People often read starter content as personalized output

What it is

Starter content is often written as if it already thought about “you”: second person, industry nouns, charts that look like analysis. People read it as a result the system generated from their situation, not as the same seed everyone gets. The cost is over-trust (deciding on fake numbers) or over-disappointment (feeling tricked when everyone has the same file). This entry is not how seeds and handmade items look different in a list. It is what happens when a seed is taken for personalized intelligence.

Why it happens

Generative features and setup wizards have trained “for you” as the default expectation: after a company name and an industry, the document that appears is attributed to those inputs. Starter templates often arrive at the same beat—three questions, then “your weekly report” is on the desk—and attribution folds the template into generation. Second person and apparently private data (“Alex’s conversion rate”) are strong cues; people do not first think “Alex is a placeholder.” Once it is generation, motive to edit drops: changing “what the system gave me” feels less legitimate than changing a draft; meanwhile fake metrics may be forwarded. Breaking the attribution needs two sentences on the object: this is a template; the names and numbers are not yours. “Template” alone is not enough, because a template can still be “filled with your data.”

Where it stops holding

Objects actually generated from user data (a real report after connecting a warehouse) must be protected from the opposite error: mark source and time so people do not treat a true result with the suspicion due a seed. Structure-only starters with no copy and no numbers are rarely taken as generated. If a marketing landing “sample report” looks like the post-login seed, attribution rides from the door into the workspace. In high-trust settings (health, money) the harm is larger; starters should avoid numbers that look like personal results.

Applying it

• Do not title starter copy “your report is ready”; use “this is an editable template; the numbers are placeholders.”
• Use obviously fake values for placeholder names, metrics, and dates (a sample company, zeros, or a “sample” mark)—not plausible conversion rates and real-looking names.
• If preferences were just collected, say “this does not yet use what you entered,” or actually use them and name which ones. Do not collect, ignore, and still speak in “for you.”
• Verify by asking someone who just saw the seed whether the numbers were computed from their situation. If yes, strip person, metrics, and generative tone and ask again.

Related

• Within the group: H2.08.1 Starter content is an editable start; samples stay out of the record · H2.08.2 Starter templates must look unlike user-created work · H2.08.4 Starter content must resemble real work or it fails as a model
• Adjacent: H1.09 Smart prefill · H2.06 Sample Data · H4.02 Purpose explanation
• Search terms: fake personalization · template as generated · placeholder data