Unlocking Growth in the Internet Economy: a Perspective from Stripe Head of Invoicing, Suzanne Xie

Who here works at a startup? And who here is familiar with Stripe? Great. I'm keeping my hands up for both because Stripe still feels like a startup to me. Uh, talk— seems like we're talking to the right audience here. So if you're familiar with Stripe, which it looks like most of you are, you're probably aware that Stripe has really expanded its product suite recently. And we're specifically building for SaaS companies, and hundreds of thousands of them are using our different products, everything from billing and invoicing to revenue recognition, tax, card issuing, and even verifying identity, and obviously payments. But before Stripe, I was also a founder, and most recently with a SaaS company. So I'm really excited to be here today to share more about my journey and what I've learned in the last year at Stripe. So after a decade of my own startup experience and the last year at Stripe, I wanted to talk about 3 main things today. First, overcoming the pain points of growing a SaaS business, but specifically with your financial stack. Second, some great trends that we're seeing in the market today. And third, ways to unlock growth in the internet economy, if the talk title didn't give it away. So it's 2021. It's been a crazy year. And while tech has enabled us to do so many things that weren't possible 5 years ago, SaaS is still really difficult to automate. It's hard to build a SaaS business. Why is this? And the simple— and it's a simple reason. We see it all the time. It's your financial stack. SaaS is notoriously complex to build. Things like setting up billing systems, getting enterprise sales up and running, automating your revenue. These are all very difficult pain points as you're scaling company. So let me start by sharing a little bit more about my own journey as running— growing my startup, and also hopefully you'll see that there's some similarities with either what you're experiencing today or what you may experience as you grow. So more than 10 years ago now, I started my first startup. That's me on the left with our startup corgi. Her name is Ro. She still likes to sit on people's laps when they're working. It's not very efficient. And on the left is where it all started, with a purple panda. Her name was Pandora, and she was the main hero and inspiration for us to build Lightwell Software, which Twitter acquired in 2019. So we started off by building interactive story apps for kids, and they were all around Pandora's epic adventure. We had an entire design, engineering, art team building these apps from the ground up. We even had voiceover artists— voiceover actors, and that was actually just us and the engineering team doing the voices. But it took months for us to build these apps from scratch, and pretty soon we realized we needed to invest in our own internal tooling to make this app development faster. We ended up building a tool to create interactive apps without code so the design team could create animations and interactions, and export them directly to native iOS code that our engineers could use. The first app took us 9 months to build. And pretty soon, once we had developed the tool and launched our last app, it took 1 week of engineering time. That was a pretty big aha moment for us. We realized we should just invest in this software and make it easier for other mobile app developers to build apps. And this trend towards no-code tooling and faster production maps really well to what we're seeing in setting up SaaS companies. But taking a step back, even though it was easier for us to creatively develop these apps, building our business and financial stack still wasn't. Let's take a look at why that was. So we have a lot of startups in the audience. And as a startup, you're focused on so many things. You're talking to users. You're raising money. You're selling to customers. You're building great products. And the last thing you want to spend time on is setting up your financial stack. So when we started selling Lightwell software, first it was hard to even set up a billing system where we could accept simple monthly subscriptions on our website. Took a lot of engineering time. Then we started to sell to some one-off enterprise customers, which is great for the business, but it required a ton of manual work. Things like custom contracts, manually invoicing, following up on payments manually. And ultimately, it led to what we call a patchwork billing solution, which some of you probably have as well. And accounting for revenue as you grow doesn't get any easier. It's really easy with a few customers, but as soon as you have hundreds, it becomes much more difficult. So we were not a favorite client of our accountants. And this squiggly line really only shows the operational side. Of your financial stack. There's also things like managing churn for completely solvable reasons like credit cards expiring. So when scaling our own software, setting up the commercial infrastructure was really the hardest part. And I really did think I was crazy. I thought this— it couldn't be this hard and it shouldn't be. And it's been sort of comforting being at Stripe and seeing that this is pretty common. We see the same things time and again from founders. The complexity of running a business actually compounds as you grow. And your financial stack sort of looks like this line. And many of you are bogged down with your infrastructure, which takes away from your core product development. So payments. Something as simple as payments, like accepting and booking revenue, it still remains a hard challenge for internet businesses to solve. We see this every day at Stripe, and it's what we like to solve for. If we look at B2C or consumer payments, it's gotten a lot better. It's pretty easy once you set up a web checkout. It's 2 steps, you get paid. But as soon as you look at B2B payments, where you have 2 sides of the equation with 2 different workflows, 2 different processes, a ton of different tools, it becomes a very different set of problems. So this gap between B2B payments and B2C payments are becoming increasingly stark. Making money really shouldn't be this hard. And then if we look at SaaS businesses that are smaller, as you're getting started, self-built billing and invoicing systems tend to be incredibly hard to scale. We talked a little bit about this already, but if you're starting off with regular subscription billing and then you're switching to usage-based billing, If you have an inflexible system, it's hard to make that change. Or if we'd mentioned collecting unpaid invoices, that's very easy with one customer, but much harder with hundreds. So if we look at large legacy enterprise companies, they tend to have entire teams that are just focused on collecting money. They'll have an AR, or accounts receivable, team that's focused on generating and sending invoices and then collecting on that payment. You have accounts payable teams or AP teams that are matching those payments to invoices. So for all startups here, likely you don't have either of those teams. You probably have a co-founder or a biz ops person who's doing all of those jobs at once. But as soon as you start to scale, this can cost millions of dollars a year. Many businesses have even told us they spend upwards of 200 to 300 hours a month just making their books compliant with accrual-based accounting rules. You could say it's a cruel manual process. A few founders have recently even shared their spreadsheet of shame with us, which has the manual tabs, imports, and rules in Excel spreadsheets that manage their revenue. And by the way, if anyone out here also has a spreadsheet of shame, Come talk to us. We really do want to help. So now that I've talked about all the pain points and difficulties, let's talk about some exciting trends that we're seeing in the market that are promising for internet businesses. So SaaS companies, all of you, are leading the way in making it faster and easier to grow a business than ever. Now, it's no surprise the internet economy has exploded in the last year. Or the last few decades, but also the last year. E-commerce spending in the US alone is expected to double from 2019 to 2023. That's $600 billion to $1.2 trillion. And if we look at these logos, which most of our— most of us are very familiar with, clearly there's an abundance of new tools and technology to enable businesses that didn't even exist 5 years ago. Our research has suggested that the cost of software tools needed to run these businesses on the backend has completely collapsed in the last decade. That's gone from roughly $4 million in 2011, which is insane, to $4,000 today, which still sounds a little expensive. And at Stripe, we have access to a lot of data, and we've seen hundreds of thousands of SaaS companies scale on our platform. And we've also seen in the last decade that it's cheaper and faster for them to build as well. Now, I've been keeping an eye on 3 trends that help explain why this is possible. First, no surprise, my favorite, the explosion of low and no-code tools. You can now outsource pretty much every business function you have. You've got Shopify for e-commerce operations, Wix for no-code website building, Xero for automating accounting. Webflow for web development, and Stripe for automating revenue and financial operations. Things that used to take months and tons of manual work can now be done so much more quickly. And this is all possible thanks to the increasing power and sophistication of no-code software. Basically, things are easy and cheap to build. Second, we're at SaaStr, so we all know this, but we're moving away from license-based software to subscription and usage-based business models. Which is great for customers and for companies. Not to date myself, but I remember buying Adobe software on CDs in school, and it's much easier now with the click of a button. In fact, at Stripe, we've seen that subscription businesses have grown 50% in just the last year, and the overall subscription economy has grown 6 times in the last 9 years. And these businesses are reporting growing up to 8 times faster than traditional businesses. So it looks like we're all in the right business. Third, many of these software businesses are evolving into platforms, which enables access to software that's even more integrated and easy to use. Software companies have woken up to the fact that they can provide a more valuable service and higher valuations by serving more horizontal use cases with vertical functionality, like payments and financial services. Substack, as an example, started out offering subscription management services for newsletter writers, but now they've expanded into podcast monetization and revenue sharing. Lightspeed started out offering point of sale, but they've expanded into enhanced online and in-store merchant capabilities. This all makes it easier for you as a startup to find an integrated all-in-one solution that just works. And it's also hopefully really inspiring to see that SaaS is taking on a new shape. Ultimately, business functions are being commoditized and made available as automated services. You can build on APIs to run many more specialist business and revenue operations entirely online. You can deploy subscriptions as a service, tax collection as a service, billing as a service, fraud prevention as a service. You get a service, you get a service, you get a service, everything is a service now. And all this to say, it's never been easier to build a product in business than it is today. So I talked a little bit before about how we're going to go through some tips and principles on how to build great SaaS companies, how to take these products to market. First, it's really about designing and deploying that killer product. It requires a rapid iterative loop with the right customers. We're going to talk a little bit more about that in a little bit. Second is helping to run your business as it grows. You have to digitize processes from the start. Now, this might sound counterintuitive for those of you who heard the startup advice, like, do things that don't scale, do manual processes. But what they mean to say is don't— is to do things that don't scale for your core business competency. When it comes to things that aren't your core, you should digitize. It's like trying to build your own servers instead of using AWS. These two principles have guided how we build products at Stripe, as well as my own experience of company building. So first, let's start with software design. You have a killer idea and you want to ship an amazing product. First, you have to be incredibly, incredibly specific with who your customers are. At Lightwell, we were our own customers. We built for interactive story app developers and we solved a really hard problem for them, which we'd already solved for ourselves. And even though we were our own users, we still talked to what we called our canonical app developers on a daily basis. That meant calling them up in the morning, showing them a new feature, fixing a bug, and getting their feedback. Y Combinator is famous for telling their startups It's better to make a few people really happy than to make a lot of people semi-happy. So my guidance to you would be really similar. Find the most discerning customers you can and build to their spec. Start with a product that's really useful for a few users. Then build a relationship with them and really understand and develop a mental model of what they want and why. Then iterate intensively with them, with that small number of discerning users, almost like you're a professional services firm building to a spec. And if you don't know specific people in your target audience, I'd suggest going and trying to find them and talking to them today. Second, so you know who your users are, go deep to understand your customers' and users' problems. You should be asking and re-asking yourself, What are your users really trying to do here? It's like the Henry Ford saying that we all are aware of. If I'd asked people what they wanted, they would have said faster horses. So a recent example from Stripe Atlas. We have a form on our website where we actually ask users, what should Stripe do next? And we get a lot of engagement. So a lot of customers started asking for help in getting started in other countries that Stripe wasn't in yet. So we thought this sounded like we should add payment methods in other countries. We were wrong. We clearly wanted other payments methods in other countries, but what they really meant to say was that they wanted Stripe's infrastructure. They wanted a way to set up their business from other countries. So we launched Atlas to do just this. We let founders incorporate and start taking payments in days, helping them launch thousands of new businesses from 140 new countries. So here, by digging in with questions, we got to the real problem that they were trying to solve, and we came up with a somewhat unexpected solution. Third, build a product that aligns you with your customers. So what does this mean? You know who your customers are, you know what they need. You want to design the best product that is good for your customer and for you. So we saw this with Stripe Invoicing recently. So cash flow is super important for SaaS businesses, right? Well, traditionally, invoices have 30-day due dates, but we wanted to help our users get paid faster and improve their cash flow. The— here, the more the business collects, the better it is for them and for Stripe. So we added a feature where they could mark invoices as due today. It's a very small, simple improvement, and even we were surprised by the results. In a few weeks, we had tens of thousands of businesses using this feature to mark 50,000 invoices as paid as due today, increasing how quickly they were able to collect. So going back to the model for the example that we just talked about in building great products, we designed Due Today, a simple feature for SaaS businesses. We knew they wanted better cash flow and predictability and that 30-day due dates could be improved. We built a simple feature that aligned us with our users on top of a product in order to improve their cash flow. And while it's still early, we've seen tons of good adoption and most importantly user satisfaction. All right. Let's shift over to running your business. And the first thing that I'm sure all of us want to talk about is costs. So I talked a little bit about— before about how it's still too difficult, but really too expensive for founders to get their financial stack up and running. In fact, one founder had shared their old models with us, and we ran the numbers, and the rough math came out to $0.10 on the dollar of overhead on just manual billing, manual collections, revenue recognition, and analytics. This founder had what they called AR days for when their accounts receivable line grew too large, and they called on the entire team that's the product, the sales, the support team, to actually pick up the phone and dial for dollars in order to collect on these aged receivables and invoices. Can you imagine? That's a full team dedicated to collecting on unpaid invoices when it could have been fully automated by software. Another example is subscription-based tools, which should be getting better all the time. So at Stripe, we use machine learning across our network to make our products more effective with no work required from our users. Things like smart retries or Radar fraud protection. Software is and should be getting better all of the time. So, last year at Stripe, we deployed our core production API thousands of times. That's our core API to manage payments and money movement. And here, without any work from our customers, the platform was getting better all the time. So, second, and actually our favorite thing to talk about, revenue. Software is making it easier than ever to experiment with pricing and packaging. So last week, I spoke with a founder who is building software to help companies scale their support teams. It's a new SaaS business, but they had some pretty big customers already. They're using Stripe Invoicing in the dashboard to send invoices, and they were adamant about not wiring up or writing any code because they wanted to understand every single conversation and get feedback on that pricing model. He was actually such a fan of price testing himself that when I asked about their billing system multiple times or plans for a billing system, he just said, "I am the billing engine." I think he really did believe it. So a good pricing strategy does a few things. First, it aligns your product with your users' needs. Really similar to aligning your product like we talked about before. How do they derive value from the product? This is really where usage-based pricing is important to test. Second, it optimizes value capture by reflecting your users' willingness to pay. So are you charging them for the things that they want to pay for? And third, it simplifies your first time to sale. Things like free trials and discounted annual plans. But ultimately, no matter what pricing model or plan you choose, testing is key. Third, let's talk about more revenue, but specifically optimizing what's recurring. So you have to understand your key stats. These might be MRR, ARR, ARPU, churn, or LTV. And like I said before, it's so expensive to acquire these new customers that once you have them, you have to do everything you possibly can to keep them. And price testing can really help with this. For example, we're seeing businesses increase renewals by offering discounted annual plans or the ability to pause subscriptions instead of cancel them. It's also worth looking at ways to eliminate involuntary churn. So, for instance, at Stripe, we automatically update expired or replaced cards, and we have a smart retries feature, like I mentioned before, which uses machine learning to retry cards at the exact time that they're most likely to succeed. Twilio is an example. They deployed this recently and increased their authorization rates by a full 2%. All of this to say, it's another argument for outsourcing your financial operations, and providers like Stripe can help afford to make those optimizations for your recurring revenue much more likely and likely things that you wouldn't invest in yourself. Fourth and last tip, plan for the regulatory implications of your growth. Now, this is probably the gnarliest one to consider as you scale. Expanding to different countries and markets can have huge compliance implications. I'm sure you've heard the saying, "The internet may be borderless, but money is not." And the cost of compliance is increasing year on year, and many businesses are struggling to keep up. So a few examples that come to mind that we think about: in Italy, there is an e-invoicing regulation where business-to-business invoices have to be sent to the Italian government in order for them to be compliant. If you're selling SaaS, SaaS is taxed in 138 countries, and you have to collect and file in every country. If you're doing business in Europe, you have to make sure your payments are compliant with SCA, or strong customer authentication. These are just 3 small examples. There's a lot more. So if you think about it, it's a little bonkers that as a startup you have to think about all of these things. So really going back to my first tip, think about the ways that you can use software to manage compliance for you. So back to you as a startup, what does this mean? Just this: your time is valuable. Stay focused on where your competitive advantage is, your killer product. Leverage software to grow, try out new pricing models, optimize your revenue, and And keep in check the things that aren't core to your business. So we talked a lot— talked about a lot today. And to wrap up, we have some big trends in software for SaaS, building killer products by staying close to your customers, and leveraging software in order to scale your SaaS business. And the key takeaway that hopefully resonates with all of you in the audience, because you're driving this change with your SaaS businesses, is that it's never been easier to use software to set up, scale, and run your business. So last time we were at SaaStr, this number was 3%. That was in 2019. Now, 5 to 8% of global GDP is currently being run online. It's growing fast, but it's still so small, and there's so much opportunity ahead. And it's really exciting to think that software is still in its early days. Think about how many trillion-dollar businesses are still yet to be built to take advantage of this. Imagine what the economy will look like when that number is 50%. SaaS, and all of you building SaaS in the audience, will get us there. So use it in your business today and enable the companies of tomorrow. Thank you for your time.