User Testing Report

Test brief
ACCESS: Public, no login, no gate. https://culdesac.site/ STATE: Live production. Known and already logged, do not re-report: the catalog renders client-side only, so there are no crawlable entry pages and no directory index. That is in… Show full briefHide full brief

ACCESS: Public, no login, no gate. https://culdesac.site/ STATE: Live production. Known and already logged, do not re-report: the catalog renders client-side only, so there are no crawlable entry pages and no directory index. That is in progress on a branch and is out of scope for this run. AUDIENCE: A first-time visitor who arrived from a link with no idea what this is and no particular intent. Not a fan of the small web, not someone looking for a directory. JOURNEY: You have 30 seconds on the homepage before you decide to stay or leave. Then press the button once, look at where you land, and come back. Tell me: what is this, who is it for, what would you do next, what stopped you, and whether you would press the button a second time and why or why not. OUT OF SCOPE: SEO strategy, monetization, what sites I should add to the catalog, what I should build next. Do not design the product. REPORT AS: steps / expected / actual. Classify every finding as BROKEN / MISSING / INACCESSIBLE / DISABLED. Rank by impact x confidence. Anything you could not reach, report as NOT TESTABLE with the blocker, not as a product deficiency.

Reviewer name

Steven

Date

Issues

1

[1 — High impact × high confidence] MISSING: The homepage does not establish a compelling payoff for visitors without small-web nostalgia

Steps: Arrive from an unexplained link and review the homepage for 30 seconds as a visitor with no interest in directories or the small web. Expected: Understand what this is, who it is for, and why trying an unknown destination would be worthwhile. Actual: The page explains the mechanism clearly as “One button” leading to a “hand-picked dead end of the internet,” with “No feed, no account,” but the benefit is framed as finding “the kind of place you used to find by accident.” I understood it as a curated random-link launcher for people seeking unpersonalized, nostalgic web serendipity, not as something aimed at a general visitor. My natural next step was to try the button once. The first result, “Salman Ansari,” was a public notebook of rough process notes. It worked, but it did not give me a reason to continue. As this audience, I would not press the button a second time because the first result felt arbitrary and the homepage had not established a broader payoff that outweighed another random attempt.

Watch at 0:05
2

[2 — High impact × high confidence] BROKEN: “take me somewhere” does not perform the action promised by the homepage

Steps: Read “It sends you to a hand-picked dead end of the internet,” then press “take me somewhere” once. Expected: Land on the selected external website, with the original page available to return to. Actual: After a brief “dialing” state, I remained on Cul-de-sac and saw an internal “Salman Ansari” preview with separate “go there ->” and “somewhere else” controls. Reaching the destination required another click, which then opened a new tab. The preview is useful context, but the primary journey requires two decisions while the copy promises one button that sends the visitor away.

Watch at 0:43
3

[3 — Medium impact × high confidence] BROKEN: Sharing repeatedly ends with “COULD NOT COPY”

Steps: Press “SHARE THIS ONE” on destination previews or “SEND THIS TO SOMEONE” in the footer. Expected: Copy a usable sharing link or open a working sharing method. Actual: The control changed to “COULD NOT COPY” without providing a link or fallback. This reproduced on several destinations and with the site-level sharing control on desktop. On mobile, “SHARE THIS ONE” for “Rent Check Me” produced the same visible failure, leaving no way to complete the sharing action.

Watch at 9:19

Session recording

Discuss the findings.

Share the report and compare notes with other product builders.

Join Discord