Acceptance Checklist: Run Through This Before You Sign Off
For AI-delivered UI and images, walk the list item by item: hierarchy, spacing, restraint, consistency, detail. Interactive checklist you can tick, plus further reading
THE QUESTION THIS PAGE ANSWERS
ANSWER FIRSTWhat is the key idea behind “Acceptance Checklist: Run Through This Before You Sign Off”?
For AI-delivered UI and images, walk the list item by item: hierarchy, spacing, restraint, consistency, detail. Interactive checklist you can tick, plus further reading
Turn taste into a behavior the product can repeat. The useful outcome is not a nice opinion. It is a visible rule, a small example, and a way to tell when the experience falls below the bar.
Capture one before-and-after example that shows the quality bar without extra explanation.
Polish that improves the surface while leaving the user's uncertainty untouched.
AI can ship a draft in ten minutes; you spend three minutes accepting it—that math always wins. The stubborn enemy is 'good enough': the draft looks fine at first glance, so your hand wants to click through, and the rough spots only sting after launch. This list reins in that hand: fifteen items, one action each—run through them before you sign off. Check progress lives locally; come back with a new draft and keep going. Several criteria have sources—after you check them, scroll down; each comes with the rationale and a practice drill.
The new restraint item has a source. From About Face 4, Chapter 17: every element needs a reason to exist, and every difference needs a reason too—if you can't give one, delete the element or flatten the difference. The paired move is the reduction test: strip the suspect element and see whether information took a hit. The book goes harder: keep cutting until the design breaks, then put the last piece back. Try one drill.
The detail-group response-time criterion has a source too. How comfortable a UI feels depends heavily on response speed, and the thresholds are well known. About Face 4, Chapter 17 cites Nielsen's split: under 0.1s feels instantaneous; finish within 1s and thought stays unbroken—a subtle cue is enough; near 10s you must show it's running, like a spinner plus an estimate; past 10s explain what's slow, give progress updates, and ping when done. Three buttons, one threshold each—tap them yourself and feel it.
Data viz is in scope for acceptance too. Tufte's principle: quantify what can be quantified—charts draw the trend, and the numbers themselves must also be present. About Face 4, Chapter 17 cites Windows disk properties: the pie gives a rough impression, and used/free bytes are still listed. AI-generated report pages love this mistake: a pretty curve, zero numbers up close, so you dig raw data again when reporting. Two weekly cards—tap the better one.
In the opener you voted by gut among three plans. After 12 lessons, same question, same three plans in miniature—vote again.
Opener: vote by gut first. Lessons 2–5 break "looks good" into four handles: hierarchy, spacing, color, detail. Facing a page: find the one hero, measure spacing, count colors, then pick at the details.
Lessons 6–9 write the variables you see into the chat: swap "make it premium" for primary color, radius, density—words AI can execute. Name the variable when revising so AI knows where to push.
Lesson 10 tears down references in a fixed order; lesson 11 feeds references with reference images, description, and variables. This page's 15-item list guards the exit: run through before you sign off.
You can see: hierarchy, spacing, color, detail—four handles, tear down in order; stop tourist gawking.
You can say it: translate feel into variables and references for AI; leave adjectives for social captions.
You can accept: the 15-item list guards the exit. Next AI handoff, run the list first—send back the failed item in its own words.
Criteria hold only with sources: existence needs a reason, response clears three thresholds, quantify what you can—sources are in further reading, so you can answer when asked.
Source: Original to Xiaoshan Academy's Taste Engineering series; some design principles adapted from About Face 4, Chapter 17 (Alan Cooper et al.); response-time thresholds cited there from Nielsen; quantification from Tufte.
Turn the feeling in “Three minutes of acceptance always pays off” into a judgment
“AI can ship a draft in ten minutes;” points out that AI has lowered the bar for making something usable. The skill readers need is noticing what is wrong and turning that feeling into an actionable requirement.
Watch the user's next action, not just the surface
Turn “The new restraint item has a source.” into observable questions: does the user know what happened, what to do next, and how to recover from an empty or failed state? Does the hierarchy make the important information visible first?
Pretty is not the same as usable
Apply “Criteria hold only with sources: existence needs a reason, response clears three thresholds, quantify what you can—sources are in further reading, so you can answer when asked” to a second screen or flow. Record one moment of hesitation and the user action after the change; observable behavior is stronger evidence than polish alone.
From “Three minutes of acceptance always pays off” to “Try it · Acceptance checklist: walk all 15”
“Three minutes of acceptance always pays off” grounds the problem in “AI can ship a draft in ten minutes; you spend three minutes accepting it—that math always wins. The stubborn enemy is 'good enough': the draft looks fine at first glance, so your hand wants to click through, an…”. “Try it · Acceptance checklist: walk all 15” then moves it toward “Handoff acceptance checklist 0 / 15 Hierarchy · 3 items Name the one hero on this screen—if you can't, send it back Squint at the whole page—does the first glance land on the hero? Besides the primary button, i…”. Together, they show that the lesson is not just a conclusion to remember, but a claim with conditions.
Carry the judgment into the next situation
For experience work, turn abstract impressions into user actions: did the person understand the state, find the next step, recover from an error, and want to continue?
- “Three minutes of acceptance always pays off”: AI can ship a draft in ten minutes; you spend three minutes accepting it—that math always wins. The stubborn enemy is 'good enough': the draft looks fine at first glance, so your hand wants to click through, an…
- “Try it · Acceptance checklist: walk all 15”: Handoff acceptance checklist 0 / 15 Hierarchy · 3 items Name the one hero on this screen—if you can't, send it back Squint at the whole page—does the first glance land on the hero? Besides the primary button, i…
- “The closing point”: Lesson 10 tears down references in a fixed order; lesson 11 feeds references with reference images, description, and variables . This page's 15-item list guards the exit: run through before you sign off
The final “The closing point” brings the discussion to “Lesson 10 tears down references in a fixed order; lesson 11 feeds references with reference images, description, and variables . This page's 15-item list guards the exit: run through before you sign off”. The useful thing to carry forward is knowing which judgments must be revisited when input, scale, or risk changes.
I turned one judgment from this article into a small experiment I could run today. Knowing what to observe next is more useful than simply remembering the conclusion.
After reading this, I first looked for the conditions behind the idea instead of copying the method into a project. That order made the later trade-offs much clearer.
When this judgment reaches real work, which constraint should be added first? I am curious which step matters most between reading and the first practical attempt.
No discussion on this article yet.