TaxTech Spotlights: Column Tax

Building the future of personal income tax APIs.

One of the investors in my current startup ModernTax writes think pieces on fintech. He wrote one earlier this year that will have an evergreen effect on startups being built to dismantle TurboTax for years to come.

This week, Column Tax led by Gavin and Michael announced their $5.1M seed funding and it's a very valid model that has a great opportunity to scale.

While I am currently building the consumer-permission access to tax data at ModernTax (among other things to be announced soon), Column looks to build the tax refund and filing API on top of payroll APIs that challenger banks can use to get more deposits.

To be honest, a lot of folks invested in their seed round passed on us (not taking names from any means : )) and we share a few similar investors including Ayo who wrote the think piece mentioned above.

I tend to not be at war with other startups because the real enemies are the incumbents. I wanted to give my insight on how I think about the Column Tax model now that they have made their public declaration.

As someone in the weeds that thinks about these things daily, there is a lot happening. There are at least three other new challengers focused on personal income tax and it’s fairly exciting to see some great minds innovate around one of the oldest forms of finance: taxes.

What is Column Tax building?

In a nutshell, Column is setting out to build a real-time tax product that adjusts employees' W-4 withholding to increase their paychecks and help them file at the end of the year. They are building it on another venture-backed API provider Atomic which has built scraping tools into hundreds of payroll providers. 

Additionally, Column is building an embedded filing solution that challenger banks can implement into their current offering and offer to their customers. The challenger banks are Column's clients because these companies are fairly flushed with venture capital and the thesis for Column is that these companies will pay to increase deposits (on average it costs about $400 to acquire a customer and likely more to get deposits and Chime spent $48 million on customer acquisition in 2019).

The two products Column is offering and plans to offer are positioned to be marketed to middle-to-low income taxpayers via challenger banking platforms that they already use.  Column’s initial goal is to encourage users to get advances throughout the year and file taxes through challenger banking apps they already use versus standalone platforms like TurboTax or a paid tax professional.

Column launched with two design customers Yotta Savings and Klover.

The winning case

If Column can get big challenger banks like Chime and Varo to implement their API, this can be a huge opportunity and Column can position itself as a pure API solution that helps these up-and-coming banks get much-needed deposits directly from the US treasury. The key goal here for the challenger banks you may ask? Increase the interchange fees they get for each customer. It is a fairly simple value proposition.

The government has issued $350B+ in refunds in 2021 thus far. A large percentage of these funds are funds taxpayers are overpaying on their paychecks, the other portion are tax credits the government issues middle to low-income taxpayers. You can imagine a world where Column is driving a percentage of this money to the challenger banks on a year-round basis. 

Challenger banks already have users so Column is well-positioned to be a useful API layer right behind 1) Plaid and 2) payroll API companies like Atomic and Argyle to create an uptick in deposits for these banks. Interestingly enough, Bain led the seed round for Column and the Series A round for Argyle, but Column chose to launch on top of Atomic’s new Withhold product.

The relationship will require no direct-to-consumer acquisition cost to Column and they can both counter-position their product where incumbents cannot win with their current business models. TurboTax and legacy bank products owned by GreenDot and other banks typically have big in-season marketing budgets, charge high fees to the end-consumer and paid tax professionals, and have seasonal businesses.

Column's model is to monetize the challenger banks based on the amount of value (i.e. deposits) it drives by adjusting the user’s W4 withholding year-round.

The lead investor does a very good job of explaining a more expansive winning case (which is expected), so I do not need to elaborate here.

If Column closes the top challenger banks and gets massive adoption <10% this can be big just in the first go-to-market product. They can generate millions on this one product with the right engagement.

The outcome can be larger than a refund advance bank product because the software multiples are much more intriguing, the gross margins are higher, and Column can be valued like a Plaid versus Santa Barbara Tax Group.

The losing case

Column has chosen W4 arbitrage as the way to "advance" consumers their own money. What that means is Column is not necessarily advancing consumers' funds, they are simply changing the withholdings to increase the take-home pay on their paycheck. Initially, Column users will not receive tax credits but only overpayments they would be making to the IRS.

There are two issues here. The target market they are focused on likes big tax refund checks during tax time. The lower-to-middle income taxpayers Column is targeting typically receive their largest one-time payment from the US treasury.

The psychological implications of reducing that will have an impact regardless of how much incremental increase they receive on their regular paycheck.

Six years ago I experienced this first-hand.  In my first entrepreneurial venture, I helped start and run a retail tax chain.  In 2017, I launched an online tax refund advance startup called RefundNote.

In the retail chain model, we offered tax refund advance products that were bundled with assisted tax preparation services in offices. In those days, the customers got back less during tax time because we used banks and hard money to fund the loans ahead of the holidays and collected the balances and tax service fees from the tax refunds. 

We learned firsthand that people did not generally like seeing less on their federal tax refunds when tax time came. This produced churn and sometimes a low NPS score from customers and likely Column will see this as well.

In the RefundNote model, we advanced hard money online and did not adjust the taxpayers withholding but collected it through a network of online and offline assisted tax providers who the consumers trusted.

In 2017, there were no Payroll APIs, so we relied on assisted tax service providers to collect the tax and payroll data at the point of issuing funds. Despite the easy access to funds and working with trusted professionals we still saw churn when consumers saw less money during the actual tax season.

Maximizing the funds they received throughout the year with W4 arbitrage may disappoint consumers when they file their tax returns and get less than they normally get.  There is also the issue of changing withholding programmatically if these consumers are using assisted tax preparation services. If I am using Column through a challenger bank, but want to file my tax return myself or with my tax preparer during tax season there may be questions that come up around why is my return lower this year?

This may cause a lower NPS for the challenger banks leveraging Column and if it gets too low these banks may receive complaints and they may shut off the option and Column may lose the partners and business.

The other issue is conversion, payroll API products are fairly new and somewhat complicated. These solutions have only become a real thing within the past few years and it still remains to be clear how well consumers can remember their payroll username and passwords outside of the application they trust and know.

For every 10 consumers who opt in to get bigger paychecks, there may be only one or two who can successfully or just want to connect their payroll account for Column's Tax engine to do its job. 

Over time, the conversion will increase but there is no way to know how long that will take. Again it may not generate enough net new deposits for their challenger banks to prioritize it and show the option to more consumers.

Final Call

Gavin and Michael are going after a large market and a large incumbent. They have some quality investors and some great design partners. It will be interesting to see how this product converts and what real deposit growth the challenger banks see. 

This is good for the #taxtech community and it is exciting to see how consumers adopt these new products.