Andy Su, co-founder of InDinero, on tech-enabled bookkeeping's 14-year evolution

You founded inDinero, an early software-enabled bookkeeping service. Can you talk a bit about the overseas labor arbitrage element and the opportunity you saw that led you to start inDinero?

When we founded inDinero in 2009, we weren't thinking about doing tech-enabled services. At that time there was only QuickBooks Desktop, there wasn't QuickBooks Online and Xero wasn't a big tool yet, either. We were actually trying to build an online accounting platform but little did we know that a lot of work goes into that. There's a lot of reports that need to be done and a lot of finicky things. 

The original tagline for inDinero was “Mint.com for businesses”. Back then, Mint was really big with folks. The idea of connecting your bank account to an app was really novel and the technology was really cool at the time, so we wanted to build something on there.

We worked pretty hard on that product for a couple of years and then, we pivoted into this mixed model with services because everyone we asked was like, "Can you do all taxes for us? If you can, we'll pay you 10 times more money than what we're paying you at that time." We were forced to pivot and it ended up being a good thing for us. 

We opened the office in The Philippines and we did that labor arbitrage. We were doing that to find engineers actually, originally. Then, we ended up finding a lot of great accountants out there.

Could you share a brief overview of bookkeeping startups and tech-enabled finance back-office services companies since inDinero? Tell us about the evolution of these products—Pilot, Bench, and others—from your perspective.

I'm not sure about Pilot specifically or Bench, but for inDinero, when we first started it, you’d get an online dashboard to connect your bank accounts to your app and re-catalyze your transactions. We would file your taxes every year after going into those things. 

That transitioned into us doing your accounting for you. You’d just give us your bank account and credentials and we'd connect everything for you. We'd categorize all the transactions for you and only ping you once a month for month-end closes. That's the standard now. 

The pricing model also, back in the day, was since we do your taxes once a year, you’d pay us once a year. But that's evolved into a monthly model of probably, I think, starting at $500 a month for most people.

Permalink

Pilot as tech-enabled services—vs. the previous generation—hinges on the rise of SMB finance connectors like Stripe, Gusto, Expensify, QuickBooks, etc. that enabled bookkeeping to be more “programmable”. Can you talk about the promise of connecting all your services and having it automatically do the books vs. the reality?

We're still a long ways away from having your taxes automated. The problem is if you write a check, someone needs to tell the system what that check was for, who it was to. That manual work will always be there because your bank account is just going to say check number 123.

Also, if you buy things from, let's say Amazon, you don't really know what the items are for. There isn't really itemization. It could be for office supplies, furniture, or even electronics, and someone who made the purchase or is familiar with the purchase, will need to categorize it as such to do that bookkeeping. 

We'll get closer to automating it by getting the person who made the purchase potentially to make these transactional categorizations. But within the company there might be different purchasers. So, you could theoretically have an app where you ask them, "What is this purchase for, for your sales team? What is this purchase for, for your engineering team?" And the Head of Engineering or the Head of Sales will go in and do that once a month, probably.

Can you talk about challenges in the data layer in doing the books programmatically and with a high quality standard? That problem hasn’t been solved with Plaid, while on expense categorization, companies like Expensify, Teampay, Brex and Ramp have made strides by building a workflow around virtual cards.

There's two different pieces of software here. 

One is the accounting software, which is QuickBooks Online, for the most part. Pilot uses QuickBooks Online, and I think, most of inDinero's accounts are on QuickBooks Online, as well now. 

Then, there's the separate in-house tech, which is more like a CRM tool which connects the customer, and shows them, "Here's the charts, here's the graphs, here's the high level view of your finances, here's where you go to do task items for your teams, categorize your transactions and whatnot." 

But the work itself is done in QuickBooks Online and QuickBooks Online is like the software that has the statement of truth, source of truth.

Having the CRM tool is a differentiator. A lot of customers like to have that high level touch when you're paying a premium. So, for inDinero and Pilot, they're going to be more expensive than your Chinatown CPA, but the software that they provide gives you a sense of peace of mind that things are getting done and that there's value involved.

What’s your take on the rise of AI-powered bookkeeping a la Zeni and Digits? Like with labor arbitrage, do you see an opportunity for a 10x reduction in the cost of labor of bookkeeping or improvement in the customer experience?

I don't know. I don't know if ChatGPT is going to be very useful here. 

I built Pry and one of the things I built in there was the ability to classify a transaction and then, it would automatically classify all similar transactions like it. That minimizes the amount of work you have to do, but it still doesn't quite automate it away. 

If you are always sending a check to your landlord every month, then, it will recognize that and it will, but the first time, you’ll need to say, "This is rent." Then, all other times, it'll say, "Oh, this is probably rent. We'll categorize that as such." 

That's what I had in Pry and it was pretty good. I used the feature a lot and I think, it was pretty helpful. It's really basic, nothing crazy, but to just know right off the bat that that transaction is rent requires a lot more sophistication. 

I don't think the AI is going to get there anytime soon. It could and this is potentially interesting to think about and there is potential here, but because everyone knows that every company needs to pay rent and there should be a large transaction somewhere that is a rent payment. 

However, I think there's so many different things in accounting that need to be fixed before we AI everything that we're probably not going to get to it any time soon.

Permalink

Tech-enabled bookkeeping tends to be the core service offering on top of which tax, fractional CFO, FP&A, R&D tax credit and other services get layered. Can you talk about that cross sell motion at InDinero and whether the typical customer uses one, some, or all of the services offered? What future opportunities does owning bookkeeping get you to cross sell into?

Yeah, R&D tax credits is one of the largest and one of the biggest ones because there's upside and people are willing to pay for it. 

Another one is just FP&A—financial planning. That's usually done in Excel and it's probably $500-1,000 a month at least, to set up budgets and forecasts. I think those are the two main cost sales. 

There's also different taxes, like the K1s and things like that during tax season, but those are just an extra part of taxes that overseas and international customers need to do.

So, for the R&D tax credits, I don't know off the top of my head how much, how many companies opt-in for these, but I would say that a good number of companies do use our R&D tax credit service.

Permalink

QuickBooks offers its own bookkeeping service and so did your FP&A startup Pry. Gusto does R&D tax credits. Can you talk about the reverse approach of selling services from a software offering? How is that positioned against a tech-enabled services offering? What does the competitive dynamic look like? How do you think that plays out?

So whether R&D tax credits are better done through Gusto or a firm like Pilot? Gusto has a really interesting R&D tax credit service. It's different from most other ones I believe, where they're just paying it, withholding it, and they can tie it into payroll. 

Accountants can use Gusto to do R&D tax credits and that's the synergy there. But generally you would want a CPA to make sure that the R&D tax credits are handled properly. 

I would be reluctant to just go to Gusto myself, and build and use their service without having a CPA look it over. But it's possible. There's folks out there who don't have CPAs and they do their own accounting. It's the same idea. You could get a CPA to do R&D tax credits for you using Gusto or whatever tool you prefer. Or you can do it yourself and there's risk to that.

For R&D tax credits, you're often looking at hundreds of thousands of dollars. When that's at stake, you generally want to have a CPA just to make sure that things are done properly.

Imagine Pilot automating 99% of the process of doing your books for the set of less than 20 employee companies. What does that give Pilot the opportunity to do? How would you argue the upside case for Pilot at its latest valuation?

Things can be automated until you can't. There's all these different rules that you can place in that. Like I said, you can assume that some payments are large, payments are checks. You can assume that large income is investment and there's rules that you can build to automate some of this, but it's not going to be a hundred percent. 

There's always going to be that one person who Venmos the customer's money back and you have to account for that. It's the edge cases that really kill you with that. 

But for most companies, I think, I also built a bookkeeping platform for Pry and it worked pretty well. For most companies as long as you categorize the first instance of the transactions, you won't have to do it again. That works pretty well. It doesn't take too much time.

But fully, completely, you don't have to touch it. Any company that's 20 people or less, I'm skeptical that something like that could really work.

In recent years, QuickBooks has become ubiquitous and the core of the SMB finance stack. It also makes the books portable, which reduces switching costs to another bookkeeping provider. Can you talk about the central role that QuickBooks plays? How has it won the segment? What does its trajectory look like going forward? Do any players like Pilot or Bench have the opportunity to displace QuickBooks?

I don't think Pilot wants to displace QuickBooks. The thing is that QuickBooks is a really big piece of software. There's a lot of accounting rules built into it. It's being worked on over many years now and it's quite stable. 

Another platform that does really well is Xero. A lot of companies use Xero as well, especially overseas. The way that they've become so powerful is for QuickBooks, it's their core practice that they've been doing this for many, many years and QuickBooks Online was the obvious flagship product that they had to push.

Inputs got most of their resources behind QuickBooks online. So that's one reason why they're doing quite well. They're really focusing on it. I think Bench is more likely to try displace QuickBooks because they have their own accounting system. 

But I think you have to be a bench user in order to get that piece of software. I don't think they're licensing it out to accountants or anything like that. So I wouldn't say they're going to displace QuickBooks. They just aren't using QuickBooks. They are using their own platform. 

For Pilot and inDinero, I think they have a lot of work to do on the CRM side before touching the accounting product. So it would make sense to continue partnering with QuickBooks Online or Xero for the accounting portion.

It's really not that much unless QuickBooks for instance, starts charging a lot more money for these tech-enabled services which I don't think they will. We'll still be using QuickBooks and not displacing it. There's just so much accounting knowledge built into QuickBooks that it's hard to rebuild.

Permalink

How would you argue the strongest possible upside case for Pilot which is now valued at $1.2 billion?

The market is pretty good. Almost all companies need some form of bookkeeping and accounting help. 

Pilot is one of the more premium players. They've been growing pretty quickly, but the monthly subscriptions that Pilot and inDinero have are here to stay. They work really well. They help make margins. They grow. 

For inDinero, we got margins close to 50%. I don't know about Pilot's margins, but they can probably do pretty well as well.

Permalink

This transcript is for information purposes only and does not constitute advice of any type or trade recommendation and should not form the basis of any investment decision. Sacra accepts no liability for the transcript or for any errors, omissions or inaccuracies in respect of it. The views of the experts expressed in the transcript are those of the experts and they are not endorsed by, nor do they represent the opinion of Sacra. Sacra reserves all copyright, intellectual property rights in the transcript. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any transcript is strictly prohibited.