Seel sells certainty.

A small white box appears at checkout. Two dollars. One click. The internet's most boring revolution has a price tag, and Seel is the one collecting.

Somewhere in a Backcountry tab right now, a shopper is hovering over a final-sale ski jacket. The price is good. The fit is uncertain. Old internet rules say: close the tab. New internet rules - the ones Seel is writing - say: add the two-dollar refund window, buy the jacket, sleep on it. The merchant gets the sale. The shopper gets a soft landing. Seel gets the premium. Everybody, in the strictest accounting sense, wins.

This is the part of e-commerce nobody used to think about: the moment after the cart. Returns are the asterisk at the bottom of every retail deck. Last year roughly one in five online orders came back. The cost of that round trip - the shipping, the restocking, the discounting, the dignity of explaining it on the earnings call - is what Seel has decided to underwrite.

The problem they saw

Retailers have a tell. Watch them say "we offer free returns" with the same forced cheer you'd use to discuss your in-laws. Free returns are not free. They are a marketing line item, paid for in margin, absorbed quietly, blamed on Gen Z. The category had a $1 trillion gap and no one had built a financial product to plug it.

Zack Peng and Bill Liu noticed. Both were alumni of Affirm, the buy-now-pay-later company that turned the pre-purchase moment into a business. The unspoken question at Affirm was always: why does the financial layer stop at the checkout button? Why does everything after - delivery, returns, defects, regret - get treated like a customer-service problem and not what it actually is, which is a risk-pricing problem?

The founders' bet

In 2019 Peng and Liu started building. The first product was prosaically named Return Assurance. The pitch was almost too simple: at checkout, the shopper sees a small add-on - call it a few dollars, call it the price of a coffee - and in exchange Seel takes on the financial liability of the return. The merchant offloads the risk. The shopper unlocks otherwise-final-sale inventory. The transaction goes through.

The interesting part is what's underneath. Seel's underwriting engine ingests hundreds of signals at the moment of order placement - product category, price band, shopper history, regional return rates, the time of night you're shopping (yes, that one matters). It produces a probability and a price. If you've ever wondered why the protection plan costs $1.99 on one order and $14 on another, that's the model talking.

The Affirm team had already proved that consumers will accept a financial product at checkout if it solves a real anxiety. Seel was betting the same was true for the back half of the journey. Hindsight makes this look obvious. In 2019 it sounded like extended warranty for sweatpants - which, to be fair, is one accurate description of the company.

The product, in plain English

Seel ships as a Shopify app and an API. Merchants flip a toggle. A widget appears next to the buy button. Shoppers see something like, "Add Worry-Free Purchase - get a 7-day refund window even on final-sale items." The cost is consumer-paid. The integration takes an afternoon. The pitch to the retailer is unromantic and exact: zero downside, measurable upside.

The product line has since stretched into Worry-Free Delivery (for lost or damaged shipments), extended warranties, and an AI agent layer that handles the resolution flow so a human at Garmentory does not have to. The merchant dashboard is the unsexy work of insurance: claims, refunds, dispute trails, audit logs. None of it photographs well. All of it matters.

The proof

Numbers, finally. Across Seel's merchant base, the Return Assurance widget converts at roughly 24 percent. The industry average for B2C checkout add-ons sits around three. That is not a rounding error; that is a category. Merchants report a five percent lift in overall conversion after enabling the widget - which, in e-commerce, is the kind of number you triple-check before you put it on a slide.

The customer roster reads like a Slack channel of the brands you've actually bought from. Backcountry, the outdoor giant. Garmentory, the indie-fashion marketplace. SidelineSwap, where used hockey skates change hands. CommentSold and ShopSimon, both at the live-shopping edge. None of them advertise that they run on Seel - which is exactly the point. The best infrastructure disappears.

The mission, restated

Ask Peng what Seel is and he will not say "an insurance company," even though, technically, that is what the regulators have decided. He will say something closer to "post-purchase infrastructure" - a phrase that sounds boring on purpose. Plumbing is supposed to be boring. The boring is the moat.

The deeper claim is that the post-purchase moment is its own financial category, distinct from payments, distinct from logistics, and previously homeless. Returns. Damages. Lost packages. Buyer's remorse priced as a probability rather than a customer-service ticket. Seel wants to be what Stripe was to the payment field, or Affirm was to the credit one: the company that named the thing.

Why it matters tomorrow

There is a deeper reason this category exists now. The shape of online shopping has changed - faster, more impulsive, more cross-border, more final-sale. Inventory turns faster. Tolerance for friction is zero. A 2025 forecast puts U.S. retail returns at roughly $1 trillion, with online returns at 19 percent of orders. The cost of doing nothing has become louder than the cost of doing something.

The optimistic version of Seel's future is that every checkout widget eventually carries a Worry-Free-something on it, the way every checkout button now carries a Pay-with-something on it. The pessimistic version is that one of the giants - a Shopify, a Klarna, an Amazon - decides this category is too good to leave to a startup and builds it in-house. Seel's bet, the same one its founders made in 2019, is that an underwriting model trained on a decade of return data is harder to clone than a checkout widget is to copy.

Either way, that final-sale jacket is going in the cart.

The shopper hovering over Backcountry, the one from the top of this article - she clicks. The widget loads. Two dollars and forty-nine cents are added to her order. She doesn't think about it. The merchant ships the jacket. Seel files the policy. Somewhere a data point about her becomes one of the hundreds of signals feeding the next shopper's price. The internet's most boring revolution proceeds. Nobody notices, which is exactly how Seel wants it.