The Watson Weekly Weekend edition is sponsored by Avalara - the agentic AI platform automating global tax and compliance for leading eCommerce brands.

If you launch a regional marketplace on Monday, you have a nationwide tax problem by Friday.

That's roughly what happened to Alan Gaffney at Townchest. The pitch-deck version of his business is straightforward. Schools partner with manufacturers. Supporters buy products through a branded school storefront. Schools get funded. He launched it as a regional play because the schools were regional, so the business was regional, so the exposure was regional. That was the logic.

It didn't survive contact with the internet. A school in Ohio has alumni in Texas, grandparents in Florida, a cousin in Oregon who buys two t-shirts and accidentally triggers Nexus in a state Alan had never planned to operate in. Within a few weeks Townchest was a nationwide tax filer.

I had Alan on with Nick Boetcher from Avalara and Justin Samakow from Mirakl. Three things stuck with me after.

The first is the obvious one, and the reason it matters isn't that taxes are interesting. Sales tax obligations don't read your go-to-market deck. The minute you take one out-of-state order, your pilot stops being a pilot. Most operators don't know this until their accountant tells them, and by then they're back-filing in states they didn't know they were operating in.

Second, and this came up sharper from Nick than I'd have phrased it myself. Putting a compliance layer in before you grow is cheap. Retrofitting it after growth hits is expensive, and that's before back taxes and penalties. Founders defer compliance because it isn't sexy and isn't a feature. Then they spend Q3 of year two rebuilding the foundation while the business is operating on top of it. I've seen the bill on this more than once.

The third one I'm still chewing on. Townchest isn't using Mirakl as a checkout system and Avalara as a calculator. Mirakl's Catalog Transformer is normalizing supplier product data at a scale that would otherwise need a merchandising team. Avalara's Avi agent is assigning tax codes from a global library, which is work that used to require an actual tax department. If you're treating tools like these as line items in a SaaS budget, you're paying for headcount you didn't need to hire and not getting the benefit you bought.

Most marketplace operators I talk to know all of this in the abstract. They deprioritize it because audits feel theoretical and growth feels urgent. That trade looks rational right up until the moment it doesn't, which is usually about the same time a state revenue department starts asking questions.

If I were running a marketplace today, I'd want clean answers to three things. Where is our actual Nexus exposure, not the one in the plan. What are we deferring on compliance that's going to compound. And which of the tools we've licensed are we actually using as partners, versus the ones we pretend we are on a QBR slide.

Pull a list of every state you've shipped to in the last 12 months. Cross-reference against economic Nexus thresholds. If you're over in any state you're not registered in, you have a back-filing problem that's quietly getting more expensive. That's the whole exercise, and it takes an afternoon.

Most operators skip it because it sounds like the finance team's job. It isn't. It's a CEO question, because the bill lands on the P&L either way, and it lands bigger the longer you wait.