Xerocon finally returned to Australia three years after the Covid pandemic. Three years in tech is the equivalent of a decade so it was interesting to see how Xero had changed its product strategy and messaging in that time.
The most remarkable factor was the lack of product announcements for its home market. The press release contained just three headlines; improvements to cashflow forecasting in Xero’s analytics tool, a redesigned job manager for Xero Practice Manager, and the addition of audit history to payroll.
Instead, the keynotes discussed a long-term product strategy based on three waves. The first wave was the transition from desktop software to cloud accounting. The second wave involved tighter integration of applications and services within Xero, and the third wave consisted of advisory insights calculated from internal and external data.
The transition to the cloud in Australia and New Zealand is well advanced. Xero’s paid add-on Analytics Plus was cited as an example of wave two as it includes tax feeds (PAYG and BAS) in its cashflow forecasting. However, there were no tangible examples of the second and third wave within the core product.
I sat down with Anna Curzon, chief product officer, Mark Rees, chief technology officer, Luis Sanchez, executive general manager of partner management and partner experience, and Abby Hempfling, director product & technology communications, to get the details.
We had a wide-ranging conversation that covered many aspects of the Xero platform, including the Future of Xero program, an internal project that rewrites Xero’s core code and architecture.
The following transcript has been edited for clarity and length.
DigitalFirst: I found yesterday confusing. Xero is a product company. It's been three years since we've had a Xerocon. I was expecting a big list of product announcements. But we didn't get that. Why was that?
Anna Curzon: So Sholto, maybe it's a good idea if we talk about our global strategy for a moment. As a global company, we are lucky in the southern hemisphere, because these (Australian and New Zealand partners) are our early foundation members and that's why we're talking to them about our long term vision and (asking for) feedback.
But as you know we're building out the UK, the US and Canada. In the UK, we announced Xero Go. So Xero Go is an amazing, sole-trader type product that supports the community as part of the HMRC (UK tax office) changes. We announced a whole raft of changes around that and in the US, major, strategic partnerships to solve things like sales tax.
DigitalFirst: Tell me about some of the cool stuff then. Let's start with Xero Go. That sounds a lot like QuickBooks Self Employed, which is obviously in the US market. So why would you only release it for the UK and not for the US?
Curzon: Good question. For some time we've been contemplating a product for sole traders, they've really been underserved. We've done a lot for the small business community with our acquisition of Planday and payroll and the larger employing small businesses. And we launched the Starter plan. But we felt that there was an opportunity for a pre-accounting type, think about it as an app to take care of all your financial admin.
As a global company, we're thinking, where's the best place to launch this and learn? And there was a real problem to solve in the UK because the HMRC is (introducing) ITSA, the income tax self assessment. It is for those tiny sole traders and landlords to submit quarterly (digital) tax returns. And so accountants and bookkeepers were saying, “Am I going to get a lot of these people coming to me?” And we thought, that's probably the best opportunity for us globally because that's where the need was the biggest.
So we launched in July on iOS, and it's a freemium product. We're about to launch on Android and include bank feeds and everything else. And you can have add ons. So for example, you can add on Stripe, and obviously you pay for that. You can invite your accountant or bookkeeper on and basically all the information goes through to them so they can submit the tax on your behalf if you want to.
We have had a lot of amazing feedback, both from small businesses and accountants who were saying, “I can move x amount of my clients onto this straight away”.
DigitalFirst: So the UK has regulatory pressure to introduce this. Are there plans to bring it to Australia or the US? Obviously, the TAM (total addressable market) is massive. But a freemium product would cannibalise your existing market and the Starter plans, presumably. Is that a concern?
Curzon: We have considered that. And we will obviously learn about how the product behaves in the UK market. However, the way we've architected the product, it’s probably not going to (work for) a small business that suits that Starter plan. There are some lines in the sand (like invoicing) in terms of Go versus our business plans. We're pretty confident that the Go products will fairly and squarely sit with the pre-accounting community.
DigitalFirst: It’s mobile only, there's no browser based version of Xero Go, is that correct?
Curzon: Mobile only at this stage. For the accountants and bookkeepers, it's all integrated into Xero.
DigitalFirst: Timeline for an Australian launch?
Curzon: We're just going to see how we go in the UK. We're looking at the numbers and as soon as we are ready to pivot, we can move pretty quickly into a new market.
Sales tax (USA) and inventory
DigitalFirst: What are the implications of the Avalara partnership in the US? How do you think that will impact your growth in the US market? What problems does it solve?
Curzon: Oh, my gosh, it is significant. With an accounting product like Xero you also need tax (in any market) and sales tax is a massive problem to solve in the US. There are 13,000 different tax jurisdictions. For example, if you cut a bagel in one particular region of a state and put spread on it in another region, it attracts different tax treatments. So we needed to find a partner to integrate with and Avalara is outstanding. They're the Rolls Royce solution of sales tax.
DigitalFirst: But why now? Xero has been in the US since 2015. Why not earlier?
Curzon: We've spent a long time really thinking deeply about our product strategy in the US and what we need to win. Part of that was we absolutely needed a sales tax solution. And we bought LOCATE as a solution for inventory, e-commerce and COGS. We now are very clear about what we need to do. And we're just making sure we push ahead and invest in it. We have had an integration with Avalara and now we are cementing that.
DigitalFirst: Oh, so like a Stripe integration, a deeper partnership?
Mark Rees: We had essentially a partial solution. To do a full sales tax integration and properly store and reconcile and report on US sales tax we have to add a whole set of additional data to the core of the engine. Sales tax has to show up across the breadth of the product, in credit notes and quotes and purchase orders and stuff. We also announced direct feeds for five of the top 10 banks as well. [ED: Xero also added many more bank feeds in USA and Canada with a new Yodlee partnership.] It's a complex product problem that we're progressively working through to improve our product market fit.
DigitalFirst: Thank you. I didn't realise there was already an integration and this is a deepening of that.
Curzon: There is a real concern that they're going to get sales tax wrong, especially in the US because it's so complex and (the US is) very litigious. And so we wanted to make sure that we're building a solution that is deeply integrated with Avalara and making sure that it's simple, it's intuitive, it's a really beautiful experience. [ED: Xero has clarified that the deeper integration with Avalara will be released next year. At Xerocon New Orleans, attendees were invited to sign up for the US Sales Tax closed beta. The integration will be available next year to a limited number of users before the official release.]
DigitalFirst: What can you tell me about this inventory announcement and how will it be different to just an integration with LOCATE?
Curzon: Inventory is interesting. We looked right around the world for a good inventory solution that was going to solve for small businesses. There are amazing inventory solutions out there but they're industrial strength and suited to larger businesses. We couldn't find anything that was fit for purpose.
We actually had quite a lot of our US partners like (Catching Clouds’) Scott Scharf and others say you need to talk to the LOCATE team because they get this. And they've been outstanding.
DigitalFirst: LOCATE Inventory will be launched within Xero?
Curzon: Yes, it's a full integration. [ED: Xero clarified that it is building a new inventory solution within Xero. It is not just an integration with the LOCATE Inventory application.]
DigitalFirst: Okay, so like how payroll works with Xero? Or is this a core code integration? I mean, there are different levels of integration, right?
Rees: The experience will look like it's a core feature of Xero. But we want to structure their code so their parts are separate, so that allows us to change it quickly. It will still be separate code, but it will from a user experience look like a native Xero app.
DigitalFirst: Lots of people here are going to ask when will it be in Australia, because everyone wants inventory.
Curzon: Well, again, we are testing and learning with LOCATE in the US, and then we will look to globalise it and bring it to our different regions, including Australia.
DigitalFirst: Okay. So, no dates?
DigitalFirst: There are some questions as to how popular certain things are within Xero. So like Analytics Plus – What's the user take up there? Or Xero Expenses, Projects? Can you talk about any usage rates?
Curzon: If you look at our ultimate plan, that was part of the discussion we had with you a few years ago, we were starting to think about the product ladder back then. That includes Expenses, Projects, Payroll, the advanced version of Analytics Plus. The take up of that ultimate plan has been beating our expectations, and we'd set a high bar. There is high demand for that suite of products pulled together, especially if you're a larger small business employing a number of people.
Unifying client records, upgrading reports
DigitalFirst: What is FOX, or Future of Xero?
Rees: It's an internal program, I talked about it in my part of the keynote. We have to structure the work we do to upgrade our product as time passes. [ED: Xero clarified that FOX is a rolling program that continuously refreshes core code within Xero.]
DigitalFirst: So refactoring the software?
Rees: Yeah. Some of the software was built 10 years ago when Xero was a much smaller company with far fewer engineers and a different market proposition. So it starts to become a limiting thing. So this refactor and restructure is in many ways like what we talked about with LOCATE – breaking the software apart so that we can have more teams moving on it more quickly. We can change (sections) independently of one another.
It's like a trajectory of ongoing change. When we migrated to AWS (Amazon Web Services) that was the fifth evolution of the platform. And so FOX is like the sixth or seventh.
DigitalFirst: What parts of Xero is FOX focusing on?
Rees: FOX will focus on the core application code, and then we need to work on the database structure. Ultimately, we're going to have to regionalise the way we host the product – at the moment it's all in the US. It's just part of this journey we're on with scaling and improving the technology.
Curzon: Today is not the start for FOX, we've been doing it for some time. A lot of the feedback we heard, especially from Australian partners, is to come back and fix the things that we've asked you for some time. So things like payroll audit history – you saw the response, and with contacts. And reporting was one of the first things that we included in our work with FOX.
And now you're seeing a rolling thunder of changes so we can pivot come July of next year to the new versions of reports. (Because of the restructuring work we have done with FOX to the core software,) we can respond really quickly to our customers. I was speaking to a partner yesterday, and she was saying, ‘Hey, I've got this idea around reports’. And I'm like, ‘Great, because we're already out of the gates’.
DigitalFirst: Well it's funny you mention reports. When you announced the sunset date for Classic Reports I felt a chill run through the auditorium. There are a lot of accountants out there who prefer Classic Reports, because it has more customisation options… or maybe they think there are more customisation options?
Curzon: Yeah, I will challenge you on that. Because there's a lot of customisation that we've built in. And part of the change is just trying the New Reports.
Rees: Much, much more customisation, actually.
DigitalFirst: Because the perception may be fixed around when New Reports first came out, and it was like, Classic was way better. Overcoming that perception is probably going to be as big an issue as product development.
Curzon: 100 percent. And we need the feedback because we've released things and partners have said, “Hey, I like this, but can you do this which I used to get in Classic Reporting?” “Great, we'll do that. What else, what else?” So by the time we get to that July timeframe the experience is going to be so much better than Classic. For many of the reports it already is.
Rees: And we can double the amount of resources that we put on reporting by us sunsetting Classic Reports. The engineering team that was working on that can now work on the new one which accelerates innovation into the future as well.
DigitalFirst: So with FOX when you are rewriting the core code, that gives you a fantastic opportunity to remove limitations in the product. What limitations are on the hit list?
Rees: There's a whole set of things that we want. It’s not so much direct limits. As part of FOX we're landing what we call the XUI, our design system. Most product companies have this, which is a standard portfolio of components that make up the UI. So you imagine the table or the date picker or those things, this is a standard version that's across all the pages. So as we evolve the UI we can just go and change one version of that table to add, say, a different search algorithm or sorting algorithm, and that flows across all the products. It enables a lot greater change and consistency across the product. That's not a hard limitation, it's an enabling change. There's more work we can do on accessibility, it enables multi-language, all those kinds of benefits come from it. [ED: Xero has clarified that multi-language is a longer term goal. "Xero’s decision to move into non-english markets is about prioritisation. Continuing to focus on the markets we are in is a higher priority."]
DigitalFirst: Multi-entity? Is that on the list?
Rees: Not at the moment.
Curzon: But the way we are architecting FOX as a means to that type of end, it's going to open up. And everything is so connected in Xero at the moment in terms of the code. What we want to do is make specific domains. That's going to give us a lot more opportunity to think a lot more innovatively.
Rees: Two things that you need to think about with FOX. One is there's the work that FOX does, and then the work that FOX enables us to do. By refactoring the code and restructuring it, like with Reporting, it allows us to move much more quickly. So it gives us the ability to support things like multi-entity. We might not do it directly in the work, but it unlocks elements for us to do that change quickly.
DigitalFirst: So you're going multi-language, single contact record across XPM and Xero Tax and HQ…
Luis Sanchez: Not quite. It’s a single client record not a contacts record.
DigitalFirst: Oh so client records are separate to contact records?
DigitalFirst: A single contact record seems like an obvious thing to also tackle. Is that possible?
Rees: The work on contacts that Anna talked about (on stage) is part of the unified contact solution which has a single store for contacts across the product and a consistent user interface.
Curzon: I talked about the fact that we need to do that across 12 different products. Contacts has been a key request for some time, especially in Australia. And we are really leaning on to that as fast as we can as part of this work. But we want to be super transparent with our community about the complexity that we're dealing with.
And I have to say Sholto, the response has been really great. The more that we can share and get the feedback, and the more people understand the benefits that are coming as well.
DigitalFirst: So a single client record at the end of '23. A single contact record, do you have any deadline for that?
Sanchez: Just to make sure we’re all on the same page with the single client record. End of FY 23 we're going to start unifying that client list across XPM, HQ and Tax, right? So that means one client list drives everything, which is great. It doesn't mean every single one of our products is completely intertwined. Like Workpapers, for example, that comes later.
DigitalFirst: Right. But you'll essentially have a client record module that is connectable and the work will need to happen on Workpapers and the other individual products to connect to that module, right?
Abby Hempfling: It probably is worth saying that the UK will have it first.
Rees: There's quite a bit of migration that has to happen. You have to take the data on those separate stores and bring them together.
Curzon: It's a really good point that Abby's made. There is a small XPM community in the UK so we will be able to test and learn there. We've got such a huge community in AU we want to make sure we get those learnings and make it super smooth (when we launch it here).
DigitalFirst: And that includes unified contacts?
Rees: No, contacts are part of the modernisation work.
Hempfling: It's just a constant updating. So we've been making updates for the last year and talked about the multiple addresses (for a single contact) element that will come next. And then we'll need to get it built into all 12 of those products, so it's an ongoing iteration.
DigitalFirst: The big question is that once you get to the stage of adding multi-entity and unified contacts and unified clients, then effectively, you've got that small business platform or an ERP for small business. Is that the target that you're setting yourself, in a multi year timeframe?
Curzon: Yes, absolutely. What we're doing is really building those foundations for the platform. But as we spoke about yesterday, in terms of wave one, two, and three, what we've got to do in tandem, is really shore up the foundation and the architecture to enable these things to happen. [The waves refer to three phases of product development strategy from CEO Steve Vamos’ keynote.] And then start to think about making a beautiful experience across the Xero platform, not just for accounting, but all those other jobs to be done.
And that was part of what we talked to yesterday in terms of seamless connectivity across the Xero platforms, the concept of embedded apps, and also insights and actions. And so we've mapped out a few like that North star moment that we want to get to and working backwards from there. And then the other great thing that Mark’s doing in the data space is starting to think not just about the experiences we can deliver but our position on AI and machine learning and data and trust.
Rees: That's worth talking about. (In addition to) the core data models that you talked about, there's another layer of core data models like infrastructure AI models. Two examples of that: You know we bought HubDoc, they have this document extraction technology which is a model for looking at receipts and bills and invoices and extracting the data. That underpins the work in Xero Go – the expense recognition and expenses – and in HubDoc as well. So there's a set of models that we have centrally that enable us to land all those different doc and extraction techniques underneath.
Another one that's really important is the graph. The connected graph of customers and their customers and their customers. That unlocks things like the App Store recommendation tool. You find clusters of similar customers, draw a circle around a bunch and then look at what third party apps they're using and use that as a recommendation. The same thing happens with a bank rec. Automating bank rec means searching the graph and finding the contact that comes in through the document you scanned, working out where that is, and matching it to the graph as well. So the work we're doing on that kind of foundational AI really does unlock a lot of the contemporary experiences we need to build.
DigitalFirst: So does the software recommendation split customers into cohorts based on say financial metrics? You would want to look at the tech stacks for the fastest-growing real estate agents rather than just what the bulk of them are using.
Rees: We define a similarity measure. It could be a set of financial metrics, location, duration, – we can use whatever similarity measures will be relevant for different scenarios. We can change that based on what problem we're solving as well.
Xero's three waves product strategy
DigitalFirst: Is that the bulk of wave two? Because it wasn't really clear what was in that.
Curzon: Wave two is really about building the small business platform for the future. That's just not about accounting, it's all those other jobs layered on top; inventory, time scheduling, attendance, etc. And that's not just the (applications) that we've acquired, it's the whole ecosystem. So that's really saying as a small business, I can log into Xero and have everything in front of me that I need.
DigitalFirst: What's the practical difference to today? You've got 1,000 apps in the Marketplace. Is that wave two?
Curzon: Wave two is really about integrating those experiences. When I log into Xero, I might be able to pull data through on my contacts page from a CRM app.
Rees: It would show up on the dashboard (data from) a mixture of third party apps, or they would appear on the site or in the navigation.
Curzon: …or cryptocurrency that you want to pull through. And then you might want to integrate that data to create meaning out of it as well, which starts to get you into wave three of insights and actions.
DigitalFirst: So the tangible parts of wave two would be showing up third-party app data within dashboards, reports, client records?
Curzon: And key jobs to be done. It's actually integrating it into your workflows as well. And we really want to make sure that we're giving our app partners the tools not just from a developer perspective but from a design perspective so they can start to build things within the Xero framework or in the Xero platform. So it's not a jarring experience (switching between apps).
DigitalFirst: Got it, okay. So Wave Two is pulling data through and showing it alongside existing data in Xero products and workflows.
Rees: We can make a much more beautiful experience by having a tighter connection, and that involves us opening up the surface area. So third parties can be more present natively in the application.
DigitalFirst: And the third wave is to take data from here, add it to Xero, and then combine it into new data?
Rees: Combine them together to generate these kinds of domains of intelligence which unlock a next wave of kind of insights. The document extraction work is a good example of that. Analytics Plus is probably the best example. We have all the high quality data in Xero and potentially from third parties. Once we know we can recommend actions to our customers, accountants and bookkeepers.
DigitalFirst: Okay, can you walk me through an example with Analytics Plus demonstrating wave three?
Curzon: It's actually made up of a number of products. Short term cash flow is an example and again, it was based on a heap of feedback from our partners. “I've got really good cash flow projection tools for the next year. I need something that's more short term like for the next quarter.” And so that's what we've focused on. And we've been constantly feeding in new pieces of data.
Now we've got payroll, we're going to feed in BAS and PAYG (into Analytics Plus) which is really important for the Australian market. And we'll add more tax feeds in there as time goes on.
Essentially, what machines need are data and time for that learning to take place. A few years ago, we started thinking about that. We've started those models. And now our predictions are getting better and better and better.
So you think about this all tied together into one experience in the future. Wave three. I've got an ice cream shop, and I am integrated with Planday. You'll get an alert (and so will your advisor) to say in two weeks we think you might not be able to make payroll. You need to have a discussion about that. And we only think you need two people on a Wednesday, not five as you've scheduled, because the weather's due to be really bad and we've seen these seasonal downturns, etc.
As I talked about yesterday, it’s about moving from hindsight to insight to foresight, and alerting your advisor as well as the small business owner to opportunities or to head things off at the pass. With that Planday example, it could also say, here are the two best people that you should have on that Wednesday, based on feedback from others. Would you like to book them now?
DigitalFirst: Right, so it is demonstrating wave three by combining data from say Planday and Xero.
Curzon: And other third party apps, like a weather app, for example. It might be a whole host of things that we're pulling together to form a view of the future. Now, you still need your advisor to sit down and provide context. And she or he might say, “Yeah, that might be true, but also have a think about A, B and C”. So we never say that that's going to replace the role of an advisor because there will always be other contexts that need to be overlaid. But what we want to do is arm the small business community with every little bit of data to set them up for success and take control of the future.
Subscribe to our newsletter
Subscribe to receive the latest stories and new guides to your inbox. No spam, we promise.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.