The number you pre-fill is the number you'll get back.

Ask a room whether the tallest redwood is more or less than 1,200 feet, then ask them to guess its actual height, and their guesses cluster high. Ask a different room whether it’s more or less than 180 feet, and their guesses cluster low. Same tree. The only thing that changed was a number someone said out loud a moment earlier — a number everyone knew was arbitrary.

That’s anchoring, and Tversky and Kahneman pinned it down in 1974 with a rigged wheel of fortune. They spun a wheel that visibly landed on 10 or 65 — plainly random — then asked people what percentage of countries in the UN were African. People who saw 10 guessed around 25%. People who saw 65 guessed around 45%. A number with no relationship to the question dragged the answer toward itself. The mechanism they proposed: you start from whatever value is in front of you and adjust — and you almost always stop adjusting too soon, still inside the anchor’s gravity.

We mostly talk about anchoring in pricing — show the $99 plan first and $29 feels like a bargain. That’s real, and worth its own piece. But pricing is the least of it, because every number you pre-fill in your product is an anchor. The default in the slider. The quantity in the stepper. The “recommended” storage tier. The suggested tip. The goal field in a fitness app that opens at “10,000 steps.” The number of seats the form loads with. None of those feel like persuasion. They feel like helpful starting points. They are both.

Watch what this does. A budgeting app that opens the savings-goal field at $50 and one that opens it at $500 will produce different goals from the same users — not because anyone was fooled, but because adjusting away from a starting point is effortful and people stop when it feels close enough. A usage estimator pre-filled with “50,000 API calls” teaches the user what a normal customer looks like before they’ve thought about their own numbers. The anchor doesn’t override judgment; it seeds it.

Which is exactly why this one deserves a conscience. An anchor is a decision you’re making on the user’s behalf about where their thinking should begin — and you can point it toward their interest or toward your invoice. The same slider default that helps a confused user land on a sane configuration can, nudged upward, quietly inflate what they commit to. The test I’d hold myself to: if the user saw the number I pre-filled and understood it was shaping their answer, would they thank me for a good-faith starting point, or feel worked?

The useful reframe is that you can’t not anchor. A blank number field is its own anchor — it anchors on zero, or on whatever the user last typed somewhere else, and it dumps the whole cognitive load of “what’s reasonable here?” onto them. So the move isn’t to strip the numbers out. It’s to choose them like they matter, because they do: set the default to the value you’d genuinely recommend to a friend who asked, label it as a recommendation so the anchor is visible rather than smuggled, and make adjusting away from it cheap and obvious. An honest anchor is one of the kindest things you can hand a user staring at an empty field — and one of the easiest to abuse.