Copy the whole field, including its zeros and case.
Review a 60-second asset-record exercise: exact fields, separate character rates, worked examples, editing and device limits. No hiring band or certificate.
This is a public guide to one 60-second record mode, not a score band or private input record. Actual results separately show final-character KPH, exact-field character KPH and field accuracy, without a hiring norm or a combined ability label.
Fictional structured records, not a preference quiz
This mode asks you to copy four source fields: an asset code, a label, a quantity written as text and a shelf code. Its 32 records are entirely fictional and remain identical when you switch the interface language. Use the displayed source rather than your own name, workplace data or real inventory. Reading a code, changing fields and checking a label are part of the timed task. The result does not assess a live inventory system or claim equivalence with a commercial employer assessment.
Attempting a record includes all four fields
Committing an edit in any field or explicitly choosing Next marks that record as attempted. All four fields then enter the accuracy denominator, including blanks. A record merely displayed without an edit or Next does not count, and later untouched records are excluded. If you type one correct code and leave its other three fields blank, field accuracy is 1/4, or 25%. If you start the clock and attempt nothing, accuracy is not applicable rather than zero percent. This rule prevents one completed field from representing a completed record.
A fresh example separates the rates
Use a fresh fictional record with AX-0042, Blue Tray, 007 and B-02. Its fields contain 7, 9, 3 and 4 characters, totalling 23. If all four are copied exactly within the fixed minute, the character-KPH figure is 23 × 60 = 1380 and field accuracy is 100%. Now suppose only Blue Tray becomes blue Tray. The final character count remains 23, but only 14 characters belong to wholly exact fields. Exact-field character KPH becomes 14 × 60 = 840 and field accuracy is 75%.
Text identity differs from numeric equivalence
The quantity 007 must stay 007; converting it to the number 7 changes the field. Codes preserve hyphens, spaces and case. Normalization to Unicode NFC handles equivalent character composition, but does not trim spaces, lowercase labels or remove punctuation. Whole-field matching gives no partial accuracy credit for a nearly correct code. The separate character rates must not be multiplied by field accuracy and presented as a hiring score: fields have different lengths, so that product is not the count of characters in exact fields.
Difference labels describe final strings
Incorrect fields use the first applicable description: blank, case or whitespace, one adjacent transposition, omission only, addition only, then other differences. For AX-0042, AX-0024 swaps the final two characters; AX-042 omits a character. BlueTray versus Blue Tray is classified as whitespace before the omission check. These categories are mutually exclusive reporting conventions, not observations of which key you pressed or why you made a mistake. Multiple edits can leave the same final string, and the tool does not reconstruct their history.
Corrections use the same minute
Backspace, replacement and returning to earlier records are allowed until the clock ends. Only final committed text contributes to output; deleted characters are not accumulated, while their editing time remains inside the minute. Paste and drop are blocked, but this is no proctored or fraud-proof assessment. Unfinished IME composition is excluded at the boundary. Leaving or backgrounding the page, or a large clock divergence, cancels the attempt. Completing all 32 records still requires waiting for the full minute, and the finite source can cap output.
Use a local practice record, not a certificate
The report gives no norm, percentile, employer pass line or certification. Character KPH extrapolates a minute to an hour and does not establish sustained performance or count physical keystrokes. A phone keyboard, a hardware keyboard and assisted input are different conditions; record those conditions when comparing your own attempts. Input stays in page memory, with no draft or server submission, and completion clears raw fields. Public-link sharing contains only the title and URL. Every result includes both complete guides, and repeating the fixed source may involve memory.