Founders negotiate price for weeks and write the return policy in ten minutes. Steven Capuano argues the second document decides more of the outcome than the first one does.
Ask a product founder what determines margin and the answer will involve unit cost, freight, and the number the retailer pushed them to. Steven Capuano’s answer is different. Margin is set by the cost of the units that come back, and the cost of the units that come back is set by three documents almost nobody drafts with any care: the warranty, the return policy, and the terms of sale.
Those documents are usually written in a hurry, borrowed from a competitor’s website, or assembled from a template the day before launch. They then govern every customer interaction for the life of the product, and they are extremely difficult to tighten later without looking like a company retreating from a promise.
A Written Warranty Is a Contract, Not a Marketing Line
The phrase lifetime guarantee reads like confidence. It functions like a liability with no end date. Under federal law, a written warranty on a consumer product above a modest price threshold has to be made available before purchase and has to be clearly designated as either full or limited, with its terms stated in plain language. The document is enforceable exactly as written, and ambiguity in it is generally read against the company that drafted it.
Capuano’s recommended discipline is to write the warranty as if it will be read aloud by someone trying to use it against the company. What specifically is covered. What is excluded. How long. What the customer has to do to make a claim, including whether proof of purchase is required. What the remedy is: repair, replacement, refund, or the company’s choice among them. A warranty that leaves the remedy unstated has effectively promised whichever remedy the customer prefers.
The Return Policy Is a Financial Model in Disguise
Return policy decisions look like customer service decisions and behave like pricing decisions. Who pays return shipping. Whether the window is thirty days or ninety. Whether opened product is accepted. Whether a restocking fee applies. Each of those lines has a number attached, and the sum of the numbers comes directly out of the margin negotiated so carefully elsewhere.
Capuano’s point is not that generous policies are wrong. Generous return terms often pay for themselves in conversion, particularly for a product a customer has never physically handled. His point is that the policy should be priced before it is published. A company that offers free returns on a bulky item without modeling reverse freight and refurbishment has not made a service decision. It has taken points off its margin and recorded them nowhere.
A return policy written in ten minutes will govern every transaction for the life of the product. Almost nothing else a founder writes that quickly carries that much weight.
The Documents Have to Agree With Each Other
The most common failure Capuano sees is not a bad policy but three documents that contradict each other. The website promises a full refund. The insert in the box describes a limited warranty with a restocking fee. The retailer applies its own return terms, which override both at the point of sale and often push the cost back to the vendor through a returns allowance or deduction schedule.
A customer experiencing three different policies will reasonably rely on the most favorable one, and a company arguing otherwise is arguing against its own packaging. The fix is mechanical rather than clever: one master document, with the website, the insert, the marketplace listing, and the vendor agreement all drawn from it and dated so the current version is never in dispute.
Returns Data Is a Document That Improves the Product
Beyond the legal function, the returns file is the most honest product feedback a company will ever receive. Every return carries a stated reason, and the pattern in those reasons is a specification document written by customers.
Capuano treats return coding as a product development input rather than a customer service metric. Returns clustering on a single component point at a manufacturing tolerance. Returns citing performance that did not match expectations usually point at the packaging copy rather than the product. Returns concentrated in one channel often point at a listing written by someone who never handled the item. A company that codes returns consistently and reviews them monthly learns faster than one running surveys, and the document costs nothing beyond the discipline of filling it in.
Write These Documents on a Good Day
The reason these documents are usually weak is timing. They get written during launch week, when attention is on inventory and advertising, or during a dispute, when the company is reacting to one angry customer and drafting a policy for everyone based on a single case.
Capuano’s advice is to write them on an ordinary day, well before they are needed, and to price them as deliberately as the product itself is priced. Model the expected return rate. Model reverse freight and the recovery value of returned units. Set the warranty period against the actual failure data rather than against what the competitor advertises. Then publish one version everywhere and review it on a fixed schedule.
The product determines whether a customer buys once. These documents determine what that first sale is actually worth, and whether the customer comes back.