A pricing page looks like a list of independent offers, each standing on its own merits. Buyers do not read it that way. They read positions — cheapest, middle, most expensive — and the position does more work than the number printed underneath it. Work on extremeness aversion by Itamar Simonson and Amos Tversky found that people shy away from the ends of a presented range and settle toward the centre, not because the centre is objectively right for them but because it is defensible. The middle option needs no justification to a boss, a spouse, or oneself. So the tier sitting in the middle of your menu is your real price, and the others exist mostly to decide where the middle is.

Picture a tools company with three plans priced at roughly a coffee, a lunch, and a dinner out. Most of its revenue lands on the lunch plan. The team notices that almost nobody buys the dinner plan and removes it, reasoning that a tier with no customers costs nothing to cut. Demand does not redistribute to the survivors. It slides toward the coffee plan, because the coffee plan has stopped being the bottom rung of a three-step ladder and become one half of a two-way choice, while the lunch plan has inherited the position of the expensive end. The company deleted the tier it was not selling and lost money on the tier it was.

The same geometry runs in the other direction. A genuinely premium tier placed above your current top plan does not need anyone to buy it; it only needs to exist, because its existence demotes your old top plan into the middle and hands it the compromise position. That is why one of the cheapest ways to lift average revenue is to publish an option you expect almost no one to choose. It also explains a failure that looks like mispricing and is not: when the tier you least want to sell is the one sitting in the middle, the problem is the order, not the numbers.