How is H&R Block Still Doing 20 Million Tax Returns?

Service as as Service

As the hype continues, I am more and more interested in discussing why certain business models aren’t changing in the AI environment. There’s the famous Bezos quote on betting on things that won’t change (faster delivery times, better customer service), but I’m going to discuss exactly why the effects of LLMs or robotics or even the internet in general won’t disrupt a business that maybe otherwise could or should be disrupted.

H&R Block prepares tax returns and offers auxiliary services to its clients. The company and its 10,000+ retail locations has assisted in the preparation of around twenty million tax returns per year for 25 years despite stiff competition from TurboTax, the federal government (1040-EZ was retired in 2017, but still). No real growth in the numbers, but still, the whole tax prep ecosystem seems to have a permanent class of 20 million people who demand H&R Block’s services, with revenue per return CAGR’ing at 2.7% per year for 23 years:

So who are these 20 million people and why haven’t they figured out DIY or a cheaper private practice? Or e-filing or anything? These 20 million people aren’t luddites, we all have smartphones and can find cheap alternatives (Reddit exists!). For a significant portion of them, H&R Block still offers value. The mechanical realities of the American revenue code and the complexity of modernity have compounded greatly in recent decades; for many, paying HRB $185 makes life much easier.

The median H&R Block client earns between $40k and $80k. To these folks, tax refunds are the largest single financial event of their year. Roughly 24 million Americans receive the Earned Income Tax Credit (EITC), with the average refund hovering around $2,800. For these people, a mistake on a DIY app risks not only an audit— they might miss rent or a car repair. They pay H&R Block for audit insurance and peace of mind. When the stakes are 10% of your annual after-tax income, free software or a chatbot feels like a high-stakes gamble. I could learn how to tile my bathroom on YouTube, but I’d rather just pay the markup and get it done right.

While the 1040-EZ is gone, the complexity for the $100K-and-under earner has exploded. The rise of the gig economy (Uber, DoorDash et al.) has turned millions of would-be W-2 employees into small business owners. These filers now deal with Schedule C forms, 1099-K thresholds, home office deductions, and other arcana of our delightful revenue code. AI tools struggle here because the input data is often a mess of shoebox receipts and fragmented digital records. H&R Block’s 10,000 locations act as a physical interface for messy data, where a professional can ask, “Did you use your car for work?” in a way an LLM prompt might not catch.

H&R Block also deserves credit for product launches and its effort to integrate itself in to its customer base’s financial services stack. Through products like the Emerald Card and their newer banking app Spruce, they provide early access to refunds (incredibly popular) and year-round banking. This creates a stickier services ecosystem. Having physical locations staffed with financial professionals is a huge advantage for cash-focused folks as well.

Flat unit volume and 2.7% price isn’t exactly a recipe for great shareholder returns, but it’s a capital-light business. I might have guessed that customer growth would have perked up in light of DoorDash, Uber, and others, but Stripe/Shopify have funneled a lot of their businesses to TurboTax & others in the ecommerce game. HRB seems like they’ve mostly just stemmed the bleeding from TurboTax and DIY filers. It’s a services/capital return story, so the price you pay here is the most critical part. And the stock is cheap, so make sure you’re not wrong!