Chris Daly, Founder of I Want That! — 25 years in retail, worked with 40+ ecommerce brands.
Every retailer treats a percent-off sale as a one-time event: run it, move inventory, go back to normal. Sociologist Erving Goffman would tell you there is no such thing as going back to normal. In 1974, he published the book on how people actually interpret reality: not as raw fact, but through learned frames that stick around long after the moment that taught them. A 40% off banner is not a price. It is a lesson. And the customer does not unlearn a lesson just because you took the banner down. You did not give a discount. You retaught their memory of what your product is worth.
Retail did not set out to teach that lesson on purpose. Percent-off became the default language of ecommerce the same way any convention becomes default: gradually, through repetition, until nobody remembers deciding to use it. It started as a way to clear end-of-season inventory, a genuinely useful tool for a genuinely narrow problem. But as catalogs grew and competition for attention got louder, the discount stopped being an occasional event and became the primary way stores talked to customers at all. Black Friday trained an entire country to wait for the calendar date attached to the deepest cut. Cyber Monday moved that training online. Somewhere in that evolution, "percent off" stopped being a tool for clearing dead stock and became the default vocabulary retailers use to say anything is worth buying at all.
Here is what Goffman actually argued, and it applies to pricing more precisely than most retail theory ever has. People do not experience reality directly. They experience it through frames, learned structures that tell them how to interpret what is in front of them. He called the foundational layer the primary framework: a person's first, taken-for-granted read of a situation, built from direct experience. On top of that, he described a process called keying: taking something already meaningful under one frame and systematically recasting it as something else, using cues and conventions everyone recognizes.
His example was a boxing match versus a bar fight: identical physical actions, completely different frame, and the frame is what tells a bystander whether to applaud or call the police. A price works the same way. $34.99 is not information. It is a frame, and "40% off" is the key that tells a customer how to read it: fair, generous, or a trap, depending entirely on how consistently that key has been applied to them before.
This is why a discount is not a one-time event. The first time you key a price as "40% off," you are teaching the customer a new interpretation of a single transaction. The tenth time, that keying has become their primary framework, their new taken-for-granted baseline for what your product actually costs. You are not discounting from your price anymore. You are discounting from the price you taught them to expect, which is a number you handed them yourself.
Retail marketing's own academic literature backs this up independent of Goffman: Levy and Weitz's Retailing Management, a standard text in retail marketing curricula, covers reference pricing directly in its consumer buying behavior chapter, the idea that customers evaluate a price against an internal reference point rather than in isolation. Goffman explains the mechanism behind why that reference point forms and shifts. Levy and Weitz confirm retail marketing has known the reference point exists for decades.
This is exactly why two customers can look at the same pair of jeans and arrive at completely different fair prices, one at $85, one at $35. It is not that one customer is cheap and the other is not. Each is reading the product through a different frame of reference, built from a different history of prices they were taught to expect for that category. The $85 customer has never been keyed toward a discount frame for jeans. The $35 customer has been keyed toward almost nothing else.
Customer generated offers invert the entire mechanic. Instead of keying a price for the customer, telling them how to read it, you are asking the customer what frame they are already in. A CGO does not impose an interpretation. It surfaces one the customer already holds. And because you are not broadcasting a keyed discount to your entire file, you are not teaching every customer the same new baseline every time you want to move one segment of inventory. Some customers will offer close to full price because their frame of reference for the category was never keyed toward a discount in the first place. Others will offer 20 to 30 percent below list, and that is fine. You are meeting their existing frame instead of manufacturing a new one for people who never needed it.
What changes on your end starts with treating every percent-off decision as a teaching decision, not a marketing decision. The question is not just "will this move units?" It is "what baseline am I establishing for the next twelve months?" Second, use allowance-based pricing to meet a customer's existing frame instead of broadcasting one discount to everyone regardless of what frame they are already in. Price Builder's market adjustment and discount allowances exist for exactly this, letting the price flex per customer without retraining your whole file. Third, if you actually want to move a customer toward a higher frame of reference, upselling them into a better product, a better material, a better use case, do it by changing what they believe the category is worth, not by discounting your way there, because a discount teaches the opposite lesson from the one you are trying to establish.
We built Vector around this exact mechanic. A customer generated offer asks for a customer's frame instead of imposing one, so you are never teaching your entire file the same discount baseline just to move a segment of it. Price Builder's allowances let that same price flex to meet a customer where their frame of reference already sits, and when the goal is to shift someone into a new, higher frame entirely, that is an upsell conversation, not a markdown.