Builder's Brain
Builder's Brain · the neuroscience of building · ◉ Evergreen

Your users are quietly outsourcing their memory to you

by Shreyansh Ojha·4 min·Working Theory

Ask someone what your product does and they’ll get it roughly right. Ask them where the thing they need lives inside it, and watch them freeze. That freeze isn’t a failure of your users, and it isn’t really a failure of your onboarding. It’s a feature of how memory works — and once you see it, a whole class of build decisions flips.

In 2011 Betsy Sparrow, Jenny Liu, and Daniel Wegner ran a set of experiments that landed uncomfortably. They had people type facts into a computer. Some were told the computer would save what they typed; others were told it would be erased. The people who believed the machine was keeping the information remembered the facts worse — but remembered where the facts were stored better. Told that a thing is retrievable, the brain quietly declines to store the thing itself and stores the pointer to it instead. The nickname stuck: the Google effect. We don’t remember the capital fact; we remember that it’s a search away.

It sits inside an older idea of Wegner’s called transactive memory. Couples, teams, families don’t each hold the whole picture — they hold who knows what. You don’t remember the plumber’s number; you remember that your partner does. The group’s memory is bigger than any head in it, because the storage is distributed and everyone mostly remembers the index. Offloading like this isn’t laziness. It’s efficient. Working memory is famously tiny — a handful of items at once — and every fact you don’t have to hold is capacity freed for the task actually in front of you.

Here’s the part that should change how you build: your product has become one of the partners in your user’s transactive memory. They are not memorizing your features. They’re memorizing that you’re the place where that thing lives — and often not even that, just that it’s “in there somewhere.” You are the external drive. Which means most of the recall you’re quietly counting on is never going to happen.

USER holds the pointer "where to look" PRODUCT holds the content "what it actually is" look-up
The user keeps the index; your product keeps the content. Design for the look-up, not the recall. Original diagram · Working Theory

Three things follow.

Design for retrieval, not recall. Recall is producing the answer from nothing (“what was that feature called?”). Retrieval is recognizing it when it’s put in front of you. The brain is far better at the second. So the search box, the “recently used” row, the “jump back in where you left off,” the surfaced-not-buried command — these aren’t conveniences, they’re the actual interface to a memory system that has decided not to memorize you. Every time you make a user recall a path, you’re betting against how their memory chose to work.

Stop treating onboarding as a lecture. A five-screen tour asks the user to load facts into a store they’ve already decided to keep empty, because they can tell the facts are retrievable later. That’s why tours evaporate. The move that survives is making the product itself be the memory: the tooltip that appears in the place and moment it’s needed, the label that says what this is instead of assuming you were told once, the “your stuff is here” that answers the where-question at the exact time it’s asked. Don’t make them remember. Be the thing they remember with.

Guard the memory like it’s theirs — because it is. The flip side of being someone’s external memory is the flip side of being someone’s memory at all: if you lose it, move it, or hide it, they experience something close to amnesia. A search that used to work and now returns nothing, a control that quietly relocated, an export that mangles their data — these don’t feel like UI changes. They feel like a partner who forgot. Stability of where things live is not a cosmetic nicety; it’s the contract of a memory system. Break it and you’ve broken the one job your user actually assigned you.

The science, to look up: the 'Google effect' on memory — Betsy Sparrow, Jenny Liu & Daniel Wegner (2011); transactive memory — Daniel Wegner; cognitive offloading more broadly. Neighboring ideas: working-memory capacity (~4 items) and the endowment of user-created data.

Liked this? Get the next one in Working Theory.

Going weekly in August (it's in beta now). One genuinely interesting read on building, the brain, and the science most people missed.

Got a reaction, a counter-example, or something I missed? Reply by email — I read everything.
◉ join in

Where have you hit this — in a product you use, or one you're building?

Threads open here soon. For now, the conversation lives two clicks away — discuss on GitHub, or just reply by email. I read and answer everything.