Behind the API Interview with Richard Fortune, Xero

Welcome! On Behind the API, we talk to the people who work to build awesome API products about their journey, learnings, and overall approach.


On today’s session for Behind the API, we are joined by Richard Fortune. Richard is a Product Manager at Xero in the Developer Experience area, as well as is very busy with advising on product strategy in his spare time.


Want to do audio only? Subscribe to the Behind the API Podcast.


To request an invite to Astral for your low-code API Portal and Developer Success Tools, blast off with us here.



Transcript (transcribed by Otter.ai)


Kirby Montgomery 0:02

Welcome on behind the API, we talked to people who work to build awesome API products. And we learned about their journey learnings. And then just overall approach to that on today's session for behind the API, we're joined with Richard Fortune, Richards Product Manager at Xero, and the developer experience area for over a decade. And he's also very busy with advising on product strategy. And in spare time, we're gonna learn about all those things. But to start us off, Richard, you want to give us a little background on what you currently do at Xero? And where you're based out of


Richard Fortune 0:33

Sure Kirby! Hello, everyone. My name is Richard Fortune. And yeah, I work for Xero and one of its offices, here in New Zealand, in Wellington, and for those of us who don't know, the Xero is a global accounting and small business platform. And I've been with them since 2010, I believe. So you know, back when it was 60 staff, and now it's at 4000. So you can say we've been on a bit of a journey. I work in the ecosystem team at Xero. And I've always have, I consider myself a bit of a platform, business native. And nowadays, I manage the developer experience product and the partner experience products. And, yeah, I'm loving it. I used to be an engineer, I worked in quality. And so I've kind of been quite hands on with all the API's of Xero built in the early days. And when the opportunity to move into product, you know, tackling the problems that partners and developers as customers face came up, I jumped at it very well, thanks for having thanks for being with us today. And you and I had had a conversation earlier and you had cultivated out to be they absolutely loved you said, you know, when a product transitions into being a platform and when platforms become ecosystems, and Xero is like at the forefront of that. It's known as one of the companies that has some of the best API documentation and you see it embedded in a lot of, you know, tools like Codat and Rutter, and things that are syncing all the systems, but obviously, it wasn't like that all the time. And so your team has built an awesome experience. But if we were going to take a step back, what were the early days, that's Xero, like, what was it like 10 years ago, when it was first, when you guys were making some of those first decisions aren't even thinking about developer experience? Yep.


Richard Fortune 2:30

Back in the early days, we had, you know, Rod, the Xero founder, who was quite a visionary, you know, he saw the opportunity and moving, you know, the ledger to the cloud. And at the same time, you know, very early in Xeros live kicked off the API team. And there was three of us to begin with. And you know, that's it's quite a shallow team to take on such responsibility. And so in the early days, it was we were winging it, like any other startup, I believe that the original developer who created our API's, and I kind of set that pattern that we followed with, did that kind of as a hack project to the side, you know, and then was moved off to another problem space. And so we inherited this thing. And, you know, as hilarious as things are in startup world, we even inverted the rest verbs. And so I don't know if anyone notices that we've got those back to front before, you know, before there was real rigor over this thing. And you know, that was not fantastic developer experience. So you know, in the early days, we were very hands on with our developers speaking to them so regularly, because you know, many of them are based in New Zealand, seeing the opportunity and connecting with Xero. And so we're very, very happy to meet with us in person to meet with our product leader at the time and 20 rule, and really grow together with us. And so that was the early stages, we didn't feel like, we felt like we were enabling others. And we just felt like we were doing as much as we could to expose the value of Xero to others. And so at that, at that point in time, we were just laying down, you know, the info we thought others needed. Probably of note back then is that API first wasn't really a thing. And so we were duplicating that core capability, that Xero hat, which is an enormous anti pattern, something that we're all very well aware of nowadays, but what that meant was, you know, even just working with developers 10 years ago, they would see something available in the UI and go, Hey, why can I get this V your API, and we're like, Well, funny story. And all that logic is embedded in the front end, and we don't have access to that VR API's. And then we're like, well, whatever it is. So yeah, so back in those days, there was a lot of education and advocacy and growing together. And, you know, something I've been reflecting on, you know, more recently is that, you know, you kind of forget how much you've forgotten. And I think that's very much the case here. You know, back in those early days, none of the very firm patterns that existed, they were around and, you know, the world is a better place for it, you know, and surely stuff we'll cover today Corbijn what that API landscape looks like and how it's improved.


Kirby Montgomery 5:01

So whenever you had that first core chunk of customers, and they were providing you feedback, and it was a team of three, when did it switch to where the company started saying, like, wow, we're getting a lot of good feedback here. Was there an inflection point where it kind of went from feeling like startup and a hack project? And you know, Richards wearing all the hats with the other two team members to like, wow, we've got a full team here, or this is becoming part of our go to market strategy. What was kind of the like, when the winds shifted?


Richard Fortune 5:34

I think, yeah, that's a really interesting question, because we will kind of build so for the first five years, I think we were just enabling growth that was a multi enabler for Xero day, you know, our CEO really saw that and so that by and existed, so mainly, the priority was, which internal or which product capabilities to be one to externalize. And so you know, and like I said, because we want this divergent path where we were replicating core experience, that's what we're doing, we're going, Hey, look, our customers are telling us they want access to invoices, or they want to be able to do this. The flip side to it was, Xero has always been kind of a part of its go to market has been around partnership. You know, we partner with accountants very early on. And that was a USP, for us. It's a partnership was very core to our DNA. And so looking at that, that partnership drove all the decisions in what we prioritize as a set of API's. And quite often, the developers we were working with, we're employees of sass companies are founders of sass companies, that was less the grassroots kind of developed, you know, very tech savvy developers that are kind of foraging throughout the world today, leveraging opportunities back then it was more people to saw an absolute gap or a gap related to what they already want to do. And you know, Fintech is a stand, that kind of platform opportunity, I think that's kind of presenting itself is very true, when you look at all the big kind of platforms out there, you know, this FinTech space is very much, you know, ripe for disruption and opportunity. And the value is very obvious to people because it translates into efficiencies or business transactions very quickly. So for us, that the early days was more about enabling that, or enabling practices to operate more efficiently, or to migrate off what they were doing, you know, there's a whole, there's a whole transfer that's happened over the last decade, you know, where on premise was the norm. And so people have to get off the cloud. And so as that you're getting onto the cloud, sorry. So as that was happening, our API's enable that whether it was mass migration, or new in your new opportunities to be exposed a standard company that comes to mind as a point of sale app that's built out of New Zealand called vend. You know, they were very early partner with us, our current EGM, Nick Holdsworth was, you know, their number two employee, coincidentally, he is, and, you know, they've grown up with us and then have become their own breakaway success. And I think exit two, I think it's at lightspeed for like 400 million, you know, two years ago or so. So, you know, that that, yeah, and it just never felt like we enable that it felt like it was a partnership. You know, and I think that's, that's always been the mindset. And so when people were giving us feedback as to give us a or b enable us to do this, it was our product managers at the time, who really understood the what that enable meant would unlock what kind of capabilities that would unlock and not just for one, potential partner or death, but for many. And so that's kind of the lens we use to prioritize what got built, what didn't or what was an emergency fix the real challenges when you're an API team, all the peripheral effort that needs to go into providing that as a service. So we're talking about gateways reporting, internal end, in the back of house administration, all the good stuff, that means you can, you know, expose API safely or have an open platform. And so those are all the things that we do in the background, that people in our forums are going, you know, we want this now give it to us now, right? Yeah, you know, no one wants to hear excuses, and we don't like giving them but that is a reality. If, you know, I think Christina was on on that talk with us as well, you know, Pandian Compendium, offer that as a service as well, the ability to manage relationships, but any, anyone who's trying to expose, you know, API's, they will inevitably either have to find someone to provide that solution for them or build it for themselves.


Kirby Montgomery 9:24

So since you guys were at such a forefront, you know, 10 years ago, and you know, you're kind of building up in this space. You're starting to talk about the ancillary things that come with having that API. Was there one like re architecture moment or two or seven or irritation moments where it's like, wow, we're growing so much. And then everyone's like, this technology is the end of life. You're the tech that's too big to carry and gonna get these like parallel situations going on. And then the Richards coming in from the side saying, oh, yeah, we have all this internal tooling we need to build as well like when you're when you're in the product management space for Growing and successful API product, when do you start to think about your pre plenty of that stuff? Or what was kind of the some of the kickers for you along the way?


Richard Fortune 10:09

So the team, that the API team, we've, that were kind of there at the beginning, obviously, we went through some crazy times together, you know, there was these massive growth opportunities. You know, for instance, in the early days, we didn't have rate limiting, you know, so essentially, there wasn't the kind of controls that you needed. We've had massive migration Xero itself migrated off Rackspace, that kind of what year it was, but moved over to AWS. And so that in itself, was this enormous, you know, organizational shift from the API perspective? Absolutely, I would say auth. One was a an enormous migration led by a standard PM, who's now gone to Atlassian, Adam Moore. And it's never as easy as you expect, you see it coming on the horizon for years. And because you've now got this distributed set of relationships, managing that and moving that and orchestrating that is a Herculean effort. And so like, hats off to Adam Moore, who led that for us, you know, he had to work with partners who didn't want to move, you have to work with devs, who were like, what I own that thing, what I haven't had, that you know, that you use, or, or even non technical businesses, you may have outsourced that problem years ago, and are still getting value from it, but have taken it as table stakes for their operation, because they didn't need to do anything with it. And then all of the out of the sudden that kind of on what I've got to update some code somewhere to go to. So we've had several, you know, the implementation of webhooks. Probably some kind of nitty gritty, moving of tech to enable operations to move faster. But I think all of that experience in the early days has really informed how I look at the horizon as a product manager for Xero. So you know, probably more here, in line with your question. We definitely look a few years, I had no way more than we used to, because you get the breathing room to think more strategically, you know, the early days, you've kind of like this needs to exist and needs to exist yesterday. Nowadays, it's like the organization of the value of creating needs to exist and needs to continue to exist for decades to come. So big ones, you know, for us would be our integration management system. That was we platform two years ago, I led that. And in the replatforming of that, we turned that into a platform product itself. And I mean platform product insofar as it is a product that offers a set of capabilities, but anyone in Xero can build capabilities into and it can grow. So outside of our pod, other teams can basically extend its functionality. So that Xeros ability to respond to scale is horizontal and kind of that way, then we can move. And we have a unified library UX library that we can leverage. And so that was built by another part of Xero. And this is something I refer to as platform patterns is like when a business creates value for itself that allows it to 10x later, that's, that's a real platform pattern for me. And so like, that's what's happened with Xeros UI library, we were able to leverage that, we're able to leverage other patterns internally. And that any team at Xero can build integration management experiences, and expose them to our pod. And so our pod is like, three engineers, it's extremely still, like, we're really thin on many, in many ways. No business ever has access to all the resources it needs, you've got to be incredibly strategic in the way you plan. And, you know, the, just your point there, the experiences, we had an early days, I've definitely informed how I prioritize, you know, because when we when we talk about building integration management as a platform, people like Woolworths, but that's a heavy a heavyweight decision to make, you know, took, you know, two years, but the payoff will be, you know, into the decades, like, you know, outside of the tech itself becoming out of date, that capability will not do, you know, and it just really sets us up for success. And as you know, I'm leaving Xero. So, you know, I won't be around the domain knowledge I have won't be around. So it's quite nice to know that we've codified our pattern, and it's there for others to leverage. So that's, that's the kind of next level thinking we're kind of aiming for in our group, it's kind of like looking to the horizon and what needs to be there. Because you know, companies grow, people move on where you still need to have that code.


Kirby Montgomery 14:15

So for a project that you as a two year build, it's like a sizable investment. And it kind of gets to be too late in your career, you've been able to work at Xero with external developers kind of listening to those first communities. and I are talking about even how to unlock the next value inside of Xero allowed people to use this integration platform. So before you even write any code on that, how, what are some methods that you and your team use to be as user centric as possible, and try to uncover opportunities for developer delight that, say, for example, other companies or product managers, like myself could learn from to be like, Okay, we're gonna be making something new here. It's going to be a big investment. How do we make it awesome?


Richard Fortune 14:57

is a great question. So RGM Danyang he's he's, you know, very much been focused on developer delight for such a long time, he's, you know, he's been really embedded with our community and and been quite high touch, you know, he's kind of had some, you know, in person events, so he's really championed and a lot of depth TV and stuff like that. So we've been quite present and trying to, you know, connect with our developer community. I think when we talk in terms of developer delight, a challenge we faced is kind of resetting some of the journeys we've been on and kind of going, Hey, look, it's very hard for us to do that right now. Because fundamentally, the delight will come when they achieve success. And so you know, whether you've kind of got to get to that base level, are our API's, unlocking that success for others? And then the next level up from that? How easy is it for them to manage the relationship once they have a relationship, because that's continuing to enable success for them. And so we've kind of come through those table stakes quite reasonably, I feel, you know, our users might definitely will have some feedback to tell me I'm wrong on that one. But there's table stakes where they can, you know, the, the relationship is functional. But more recently, we've, we've kind of stepped back a bit, I'm starting to see a lot of system service design, you know, to kind of go look, really, what are the jobs to be done here? And who are the people we're trying to enable to be successful? Are they an employee? Are they a contractor? Are they a independent Dev? who just wants to seize the opportunity of building on Xero? Or are they a SaaS company? Blah, blah, blah. And so all those different developers personas kind of have to be well understood. And so before we can do that, we need to step back and say, How are we making those people successful? The upside to that is working internally at Xero and influencing the developer community inside Xero and making them aware of that opportunity when they make decisions. And so luckily enough, our CTO has done a big reset on that. And we are moving towards being an API first organization. And that way that it makes my job a lot easier to go. When you write the deck seems like we have an API catalog internally. And so when you write the docs for your API's, if you create them an open API spec, we've got a workflow that can consume that. And now create public documentation in one nice, seamless workflow. So your team doesn't have to worry about how we externalize and attract that to build a new thing. So we kind of go like, and we can reduce the time to API consumption. And so there are two ways to look at it is how quickly can we get new value into the hands of devs to delight them? And I don't feel like I've answered your question. But you know, the touch points as well, we can definitely enhance those. I think I'm probably avoiding it right now. Because I think it's a work in progress that I can't, you know, really speak to you. I think it's okay, and you've acknowledged it, but we always feel we can do better. Yeah,


Richard Fortune 17:36

there's something that I really taken away from them this sense that a lot of times at a company, we're breaking up the monolith, we're making micro services, and then some of those micro services will become available externally. But even in your talking about your time in developer delight internally, where it's like, I have an internal developer community, and oftentimes, I feel like there's conversations around oh, we need to Dev Rel, who's gonna go work with all these external people, but not even kind of view tree in that tribe internally as their own community as well, that is just as good of tools just as good documentation. And that's a really cool takeaway. Now, outside of Xero, you also do a lot of consulting, and you know, obviously, have a product mindset. And so if a new company was going to come to you and say, Hey, Richard, like we're going to launch this API product. And earlier, you had mentioned about table stakes, table stakes are changing. Well used to be delightful, maybe five years ago is now just table stakes. We think about like documentation, tools, see it box, whatever it is, it's so what are some of the things that you would say to the startup were like, No, it's not delight. it's table stakes. And then what are some things you might recommend or like one to two pitfalls to avoid? For launching a startup launching new API? Yeah.


Richard Fortune 18:54

I guess, let's assume that they're unaware that their future depends on it. So let's just say they're not a stripe, you know, that they're like, you know, let's, let's assume that because if you're a stripe, you are going to do whatever it takes, and you have that kind of almost codified into a culture. So let's assume it's someone who has, who thinks their position is x, that we solve this for others. I would say dial forward five years and years and look at the landscape around you and say, how are you going to protect? Protect your your niche? Or how are you going to increase the value you can create? And, you know, I firmly believe that, you know, you're starting to see this go to ecosystem language emerge on LinkedIn, I think that's the very beginning of the wedge to where, you know, non technical people have an awareness of what integrations between SAS companies really, really means. And if you're a technologist building software today, you know, if you have the awareness to do that, you've got to look at how mature the rest of the world is going to be five years from now. And, you know, they'll be coming to market going. I know X, Y, and Z provides this capability. Let's speak to that. They're tech and augment our product with that. And if you've shut the door on that, now, it's gonna be very costly to fix later. And it might not be you know, there's a lot of people just building a Graph QL for the mobile experience, yada, yada, so it may not be. And so that may not be as much of a switch, as I've experienced. But I think nonetheless, having your position very firmly addressed in your mind, who are we what are we here to solve? And how are we going to continually solve it or grow what we're solving, you've got to look laterally. You know, one of the things I'm hyper aware of, which we may have touched on talked about before is externalizing value creation, if you think about partnership, as the your ability to leverage others to create more value for your customers and the partnership, you can negotiate the details of that later, you know, the more valuable you are, obviously, the power dynamic shifts, so that's less good when you're a startup, depending on what you have access to, you're probably going to feel vulnerable to that you're going to feel a bit at risk. But thinking longer term, and being aware of that pattern means that like you can be aware of who we want to partner with today, who do we need to partner with you, we definitely not need to partner with because just because they're a big name brand, they provide zero value to the customers we're trying to reach. And so a lot of this is potentially less technical and thinking but you know, I feel like API versus table stakes, then accessing people who get value from leveraging your API's is next step is that you're going to be probably a go to market. And then after that, it's how do you create success together? When I think back to that vendor who exited, you know, they created that value completely outside of us, we can take no credit for that. But we were definitely part of their go to market, you know, it was by using XERO, the point of sale services that provided were end to end much better, because the results ended up in the ledger at the end of the day without any effort. Yeah, one solution that feels like it's a no brainer now, but it was novel back then. And so what can you do the equivalent of that in your space today. And so an example of someone I worked with, I can't take credit for this, we worked together at Xero. And then they left was a data science company called I'm asking them, it's embarrassing. Anyway, we'll edit that out, we can fix it in post production. But you know, we work together at Xero, the data science team, and you know, I made them very well for ecosystem play. And then when they exited Xero, they basically started building a bunch of data science services that were exposed via API's straightaway, it wasn't just this in house thing, which meant that one of their probably can't talk to too much in detail. But one of the earliest early success stories was, you know, being able to partner across the globe with with an up and comer in Canada, which then led to them being able to get listed, you know, and this was a trajectory within two years. And I believe it was just that fundamental awareness of the ecosystem opportunity. Being API first meant that the product that they built was externalized, double, but also that their path to growth was easier because of them feeling less protective about their castle, they were like, No, we know what we have here, we've got extremely strong foundations, the castle isn't the thing of the activity. So I feel like that's just a very easy example. But But there's other examples with, you know, the company I work with in Germany, just the fundamentals of how easily they can transfer data back and forth between a product they offer, they have a product that was a time lapse offering to marketing companies. But actually, those time lapse cameras started reporting back and the data they were reporting back meant that they were getting, you know, weather data temperature data, they were then started to analytics over the activity on the camera feed. So they would say look, you know, during COVID times it was masked adherence, when it was security was like X number of people are wearing hard hats. And if you were wearing a product hat, you never would have seen the opportunity, you would never you've never even had the pipeline to push the data back into your core systems. And so


Unknown Speaker 23:53

that's, that's really interesting, because I feel like part of the coaching that you'd be giving this hypothetical startup is not only knowing who you want to partner with to where it's like Know your value so a startup super example has a unique value proposition so cool, you know your unique value proposition you know your customer so that value proposition should be true if you go embed yourself in a partner, helping them increase value to their customers or if you bring that partner into your target customer, the you know your value and risk giving them kind of ancillary or additional value on the side and so they give a little bit about the products that inspire you Who do you think are two companies that you think are just crushing it in that space of thinking about ecosystem and like knowing their value and then where their value works with partners in or going to put themselves instead of partners?


Richard Fortune 24:45

I I kind of want to say stripe but that's a no brainer, so I'm gonna steer away from it but but but one thing is I


Kirby Montgomery 24:54

saw your LinkedIn your mirror certified so I think that you might really like the mirror was


Richard Fortune 24:58

one of my heroes. Well, actually, I've got records on there. I'll do straight from an analogy mirror, and I've got one, okay. So I love stripe, because stripe is under the hood, I don't think people really recognize it. But they know what they're about their fundamental transaction infrastructure, if you kind of observe what they're doing, and I will loop back to this on the third company I talk to, so, you know, it looks like they're enabling payments. But literally, they've figured out that they're just kind of going, we don't care what the transaction is, we will enable it. And we will enable anybody to leverage our ability to do that. So it looked like they were a payments company, actually, no, they were getting brilliant at enabling transaction management and facilitation. So they are like mycelium growing underneath the world. And you know, they're just, that's why they're valid, in my mind, as amazing as they are. Mirror. I love Miro. Who doesn't, you know, I was even a mirror user back when they were called Real Time board. I've searched my Gmail to see if I still have my original. Yeah, well, they're original. But that's another story about product lead growth, because I consume them. But I died in the water because they weren't great them. But obviously, they unlocked something in the meantime, but I monitor them because I think they obviously have huge potential. I look at their pattern. And that seems to be heavily outbound. And I'm quite interested because I feel like, you know, they have a very numb, this is no sleight anyone from mural Absolutely not, you know, your worth, these words will not harm you. But my observations from looking at that is, it's a mural centric view right now. And that might be because of their maturity, they're like, look, let's get people able to conduct activities on us that they're used to outside of us, because that keeps them on us. The next level for that is for that activity to create extended value. So that's quite abstract areas. But let's just say we connect to drive, and you can now you know, see what's happening and drive. The next level that will be deeper integration with Drive. So that activity in your mural board gets updated as somebody updates something. So mural is actually where I'm monitoring my drive google drive activity, because Google Drives a dumpster fire. So let's be under no pretenses. There are seven ways that it could be improved. Amuro could be one of them. And so I think we're all moving beyond but this is all about maturity. And they are a product company with product lead growth. So they'll end up there eventually, I'm sure as they focus on inbound, and how do you create value for end users and and scenarios I think of is, this is a common one I do as a product manager, I will do a lot of screenshotting of the websites we own and then want to present that back to the team. So we will do a full sitemap, I would like to communicate that visually. So I'll take a screenshot of all of the pages that we want to talk about in that day. And I would like to manually do this right now. It's a pain in the guts, but to feed the data underneath. So we see like pageviews per day, all that data, which would be in GA, why shouldn't that be an active feed on my on my mirror report as well. And that means good? Yeah, that's a good feature, I would use that that's a good picture. Yeah, like, and wouldn't it be nice if we built it for them. So that's where that's the power of ecosystem on my mind, if you can get comfortable enough your position, your moat is there others will create that value for you. But all boats lift equally, then at that point, you could still be a $400 million company underneath their $17 billion valuation, and they should not feel threatened by it. I think a lot about this, I feel like this wasn't my third action, but you can tell I'm love this stuff. I think about I feel like Canva should purchase notion that is my top tip of the day. Why do I think that it's my lack of understanding of the Canada valuation. So anybody from Canada call me and this one is more, I feel like Canada is a communications tool. It is amazing. But what is the what is the end journey for their ability to provide communications, you know, they like to communicate visually, and or can do more later. I feel like notion has the data that would enrich it. So therefore, imagine if you could programmatically turn it, you know, they like to manage it, and that it's totally left field, I realized, but I looked at the two products and feel like one is better at doing communication. And the other is better at just completely allowing you to tear apart data relationships, and manage that. And, and interestingly, both are trying to do their API plays at the moment. And so that's why I say they should marry, even though I mean,


Richard Fortune 29:26

airtable may come in and try to do that as well. Because what you're looking at is you're gonna get these, all the content that you need to make the visuals is sitting in this kind of database thing and say, How can I then kind of marry these two together? Ultimately, when yours can? They're looking in front of that? You mentioned something about screenshots in Miro and wanting to get me to an earlier piece of advice that I was curious about. I've had this problem myself, where I'm making an API and let's say for example, I'm in a sales pitch. I need to show the value that my company provides in their price. Direct, and it sucks. I'm like taking screenshots of the product kind of like, you know, putting either some hacky mock ups? How important is it for storytelling, whatever you are talking about an embedded feature, or trying to show somebody the future that can be enabled by some of these things. And, you know, I'm thinking a little bit about even just the Canva example, they says, like, how could you start to explain, like, hey, there's so much value that can be unlocked in the ecosystem? Yeah. So talk a little bit about your career, how storytelling changed for you, because starting in the engineering space, and then kind of moving over to this bigger vision has been obviously something real for you.