System image conveyed by word of mouth and peer demonstration can outweigh what the product presents on its own

What it is

A friend showing "just tap here twice, that's it," a coworker explaining "here's what this feature is actually for," a one-line tip in a community post — all of these are part of the system image, but they aren't optional add-ons to it. Very often a user starts operating a product with an understanding picked up secondhand, before ever reading the product's own copy or opening its help documentation — and that preformed understanding is frequently harder to displace than anything the product later says on its own.

Why it happens

A successful peer demonstration carries a kind of evidence the product's own copy can never supply on its own: the user watches someone they trust perform an action and reach their goal, which is more persuasive than reading "click here to…", because it comes bundled with proof that it actually works, rather than an assertion still waiting to be tested. Written instructions ask the user to believe first and verify later; a witnessed successful demonstration skips that first step and puts the verified result directly in front of them.

Timing matters too: word of mouth typically happens before, or right at, a user's first genuine encounter with the product, at a point when no model of the system exists yet in their head — whatever understanding the demonstration conveys becomes the first model to occupy that space. When the product's own explanation shows up later, the new information doesn't get to fill a blank slate; it has to first dislodge an existing understanding that's already in use and already been verified to work, which is inherently harder.

Where it stops holding

• This influence only applies where a user has actually been exposed to word of mouth or a peer demonstration. For users with no social channel available (early adopters, niche products, private-use contexts), this simply doesn't apply — the product's own system image faces no such competition for these users.
• What gets passed along secondhand isn't necessarily accurate. If a friend demonstrated behavior from an older version, or just showed off an idiosyncratic workaround they stumbled onto, this channel can just as easily convey an outdated or one-sided model — one that the product's current system image is actually trying to correct.
• This entry covers where the secondhand channel's advantage over the product's own presentation comes from — it does not concern whether the secondhand content itself is correct. Word of mouth can transmit an accurate model or a wrong one; the mechanism is the same either way.

Applying it

• For products expected to spread heavily through coworkers, friends, or community videos showing "how to use it," deliberately design the path most likely to get demonstrated so that it is, itself, the correct and worth-copying way to operate — rather than letting users independently invent a scatter of workarounds that then propagate by word of mouth.
• Keep visible correction points in the interface for behaviors especially prone to being skewed by secondhand demonstration — so that a user who enters a flow carrying a wrong expectation from watching someone else still sees a cue that contradicts that expectation at the key step, instead of the mistaken action sailing through unopposed.
• How to check: ask new users "how did you learn to do this," and tally how often the answer is some version of "someone showed me" or "I watched someone else do it." For features where that share is high, prioritize verifying whether the understanding circulating secondhand still matches current behavior — not just whether the product's own copy reads clearly.

Related

• Same group: A7.08.1 The system image is assembled from interface appearance, documentation, error messages, and word of mouth · A7.08.2 A self-contradictory system image produces inconsistent, even self-contradictory, user models · A7.08.3 Designers can directly control only the system image; the user's model can only be influenced indirectly
• Nearby: A7.13 Cross-product transfer of mental models · A7.14 Recognizing and correcting wrong mental models
• Search terms: system image · word of mouth · vicarious learning · peer demonstration