet. So if you sell people a credit and then they forget about you and then in 25 years you still owe them that usage. If they come knocking and say hey I want to parse this invoice You kind of have to let them do it right and then your CFO is going to hate you and probably also your legal counsel. So you need to make them expire. I usually recommend to let them expire after two years. Why? Well, because it is a good middle ground between the. Liability of your CFO and your legal counsel and the customer psychology. So two years feels like enough time for the customer to use the actual credit volume. So say hey, buy 50,000 credits now, whatever you don't need, you roll it over to next year and then you spend it then. Right? Easy. So if I rolled it over three or four years, am I significantly strengthening that argument? Not really. Like one year in my experience just seems to be enough.

for almost all customers of all sizes to sort of cover that argument. And then if we have a two year revenue recognition that actually is manageable for most CFOs to then sort of handle in terms of how they run their PNL. How is it determined how many credit customers buy at the beginning? Well, that of course depends what you're selling. Are you selling audio books or are you selling whatever it is that Eden AI is selling like all the technical outcomes? I think that you need to have a relatively simple calculator or estimation tool. I usually advise salespeople to not push too hard to say hey start them off a little low say you can always upgrade later this actually usually creates a lot of trust in the sale and sort of and if they use the product then they will build the momentum. So but some sort of yeah some sort of calculator or or or estimation tool is usually a good place to get started. And then I would also say that what you need to do is you need to have predetermined credit subscription granularity. This is basically a fancy way of saying you can sell 124 credits. There's like certain set amount like 100 200 300 thousand five thousand whatever the whatever the granularity is. sell them in specific like chunks of bundle sizes and not some arbitrary amount that people dream up right so you need to sort of round off to these numbers and because the numbers usually match dollars pretty like straightforward then that also means that the invoices come out in pretty clean numbers like $5,000 $25,000 $70,500 whatever the number is. They are used In the account on a FIFO principle so first and first out basically means that the oldest credit is spent first so they expire after two years but I always use the oldest first that actually creates a it's a really favorable version for the customer because it just means that I really need to have less than half of my expected volume before I ever are in risk of having credits rot or expire.

I would then allow all customers to upgrade subscription amounts freely which means that if I subscribe to 50,000 credits I would just say hey if like during the year if you want to go for 70k just go for 70k so I would not I would not charge them some sort of penalty price or overage price where the credits that they didn't commit to at the front of the year suddenly were more expensive. Why? Because one of the powers of the credit model is that it shortens sales cycles and it becomes easier to go in. If you start to work with overage penalty prices on extra credits, suddenly you're putting people in analysis paralysis at the beginning of the sale saying, oh, now I better estimate exactly how many credits I need, et cetera, cetera, as opposed to just buying something and getting started. So I'm a big proponent of just having like a non penalized like addition of credits. So this is also the answer to what happens if customers run out of credits during the year you need to have a price point for it and I usually say whatever you pay for your current credits that's where you're going to buy more credits at that's it. But the new credits the overage credits does not unlock additional discounts right? So if you had 50,000 credits you need to buy 20 more you buy them at the price of 50 and not at the maybe Better price of 70 right? So but you're saying hey, this is then what happens at renewal. So how is it determined how many credit customers by renewal? I would usually say. That where do we have it here? Force upgrade customers to subscribe to credit volume based on long-term spend. So I sometimes call this transposed pricing meaning that whatever I spent last year now becomes the volume that I am forced to spend next year. You can also do 80 % off. I've done this in vessel tracking in maritime for example. Hey, if you spent 10,000 credits last year tracking ships in the world, we're going to force you to upgrade to a minimum of 8

3,000 credits next year right so that means that we're always going to be put in a position where we don't have to negotiate the number with customers because we don't want to have to respend the cock the time the effort to just talk to customers all the time about how many credits they need we just want the Contract to solve it and the system to solve it so the hey, we're gonna measure how many used that's gonna be your new amount for your two And then we send you the invoice and then we go right? So whatever works for you there if it is like purely 100 % of past volume or 90 or 80 % of whatever it is then just go ahead and do that. You can always have like backdoors for CS if they say the customer like divested part of the business or just something that like material reduces their volume they can call you're gonna be human and make like a sensible solution to that. Yeah, so that's beginning that's renewal runs out during the year who in the customers organization can approve to buy more credits? I would say. This is basically where you need to sort of solve for risk on the customer side. So if you say to the customer, hey, anyone in your organization, any user that this touches can basically approve spend, then you're going to have a lot of procurement departments and P &L owners that are going to back away from you saying, hey, I don't trust all my users with this. Let's say I'm a hospital director of finance and suddenly you're giving all the doctors and nurses just like free range to just spend money without check. Big problem. So you want to understand the relationship between buyer and user when making this decision and depending on that relationship there are different sort of options available to you. By far the fastest expansion of your product and the largest sort of net revenue retention growth is if you can push more people to be able to approve SPAC.

However, in most enterprise settings, you need to centralize it so it's only the buyer or the P &L owner, whoever it is, that can approve the spend. But make it a point to really try to like thin slice this decision. Say, hey, can we move it like just another step closer to the user? Can we move it to middle management? Can we move it to some sort of like approver in this function? Can we give them the ability to engage in some of the work and then having it approved after the fact? there ways where we can really push forward the spend without having the user hit a paywall? That's really what we're trying to optimize for here. And then finally, can customers get the credits back if they churn? And the answer just straight up is no. They cannot get them back. So and because you're again going to have a financial liability and you're going to have to be regulated in a way that you don't want to be regulated if you need to pay it back. Whatever you say is like hey you need to just like keep buying credits and I would also say that. There is no subscription at zero like as soon as long as you keep buying from us the credits will remain but you can't just like put your your new subscription to zero and just eat out of like the store credits that you rolled over in the past years. You need to have some sort of minimal spend with us every year. That's also determined by the. By the granularity here. Yeah and every billing period the customer pays for a set amount of credits. That's usually how it would run. Okay. That's it. We have a few questions here from Maggie and Jeffrey. So I was on a roll you two guys so so I didn't sort of I didn't get to it while we're doing so you say Maggie here. Hey Oleg it sounds like if a current SAS company on one of the basic subscription models want to pivot to credit based a big product development effort is needed to make this happen. Measure credits would you ever advise against introducing the system if it hinders product innovation?

You you're going to have different options available to you like hey we could do product, we could build new features or we could build a credit system which is going to make you more money. Make the call. I would say that I have launched a lot of credit systems that have just been run in spreadsheets for the first like two, three, four, even six months. That can totally get done even with multi purchase systems. If you have single purchase systems, it's almost easier right then you have no excuse. So it's not harder to measure credits than it is to measure usage based pricing. It's just harder to communicate and keep track of so it's more of the billing aspect and the accounting aspect that really so takes. Yeah, that really makes credit systems sort of a little bit more complicated. But. But overall sometimes like in the initial case I talked about with the icon finder like they could have developed like a ton more features in the product and none of them would have five or six X the customer lifetime value like the credit system did right. So sometimes the credit system just is the killer feature that is just going to like unlock a ton of product value as opposed to like incrementally making something like a tiny bit better inside of the product. And I think you just need to sort of see hey, if you have a license or user space model that is working like pretty well and customers get in and they're paying you and you're growing and all these things. Maybe the credit model isn't like the right thing for you, but if you're having a lot of like pushback like a lot of initial sort of like. conflicts in the sale, a lot of different sort of seasonal variability, a lot of different sort of values of the product that you have a hard time pricing, then maybe a credit system is just like the solve that is just more elegantly going to create a better relationship with the customer. And then I think like it just trumps whatever you have in your roadmap elsewhere. Jeffrey, you're gonna get the last question here.

When you say the credits roll over to the next period, is that from year two to three? If yes, doesn't this create revenue recognition issues? So whenever I'm ending a period, I'm looking at whatever credits I have. And then all of those credits that are the same more than 2 years old at this period end they just get like deleted and then any credit that is less than 2 years old they are now put into the to the next period right so so that's it so the revenue recognition part so. I have had CFOs do both things so you can either say hey we're going to put a price on a credit and then we're going to measure the usage and then we're going to recognize the revenue as we go along. That is relatively straightforward. However whenever customers start to buy more. then the new credits also have a different price, which means that the blended price of credits can change. And now they basically need to run an inventory motion on your credits, FIFO principle, where some credits are worth, let's say, a dollar each, other credits are worth less, let's say 80 cents each. And because they are spent at a different rate, now the revenue gets recognized a little bit differently. Most CFOs just run like an average on an account like whatever they have paid and hasn't been recognized yet. That is the total.

Value of credits divided by number of credits and that's what we recognize whenever a credit is spent but some purists would argue that you actually need to identify the Original purchase price of each credit and then recognize that one whenever it is spent, right? So it is a little bit there are different ways where you can recognize this I have even had ones that just recognize it In the year that is bought and say hey, this is a right to use it is up to them to use it. So here we go. I have also have some that just say hey, we're going to recognize like 7080 % in year one based on the assumption that that's how much they're going to use. Then we're going to roll over the remaining 30 % to year two and recognize it then so much more blunt approach and I have seen each and every one of these methods used. Let's say live and pass ordered. So it's, I'm not an accountant, so I can't speak to the legality of this, I can just speak to my experience on these parts.

Everyone I appreciate it. I appreciate your questions. I appreciate you taking the time to show up on the on the other end of this. Here's a few resources for you so you will get an email shortly with this. You can go to community.pricing.com for follow up questions and also some of the earlier webinars to check that out and then on willingness to pay.com. You can find me or some of my colleagues and our resource section in general. When I send you the email a little bit later it will be from my personal email address. So if you have a comment or shout out or another question just feel free to write me directly. Yeah I read all of those responses but for now thank you so much for tuning in.