The Rise of Usage-Based Pricing with Patrick Malatack and Kyle Poyar

One of the things that I think is really great and core to why we embrace this at Twilio is when customers are getting value out of the product, they're using it every day. Then you're getting revenue from those customers. It's really around just making sure the business that you're building is aligned with the success of your customers. I need some traction. You need some traction. Let's get some traction.

Hey what's up innovators, entrepreneurs, visionaries and disruptors. This is your Traction Podcast host Lloyd Lobo. We're a community of over a hundred thousand people just like yourself on a mission to help you get the methods, the money and the madness to explode your business growth. Featuring stories and tactical advice straight from those who've done it before, like Shopify, Twilio, Asana, and many more, this episode is brought to you by Boast AI. Each year the US and Canadian governments give out billions of dollars in R and D tax credits and innovation incentives to fund businesses like yours. But the application process is cumbersome, prone to frustrating audits, and receiving the money can take up to 16 months. Boast AI gets you access to research and development tax credits and innovation funding opportunities without the headache and red tape. Join thousands of North American companies leveraging Boast AI software to maximize cashback. Check out B O A S T AI. This episode is also brought to you by Launch Academy, an international tech hub that provides mentorship, resources, network and the environment for entrepreneurs to launch, fund and grow their startups. Since 2012, Launch Academy has incubated over 6,000 entrepreneurs, of which 300 have grown their startups past Seed and Series A and have collectively raised over $1.2 billion in funding. To learn more about Launch Academy's programs, check out LaunchAcademy CA. Special thanks to our Podcast partner Content Allies From Podcast Production and Promotion content allies helps B2B companies build revenue generating podcast. We recommend them to any B2B company that's looking to launch or streamline their podcast production. Learn more at Contentally it's going to be phenomenal.

Session full of great knowledge bombs here and who better in the house than the two guys who've written the playbook? You got Patrick who was early early at Twilio where there were less than 50 people, they had less than 14 people on the product team and Patrick led product all the way through Twilio's post, IPO and 1800 people. And today he's investing in great companies, great founders, and most of your portfolio companies have been usage based and Then on the other end we have Kyle who has pretty much written all the great playbooks on usage based pricing. And the stats here from openview and Kyle is the end of subscription pricing near like Research shows that usage based pricing companies grow 38% faster and of all the IPOs over the last few years 7 out of 9 of them had the best net dollar retention were all usage based pricing models. And Kyle, he leads OpenView's growth team. He's responsible for advising portfolio executive teams on strategies to increase revenue growth and dominate their markets. And the team has helped the portfolio companies generate over 100 million in additional enterprise value in the last three years. So welcome Patrick and welcome Kyle.

Patrick, Usage based pricing feels like a really hot topic right now. I think Bessemer called it one of the go to marketplace for the new normal of SaaS. TechCrunch has been writing about it a lot. But if you kind of rewind to the early days at Twilio, the model was not tried and true at that point. What gave you guys conviction to go with usage based pricing and and how did that work out in the early days?

The thing that is most attractive about usage based pricing and why we had such conviction around it early on and particularly when it was a lot less proven was just that it was very customer aligned. The core tenet of usage based pricing is you get paid when the customer is using the product and getting value out of it. It really came less from a hey this is a super efficient go to market strategy and really just from a hey, seems like it would be a good idea for us to be aligned with the success of our customers. What was happening prior to that? It was a lot of contracts that were sold. I was at Microsoft before I joined Twilio. A lot of software projects never got off the ground. The license would get sold in the project may or may not have launched. I distinctly remember I was at Microsoft in 2008 when there was the economic downturn and all of a sudden we got a ton of cancellations for SharePoint and it turns out it's really easy to cancel your license when you haven't already deployed. One of the things that I think is really great and core to why we embrace this at was just that it's aligned with customer success. When customers are getting value out of the product, they're using it every day, then you're getting revenue from those customers. It's really around just making sure the business that you're building is aligned with the success of your customers.

For me When I think back to the normal way people think about SaaS products is it's the traditional heat based subscription model. And when you look at how products are being built right now, increasingly their benefits are either around automation for users, their API based business models, or their AI products streamline what would be manual processes. The value people are seeing isn't necessarily aligned with having more people log in for the software. It's on how they're logging or how they're using the products and seeing value as an organization. And I guess my research showed that public companies are growing faster. They have a usage based model. I think a lot of that's driven by the net dollar retention and their ability to grow as their customers are being successful with the products. But it's not always the fact, it's not a de facto fact that usage based companies should be growing faster. And I think you've seen in the early days the opposite might be true. How are you thinking about how quickly companies grow and how a usage based model impacts that?

Yeah, there's a quote from Einstein that I love, which is that the most powerful force in the universe is compound interest. So who am I to dispute Einstein when it comes to forces in the universe? But the way I think about usage based pricing is that it really is compound interest type pricing. What does that mean for you as someone that's building your business? It means that growing at a growth rate in the early days, let's say the growth rate is fairly fixed early days versus later days. It means that actually it's really hard to get your first million in revenue in subscription based pricing. You, you can sell a bunch of subscriptions, you can get out there and you can really push the numbers in the early days in usage based pricing, you really aren't getting paid until the customer's launched, successful and then they're integrated into their product and those products are themselves growing. And I'd say one of those things where you backload the growth in these businesses. The flip side is hard as that first million dollars of revenue is to go out and get getting an extra incremental million dollars of revenue at 100 million in revenue size is really effortlessly. The vast majority of that growth now is not a function of your distribution efforts. It's a function of the, frankly the distribution efforts of your customer in terms of whatever it is that they're building. Particularly for these API type businesses where you're being fully integrated into the products that your customers are building. Things like Twilio, things like aws, things like Stripe it makes it so that the growth rate much later on in it's much easier to maintain these sort of high growth rates later on in the company. I think that when you look at those numbers of publicly traded companies and how rapidly they're growing, I think that's why you see it in the early days. It's difficult to just to get to that first million. But once you get things going, Compound interest is able to do this on its own. And you really end up with these businesses that have three independent growth levers. You have the expansion of the existing usage of the product that you're using today. You have net new customers that you're bringing in, which is the traditional growth lever most companies are expecting. And then you have new SKUs that you add. Right. That you can add on new products and services and cross sell into those accounts you end up with. And each of those then has the same baseline growth rate that is the growth rate of your customers. It's one of the real powers of these usage based business models and I think in particular they are phenomenal as you've gotten to scale because you can continue to scale growing those businesses over time.

Yeah, I think that's key from an investor perspective. We look at net dollar retention being able to show that a company can expand their cohorts. But not all dollar of expansion revenue is created equal. It's very different if you're going and selling to a new bu, a new team or selling another product into your customers. With the usage based model, a portion of that at least is more organic growth. It's pre negotiated. It's just the customer is successful so they're using more of it and you're naturally sharing in that success. And it's not only faster expansion revenue because you have more of these opportunities to grow, but it's even more efficient expansion revenue.

It's really neat because in a lot of businesses when your bill to your vendor is getting too large, you're like oh wow, this bill is too large. I shouldn't be paying my vendors so much. It's actually the opposite, particularly in these horizontal developer platform where your infrastructure right at the end of the day it means that business is working really well. Customers are actually in a very different spot as their usage of your platform grows. They're happy because it means their business is being successful. And this is the initial point I made around aligning with that customer success. It's really powerful to be on the side of your customer success as opposed to being a vendor where you have Procurement, trying to get the best price and you're trying to drive down your bill over time as opposed to the world of usage based pricing where it's as it goes up. That's actually generally a good thing for these businesses as well.

I think a fear that a number of folks have is that usage based revenue isn't worth as much as subscription revenue because it's not committed, it's not locked in. But as I talk to investors and even from the open view investor standpoint, I think it's more valuable than committed revenue because it means there's not shelf. We're in the business. Every amount of revenue that you're generating from usage is actually revenue driven by customers being successful with the platform. You can't hide anything with a usage based model. And I think it's analogous to going from on prem SaaS to consumption based SaaS. There's not the commitment and you're not necessarily collecting as much cash up front, but you're building a more customer aligned business model and if you have a sticky product and customers are using it, you're setting a really strong foundation for future growth that the market will eventually reward. But I guess zooming out, where does this fall down? What businesses aren't right for usage based pricing or shouldn't be adopting this model?

I don't think it applies to everything. I think it's a really interesting question though to ask of a lot of businesses, particularly even existing businesses. We talked a little bit about infrastructure businesses and why it's so powerful. When we priced a more recent product from Twilio called Twilio Flex, we did it. It's a customer port tool and your agents typically you have a monthly agent fee but at Twilio we actually priced it per agent hour because it unlocked a whole bunch of new use cases which is if you had a bursty contact center all of a sudden there are a few hours a day where you need a lot of people manning that contact center to be able to do that in a way that makes sense. I think there are though places where the initial configuration and sell in is has a high fixed cost and making sure that you aren't underwater because you don't know that every customer is going to expand. And in fact I think one of the powers of the usage based business model is that you can have a generous free tier and you can see which customers are successful at the product and grow over time. But not every business looks like that. Some businesses do have a lot of upfront costs for the company to just get the customer started for those things, it might make sense to have an initial subscription fee then paired with a usage based component on top of that as they succeed with the product. And so I do think you need to look at the problem you're solving and where you're creating value for your customer and figure out how you meter appropriately. And it may in fact be a combo of some initial subscription with then usage that sits on top of that over time to map to the value that you're creating for your customer. I think the other thing too that I've seen folks do wrong in building out their pricing models is that I think a lot of times folks will look at where their costs sit and exclusively try and meter either a subscription component or a usage based component around where they're worried about their cost exposure. And while super important to have an understanding of what it costs to run any service that you're operating, I think that the more important thing is trying to figure out where you're creating value and making sure that you're metering on those units of value. The example I give is like you could have metered on a Twilio API request, but that doesn't actually create value for the downstream customer. What creates value is them getting a text message that reminds them to schedule their next appointment. Right? That's actually where value is being created. Making sure as you think about, even if there's a combination of subscription plus usage, that ultimately all of those are getting back to where the customer is deriving value and not where you're deriving cost and certainly need to be aware of that. Ultimately I think it is far better to meter on the value that you're creating for customers.

Yeah, I totally agree. And to me part of that is even ideally that metric shares in the success of your customers and the business outcomes they're trying to achieve. And I think like a classic example is HubSpot, which is one of those non developer centric companies that has started to monetize more and more based on customer usage. And for them they used to just have fixed packages like a small, medium and large and they realized their expansion was abysmal. They were at like a 75% net dollar retention and they decided to add this additional layer to their pricing about contact based pricing where as you generate more marketing contacts from all of the inbound marketing you're doing, you're going to pay more and that's going to get HubSpot aligned with helping you generate more contacts. And as you generate more contacts you're going to be able to grow your business and that's really aligned because your customers are trying to do the same thing that you want them to do and it doesn't feel like a tax or like you're penalizing them. And it's not necessarily just because you, your costs are structured in a certain way. It's actually based on the value that they're seeing and what they want to do. On this note, how do you advise companies on picking the right metric for usage based pricing and what does that process look like in your experience?

Yeah, it is super tricky. I think that there are a bunch of different metrics that I really drill the teams at Twilio to focus on. I think figuring out what that unit of value I think is a product by product specific exercise. I think understand all the products in the world that could be created and know what value you're creating for customers is very tricky. I think it is important to try and experiment particularly in the early phases of the product and sometimes you choose the wrong usage metric, then you realize that it doesn't work for this whole class of scenarios that you really wanted to be able to solve. All of a sudden you find out oh wow, this is super expensive or we make zero revenue here and we're doing a lot of work. I do think it's important to experiment a little bit in early days in terms of trying to figure out exactly what the right metering metric is. I think after you've arrived at what that metric is, I always focus folks to look at their business through what is that usage metric that you've arrived at? How are you getting more trials into that particular meter, whatever that is? How are you getting more and more folks started and is it easy to start? Do you have the appropriate free tier at whatever those usage threshold in the terms of HubSpot? The other advantage that you get from contact based pricing is when the business is small, you're not asking for very much money and when the business is large and you have a big audience, then all of a sudden you're asking for more money. And I think you'll see that play out in how you set up those trials. How are you able to get folks in the door getting the start of some small amount of usage and having a trial there? I also have folks look at new SKU usage. Is there cross sell opportunities within particular SKUs? At Twilio we saw for sure folks would come in, they'd start using Twilio messaging and that was the right land product. For most companies they needed it and wanted to talk to their customers that way. But then over time they would start using other products, they'd start building an IVR for site and all of a sudden we'd get voice usage and we get cross sell to other SKUs over time. And I think having an understanding of what that looks like in your business is super important because it might lead to an insight of great, let's lower the price point and how we think about metering for the land product because we know it opens the door on the demand generation side for a lot of other customer relationships that we can build trust with and have folks take a look at that. And then obviously expansion rate is super important on whatever that meter is that you've put in place and having an understanding of what that looks like. But just as important is dollar based and account based churn right when you're having things fall out, why are they falling out, what's going wrong and all those things together. If you have an understanding of sort of baseline expansion rate, new SKU use, ability to get new folks in the door, that will all lead to essentially a situation where you have the ability to have a pretty good understanding around what revenue ultimately is going to look like. Revenue to me is really just an output and I know a lot of folks like to focus on revenue, but for me revenue is just an output, not the core thing to be measuring early on.

Yeah, I think those are great call outs and for me a lot of the testing can be done in more of an interview setting, like qualitatively asking your customers around different metrics you're thinking about charging on and there's some element of if your customers don't buy into it in an interview or you can't simplify the message clear enough in a 30 minute conversation, it's going to be really hard for you to sell the product. So it's good practice before you even start to put your existing revenue at risk. From my point, just to double click on a couple of those things, predictability ends up being important for the metric. Knowing that with a usage based model, usage tends to grow over time and what someone's forecasted usage don't always hit where you could dramatically exceed it. But the customer needs some element of predictability to have some peace of mind. I think like with Snowflake, while you could say their pricing is hard to predict because it's based on compute and it's a little bit technical to figure out, they have sales engineers that will help them based on what the specific workload is like, they can pretty quickly and very accurately help the customer figure out. Okay, based on your workload, here's what you can expect your usage behavior is going to need. And we can give you 5%, 10% headroom in case you need to go beyond that. But creates some predictability which puts the customer at ease even though it's at the end of the day. A usage based revenue model and there's going to be some inherent unpredictability. I think the other thing I'd add is the feasibility element. Like you have to be able to actually measure it, monitor it, report on it, charge based on that. And that can be an operational headache. You can have systems that are giving you different insights about the metric or might be a metric that the customer has access to, but you can't measure yourself. And all of that just creates a ton of pain in realizing the benefits of a usage model on this point. I know there's just a lot of operational, financial and strategic challenges with adopting a usage based model. Maybe we'll start with one on my side. I guess it's the element of enterprises and enterprises potentially pushing back procurement, wanting control, predictability in their spend. It's hard for them to figure out how to budget for something if there's not a fixed cost. How have you handled that?

The points you made around both making it predictable and understandable for customers is key. I think. You know, one of the downsides, and certainly we had some challenges with this at Twilio, is you can get to a spot where your pricing is not particularly simple. You have tons of SKUs, all of them have variability. And you end up in these customer conversations where they're like, what's that going to cost me? And you're like, I don't know. And that's not a good spot to be in for your sales team. It's not a good spot for the customer to be in. And we try to invest a lot in understanding for various different use cases you might build out on top of Twilio what the cost of that would ultimately be. And equipping your team to be able to go in and say like, here's roughly what this is going to cost. If you start doing reminders, you will see that appointments in your business will drop, appointments that people don't show up to will drop by 20%. The value you're going to create is you're going to have a lot more people making it to their appointments and ultimately you're going to be in A spot where it will be a worthwhile investment. But we did calculators too, for our sales teams, just that they could plug some things in. And then I think the other thing is working with your team to be willing to say, data, you can share us about your business today so we can come up with an actual estimate. Certainly set of things about your customers and set of things about your business that would be useful for trying to be able to predict what this would look like if it was integrated into your product and not being shy about that. And in many companies, they won't share the direct data, but they'll give you some proxy for what that data looks like. And we'd have our sales engineering team sit down with them and be able to produce a model for them over time. I think one of the most powerful things, though, is coming up with a model that shows them the value that they create from adding whatever this particular component is to their business. And what that the business value that creates. Hey, all of a sudden a bunch of declined appointments aren't happening as a result of your Twilio. And typically you'll find that building something better and improving your processes in a particular way is actually just a lot more valuable to the business than what the ultimate cost of that Twilio bill in this case would have been. I think trying to find those opportunities and coming up with playbooks where you can talk from customer to customer. There's many businesses in the scheduling example that have something that is scheduled where there's a decent number of no shows at the scheduled events. Our sales team then had a playbook. They'd be able to go talk to a lot of different businesses, a lot of different folks, and be able to articulate the value of Twilio. That was a lot of ways independent of what the price of Twilio would be. Because the price of a penny for a text message relative to having a 20% reduction in no shows is like, not even comparable. Those are the ways that we tried to look at it is to get folks thinking less about, like the cost of the underlying product, but more about the benefit they would receive in the long term by having that the product is that they're adopting. Those are some of the Playbooks we would come up with. But I do think it's a challenge, right? And I think the best way to do it is to come up with these for this particular use case. This is what we see from other customers and then also really being aggressive about saying, hey, we want to be aligned with your Business success. We don't want to be taking your money if you're not successful. Why don't we try and figure out how we do a pilot here, get you guys successful. We think there's going to be a ton of value that's created here and that in the long run you're going to be a lot more focused on the value create side of this rather than the cost side of things and get that going and get them successful. And I think that's the magic of usage based billing, which is at the end of the day, if they don't like it, they can stop using it and it's not a giant commitment for them and the business overall.

I think those are great points and that helps alleviate a lot of the customer potential concerns. And then taking that example, so you might have a customer that starts with a proof of concept. Maybe they're spending a thousand bucks. Over time, they might spend millions of dollars a year. How do you think about the financial aspect of managing a business with usage based pricing? When there's that challenge around forecasting, all of a sudden rev rec is more complicated. The KPIs you're measuring as a business aren't just typical MRR ARR, but there's a lot more nuances. What challenges have you had to navigate there?

Yeah, we had a ton in the early Twilio days just trying to figure out what the right model was. Particularly as we were going at it at Twilio, there wasn't public companies that were also doing this. We were trying to figure it out a little bit on our own. When AWS was broken out of the overall Amazon bill and I think spring of 2015, that was for us a really great moment that validated a lot of the hard work we were trying to do to get this model just to be able to see, oh wow, this is how it grows with a publicly traded company. Our aspirations at the time, the thing I think first just keep in mind is that this is not a product problem, it's not a finance problem, it's not an accounting problem, it's not a sales problem. This is a whole company challenge that you all need to come together around to really embrace building and offering products to your customers in this way. For example, let's just take a few here. The predictability of these businesses. It means that the CFO needs a lot more data scientists, a lot more analysts to try. And without a paper with a signed commitment from a customer that they're going to spend X amount of dollars over Y amount of time frame they need to be modeling, they need to be looking at this as an analytics problem and coming up with a model in a range of outcomes that they think could be possible and trying to establish a budget within that range so that you can both aggressively invest in the business, not being too aggressive about your investments going forward. On the finance side, it's a lot more analysts, a lot more number crunching. It's not just looking at what do we see in salesforce, how do we think that's going to play out over time. On the sales side, a lot of the hard work your sales team is doing really isn't necessarily going to be recognized at, you know, contract close time. Because contract close is, hey, we got a pilot going, we got something very basic started with this particular customer. And these businesses grow at really robust rates. You need to think about that, right? You need to keep your sales team sales. Getting that pilot going is immensely important for your business. And even if it isn't a giant revenue win on day one, you need to make sure that you're compensating your reps, take into account what that account could be over time. And getting customers successful is a large part of what the sales exercise is here. And making sure that your reps are compensated on getting that pilot out the door, getting that initial usage based thing going. So you get to start the clock on that compound interest. And you want to make sure that you're compensating with them in a way that they're incentivized to go ahead and get customers started and get them successful. You also want to be thinking about how do I compensate reps on cross selling to a new sku, a new use case within the organization in a way that is quite different than how you would have done this in a more traditional subscription business. And a lot of the work is done not just to get the ink on a contract signed, it's really in getting those customers successful. And the type of reps that you hire actually even look quite a bit different. Right. They're ones that are really getting deep in with those customers that understand their business, that are consultative in their approach. And none of these things are a specific problem for one part of the organization or another. The whole organization needs to be working well together. Our finance team was spending a ton of time getting the product usage metrics out of the systems that our product teams were building. You had FP and A folks that were paired deeply with our product organization to help try and do forecasting and forecasting. Wasn't single account forecast in the way that it typically is in organizations through your sales team. It was a combination of that with a combination of just looking at the numbers and seeing how they grow over time. Trying to be predictive in terms of what that expansion is likely to look like over time.

Yeah, I think those are all great points. Just to add to it on the sales comp side, I think we're all used to compensating reps based on booking. So how much? What commitment did you get from the customer? But in a usage based model, the revenue might scale over time. You might and I actually be best to start with that pilot or proof of concept. And if you just compensate based on those upfront bookings, your reps are going to be encouraged to go for too big of a deal that might be larger than a customer really needs where they have a bunch of extra usage at the end of the period. And that has the challenge of slowing down a sales cycle. I think compensating more of a balance of based on consumption revenue as it's recognized plus the upfront commitment, just being mindful of doing what's in the best interest of the customer aligns the rep with the behavior that you want to drive. I think is better than taking the.

Standard approach just to really double click on. A lot of this alignment is around the usage based billing model aligns companies with their customers really well. And I think from that standpoint you then need to work backwards and say great. How do we make sure all the other parts of our organization are similarly aligned with customer success? That's why I love these types of businesses. It's why since I've come over to the VC side of things, that's where I've been spending time. And the type of companies that I've been investing in have this component because I think ultimately it's not just a really good business where you see these great growth rates in these publicly traded companies, but it's also these companies also have super high NPS scores. They have really great customer sat because they are aligned every part of the organization with the success of the customer. The sales team is aligned with getting usage going and getting their product whenever their integration is launched rather than just waiting, getting the signature and then walking out the door. All these things are really about just shifting your entire organization to just be very customer aligned. Which is I think great for building a long term defensible business.

It's almost like why do you need a renewal conversation if the customer is just paying as they're seeing value renewal is a non event. Customer will pay less if they're not successful. Then they'll pay more if they are. One thing that I've noticed and curious if you've seen this too is that with a lot of usage based businesses it becomes a whale and tail phenomenon where you're attracting a bunch of folks. A lot of people end up never paying a whole lot of money. Maybe they're spending tens, hundreds, low thousands of dollars a year, but then a subset of them grow rapidly and end up spending hundreds of thousands or millions of dollars. It's hard to necessarily predict going in who's going to see that rapid expansion versus who's not going to going to expand. And I guess wondering if you see that and if so how do you think about accommodating those kinds of patterns.

Certainly we saw that I'd say in the very early days at Twilio where you had a set of customers that got successful and they grew really large. I think one of the things that's important as you invest in the business is making sure one, the value capture that you're taking from each customer is right size to the size of the account and the opportunity. And then I think also looking at what made those customers successful, why did they grow the way that they grew and what are other customers that would similarly see success with this product and then focusing your distribution efforts on capturing more and more of those so that you don't find yourself in one tail exclusively, that you actually have pretty heavy middle because you're repeating the pattern you've seen with the successful whales and you're applying that to other businesses and other opportunities that are out there. I think the long tail though is fantastic part of these businesses. I think one of the things that is really important is sure you're not monetizing that long tail particularly well, but it also means that those customers are happy to just have you for some small use case within their team or their organization. They're trialing you out and those become future opportunities for the business. Think about particularly as these horizontal developer platforms that are core infrastructure. Your customers teach you where the market opportunity is. You're solving for a basic need, storage, communication, computer and customers are going to show you where the best opportunity for compute is. And these change over time. Initially what folks were doing with AWS is very different than how they're using GPU instances and what things are being used for machine learning workloads that change over time. I think customers will show you new opportunities particularly in these horizontal infrastructure businesses like aws, like Twilio, they will show you new opportunities for new types of workloads. And frequently that starts happening first in that long tail. Someone's innovating, someone's trying something different, they find success and one of those starts to grow. And then you realize, wow, there's a lot of other companies that actually have this challenge. There's a lot of other companies that are seeing these particular problems in the market. A good example of that was at Twilio. We saw two factor authentication become increasingly important use case. And certainly when we started out building out the messaging product, we weren't targeting for two factor authentication. Our early customers were Beluga and GroupMe and folks that were doing group messaging and two way communication. And then we started to see companies like Box that were launching things like two factor Authentication. We're like, oh wow, this is the need for everyone out there that has SaaS offering. And so you then go and find great, who else has that need? Oh great, they already have an account, they've already done some testing with us. Great. Now let's go ahead and engage with our sales team and figure out how we can make them successful on the platform. A lot of what that sales effort is is not net new knocking on doors and it's actually just going back to that tail and seeing who amongst that long tail really should be a whale and figuring out what's blocking them. Is it someone on the engineering side that says no, we can't use Twilio? Is it someone in legal that says no, our data needs to stay on prem or our data needs to stay in this particular region better understanding what those use cases are and then working together to get them unblocked.

Now that you have the investor hat on, what are some of the ways that you evaluate usage based companies? What are the things that you're looking for or that you think they should be tracking?

I think though just at its core what we're doing at Matrix, we primarily are doing seed and series A investing. This is very early stage and I think it's at its core it's a lot harder for someone to spend their time using your product than it is often to get revenue. I always tell folks, figure out first how we get folks to engage with us to start driving usage on our platform and solve a hard problem for customers. And over time I feel confident if the companies that we're investing in are solving hard problems for customers, are willing to spend time and use the product that you can monetize that over time, I really focus early on, which is just, can you get folks using the product? Rather than can you get a I will use it. And here's a 50k LOI. I'm a lot more interested in how companies are spending their time and where they're voting their most precious resource, which is their time, not their dollars these days, rather than just what contracts you can get signed early on. And as I think about it, that's the stuff I look for really early on. It's not a question of how many accounts do you have. It's do you have a few accounts that are. That absolutely love the product and are using it quite a bit. Those are sort of the early indicators of success. And from that base, you can go take that story of a few accounts that are wildly successful and using the product a ton. You can figure out, all right, great. How do we maybe modify our pricing to drive more revenue? How do we reach more of these types of customers over time? But I think that's the core I'm looking for is just wildly happy customers that are using the product or service that you're offering in the early. And I think revenue and some of the other stuff that investors typically look for, I think those can happen over time and more than happy to take a bet on a founder and work with them on how you figure out those other elements of the business over time. If you've got wildly successful customers that are voting with their most precious resource, their time, and actually using the product.

There'S a bunch of questions. A number of them are actually about blending subscription or recurring revenue with usage revenue. I think there's a couple of flavors of the question. One is whether to have some hybrid model where there's a base subscription fee with usage on top of that. And another is around this idea of when do you flip the switch from usage to subscription revenue and should you try to move deals into more predictable recurring revenue streams? We'd love to hear your thoughts on that.

I think. Let's take the first question, which is around subscription versus usage and when to choose each. I think to me it comes down to a, you want to lower barriers of entry for products, particularly ones that have a ubiquitous need. I think this is a little bit different. There's, you know, products that are ubiquitous in need, I'd say communications, payments that are quite ubiquitous. And there's a set of products where there's a small number of customers. And I think for those products, you know exactly who the accounts are, you can reach out directly and an outbound motion works fine. And I think starting with a subscription model is okay if it is a less ubiquitous model. But I think for products and services that are ubiquitous in nature, I honestly think usage based is the right approach to go. And a lot of what you're doing is lowering the barriers of entry, where you have a free tier where folks can just get started. And if it's a product that's applicable to almost everyone, there's really no downside in you doing that and seeing who gets successful with your product and who doesn't. Worst case scenario, people aren't getting successful and you go call them up and say, hey, why didn't you use? Why did you churn? Why are you stuck at this rate? And you get to learn something as you're building out that business. I sort of think about it from that lens of is this a ubiquitous need or are there just a nameless of customers that have this problem? We can call all of them up and we know exactly what the value is to each of these customers. I think for that, I think actually subscriptions are a plenty fine way to go about acquiring those customers. You understand what value is, you're able to price effectively and you're able to go outbound and offer them a good price per value opportunity. I think on the usage side, when it's a ubiquitous use case, you really don't want to, in my opinion, have any barriers to using the product or entry. And I think sometimes folks throw an initial subscription on, which is, hey, I want to get some dollars out of my long tail. I think it's a lot more important that you get your long tail successful and spending big dollars than trying to monetize the long tail. I would highly encourage you, if it is a ubiquitous need, not to worry about the revenue leakage in that tail, instead to worry about how do I make that tail successful. Whales and anything that is a barrier to doing that, I think is something that you want to be figuring out and solving for the business. Sometimes I think a artificial subscription fee might trick you into thinking you're actually getting successful. And really you're not. You're not succeeding in making customers successful, you're just succeeding in monetizing your tail a little bit better. Which I think ultimately isn't what generates really big business over time.

Yeah, I've noticed similar things that it speaks to this like whale and tail phenomenon where generally those small paying customers are not actually generating the majority of your revenue. You're trying to move them onto some subscription with a Minimum, that's actually not going to move the needle on revenue all that much. It's better to open up access to your product, allow more people in, and then really work on product engagement. Customer success get those tail of users successful knowing that they'll expand pretty dramatically in time and so trying to be punitive and add subscription fees for them is a short term way of optimizing for revenue versus looking at the business over the long term.

Just practically like your sales team trying to sell into someone that has a trial going on. Maybe for a small use case in their organization. It's way easier for a sales team to sell into that type of a customer than it is for them to sell net new sell folks with some starting price is a barrier focus on getting them successful even if it's for a small use case. Driving that success will drive trust in the nature of the relationship between you and that particular company and then it will be much easier for your sales team thereafter to help expand from there. I do think it's important though that it does require you to be aligned with your investors on this because as I said, it's compound interest. And it's really great when the numbers are big because you're able to grow them at a very healthy clip. And you see this when you look at the AWS line item. You see this when you look at Twilio and you see this when you look at Snowflake. But early on, that same growth rate actually is applying early on where you would expect it to be super high. Unlike the way this used to happen where you would have these super high growth rates and then they would tail off because the majority of the growth was generated by your distribution effort. This is a little bit different. You need a set of investors that are patient that understand the model that are willing to say great, I'd much rather have a high growth rate at 100 million in revenue than a high growth rate at a million in revenue. And that requires patient investors and your management team to be patient too.

Yeah, I think that's true of most bottom up or product led businesses. That phenomenon. There were a few questions around metering and billing. So the usage based product does require more probably in the MVP around thoughtful usage metering, sharing analytics around usage behavior in app, making sure that there's meter billing in place. What's part of the MVP for that for a usage based product and what's the investment that folks should be looking at?

It is not trivial. Usage based metering is quite difficult. We had a pretty sizable team at Twilio that was exclusively focused on this. It is not a simple investment. The way I tell folks, particularly in the infrastructure space to think about this is if you're building infrastructure that others are going to depend on, like Snowflake, like Aws, like Twilio, like Stripe, probably worth the effort to over invest in metering and measuring your services. Not just for the ability to bill, but for the ability to understand what's happening, to understand if there's any downtime. Because a big part of what you're selling in these products is frankly trust. Right? They're outsourcing. You become one of their microservices when you build these products and so you're really part of their product. I would over invest in how you meter your product both from better understanding uptime and availability of the services that you're you're offering and from also metering those products effectively. And I think it just takes more time. I think the era of hey, we got our MVP out in two months and we're solving a really hard problem for customers. I don't think that era exists anymore that you know, two months when we built this mobile app and now it's blowing up. A lot of that lower hanging fruit has been picked. I think it does take to get to that initial MVP where you're really able to prove it out and solve a real hard problem for customer. I think increasingly we look at our seeds at Matrix. We expect, hey, it's going to take 18 months, sometimes two years to really nail that MVP. And certainly that's what we've seen play out. And I think there's risk in trying to get that out there. When you haven't actually delivered on the core value prop of the product. We tell folks, take the time and do it right and then the other thing is right, the thing I really care about is usage in the early days, not even the revenue. If your ability to meter at fractions of a penny and drive revenue dollars from that is tricky. And figure out how I do taxes for that. There's a lot of complicated things you need to do that correctly. Great, let's just put them on a fixed fee. If they go over, then the customer gets the benefit of it for now. And let's get them using the product because the thing I really care about is how they vote with their most precious resource. It's their time. So if they integrated with that API and they're using it every day, that to me matters a lot more than the revenue. If you're going to skimp on anything. We can figure out revenue collection later, figure out where the revenue leakage is in our products as long as we're getting customers successful and they're happy using the product.

One thing I'd add is that a lot of the large at scale companies with usage based billing have pretty sizable billing teams and think about the customer's usage and billing experience as a core product function. But there are more and more startups that are helping companies address this in a more turnkey way. Chargify and chargebee have both invested substantially in their usage based billing capabilities and the nice thing is that there's now more off the shelf tooling than there used to be. We got a few questions around investor pushback and I thought that we've done a good job of getting people not as concerned about this. But folks have said in investor discussions the investors only seem to care about mrr. How do you manage those conversations and how do you make this story resonate with other investors who are not don't think this the way we do.

I think that a lot of that comes from folks just pattern matching on what was successful in the last generation of SaaS, which is exactly the metrics that many are focused on now. And I would just really encourage you to share with your investors the quarterly statements that come out of companies like Twilio, companies that have embraced this model and really have them take a look at how those companies are doing it and what their growth rates look like at very large numbers. Twilio and AWS have been able to have huge growth rates at very large numbers because a lot of that work on the growth side is done by their customers. Right? It's the success of their customers and there's customer businesses that are growing. For me, that's where I would start is just showing them public companies that are executing on this isn't some secret sauce anymore. I think in 2015 at Twilio we were struggling because we didn't have a lot of public comps to point at. We thought a lot about different t shirt sizing of the product to try and map to what investment world knew. But I don't think that's the case anymore. I think there are a lot of successful public companies that you can point at that have really embraced this model and you'll see a whole bunch of things. It means you have a super efficient cost of sale because that tail right when your sales team is going and selling, they're selling into the tail. They're not selling into net new or they're cross selling into an existing account. It's how you use your sales team that changes ultimately and you can make businesses that are quite efficient from a distribution standpoint. So that would be my advice is I'd encourage you to just share that with your investors and really internalize what those companies are doing and how it might be applicable to your business model.

One Another question is around free products and how they fit into usage based models. Is it better to have small limits or limited features for a customer? Have a trial period on usage based models? Just a few words on the best way to layer in a free plan into these models.

Yeah, we certainly started with credits at Twilio. The path that we went down. Credits are tricky because it's another level of complexity in your billing system that you need to solve for. Because I need to track how many credits you had. Even if you're giving away free credits. I think ultimately the model that I like best is just the generous free tier. And to not try and meter and put pricing on too many different knobs of the product, it's just let them use a free tier. If they're trying it out, it should be free to try it out. For some products that means like great for the first five users, this is going to be a free service. For some it's the first X number of transactions, this is going to be a free service. But rather than limit what the product can do, I think it's way better just to limit them to meter on the number of transactions that they're doing within the product. Because what you really want as part of that trial offering is you want them to understand what the boundaries of that product can do. You want to be able to understand, oh, I could do this and I could do this and I could do that. And you want folks to be able to explore all the capabilities your product has to offer. And if they're not launching with it, you shouldn't be trying to make money off of them. You just want to encourage their exploration of what your product offering is. I'm very much in favor of a free tier. In fact, I really like a free tier in perpetuity. The same way that AWS really should have pioneered this where it's like if your usage is over X per month, then it's great you continue to have free usage of the service offering because it then stays top of mind. Particularly for developer platforms. A lot of developers have ideas and they play around with them at home. And like my first Twilio app for Example was I built like a doorbell buzzer for my apartment building in Seattle when I was working at Microsoft. That was actually how I found out about Twilio. And it was my friends would I needed a Seattle area code and I bought a Twilio number in Seattle. And then I had a mechanism for them to basically buzz themselves up to my apartment building. And if that wasn't accessible and able to just be there and operate for little or nothing for me, then Twilio wouldn't have been top of mind. And then later on as you encounter problems where oh wow, communications could actually help with this business use case, it was top of mind. I think generous free tiers are quite useful and as I said, you want to make sure that when customers launch, in fact when customers launch they want to be paying you too because they want to be able to call you up. They want to have a support agreement in place. They want to know that you're going to answer the phone and that you're motivated to help them be successful if there's ever a problem. And yes, I'm very much in favor of a generous free tier that lets them use all the parts of the product because it's part of your sales process.

I agree with that. I think you and I agree on a lot of stuff. When you even see that in models like Slack which charges based on the user, they'll have a generous free tier where you can do a lot if you can invite as many people as you want into Slack. And Slack goes from this place where teams casually interact to a place where your mission critical business information lives. That's when you run into the usage based paywall which is how many, what volume of messages they're going to store for you that you can search through and have access to. That's a great specific compelling event that lets the customer know okay, now it's time to convert. And they're able to do that while still offering a pretty generous free experience. I think that you can make the customer feel like free is great, has a high NPS without hurting your free to paid conversion rate. When you have these models that are like compound interest where the customer starts small, it can be difficult to measure things like CAC or LTV to cac. And how do you think about CAC specifically for a usage based model and what healthy is?

I think in the I separated between phase of the company in the early days, I just don't think you're really going to have a handle on that. Just realistically, so much of the Growth comes two years out, three years out, and when the company's been around for the product's been launched for eight months, it's really hard to have a handle on what that looks like. I think a lot of this is just a time horizon you're measuring so early on it's are customers spending time, are they wildly happy and successful with the product? And really focusing on that in the early days, I think later on what you'll find is that like cact, LTV is super efficient if your expansion rates hold up. That's why trying to measure what the expansion rate is early on is I think the most important thing. Because that ltv, if you have compound interest, the good news about it is from an LTV standpoint, it generates a ton of value for you later on. And so I really care about what that, what we think that expansion rate looks like for most customers. And then you'll find that actually you can afford a lot of dollars on the CAC side because LTV becomes very large at the two year mark, at the two and a half year mark, at the three year mark. If you have low churn and you have a high expansion calls for itself. So that's why I focus on those early on.

For me also when you're looking at usage based models, product and engineering might also be part of your cac, that's just something for folks to consider is that your product efforts are meant to help drive more usage, more use cases, customers being successful. There's a new set of metrics where you probably want to think beyond CAC towards things that are more relevant for the business. Patrick, for folks who want to follow you or learn more, where should they go?

If you ever are interested in chatting, my email is patrickatrixpartners.com you can also follow me on Twitter. Just PatrickMalletakMalatak is the best place to get in touch with me. And if folks that are embracing usage based billing, I'd love to chat with them. And companies that are helping drive the usage based economy I think are really interesting businesses to invest in and please send them my way and really happy to be able to do this with you and it was a lot of fun.

Yeah, likewise. And then for folks interested in more, I'm posting about usage based pricing pretty regularly on LinkedIn. You can find me at Kyle Poyer and we've got usage based pricing playbook at OpenView that I can share. All right, thanks everyone.

Thank you for listening and we hope you enjoyed this week's episode of the Traction Podcast. If you enjoyed the show, please leave us a five star review and you can find more information and all the resources mentioned in today's episode at Boast AI. That's B o a S t AI Blog.