Guide
Remove identifying details and put one comment on each line, labelled F1, F2 and so on. Count only the pasted sample. Repeated comments may come from one person; frequency alone cannot show how widespread or serious a problem is.
Your worksheet stays in this page and is lost when you leave or reload. Copy it before leaving. Only guide completion is saved. Do not enter sensitive information.
Use the fictional export comments below to prepare an investigation brief. It should show what was said, what is repeated, and what still needs checking. You can group the items yourself or ask one approved assistant. For your own material, remove contact details, account IDs and identifying context. Four comments do not become four customers just because a table has four rows.
Use the fictional export comments below to prepare an investigation brief. It should show what was said, what is repeated, and what still needs checking. You can group the items yourself or ask one approved assistant. For your own material, remove contact details, account IDs and identifying context. Four comments do not become four customers just because a table has four rows.
Fictional feedback sample — not real user research. [F1] Export fails when I select all rows. [F2] Please let me choose export columns. [F3] Export fails when I select all rows. [F4] The CSV download is useful, but I cannot tell whether hidden rows are included. F1 and F3 have exactly the same wording. Respondent identities, collection period, product version, devices, error messages, file size and reproduction results are unknown. Keep both labels; a duplicate may be a repeated submission or different people using the same words. The sample cannot tell us which. Do not attach guessed identities or combine these comments with private account data.
Use only the supplied feedback and preserve each message label and exact wording. Group into reported defects, feature requests, positive observations and clarity questions. Mixed comments may appear under multiple themes with the same label. Report message counts separately from distinct wording and respondent counts that the source does not establish; do not count theme memberships as different people. Flag actual exact-text duplicates without guessing identities. In the fictional F1–F4 example F1/F3 are duplicates; do not assume those labels or duplicate pairs in other input. Give themes, evidence limits, a proportionate proposed follow-up for reported problems and neutral questions for remaining gaps. Do not invent a failure when none is reported, root cause, business impact, user-base frequency or completed tests. Keep proposed investigation order separate from severity and roadmap priority. Feedback:
Draft investigation brief — fictional sample Sample accounting 4 labelled messages; 3 distinct wordings; number of respondents unknown. F1/F3 are exact text duplicates. The themes below contain 5 label memberships because F4 appears twice; that is not 5 messages or 5 people. 1. Reported defect: export after selecting all rows Evidence: F1 and F3, both “Export fails when I select all rows.” Count: 2 messages, 1 distinct wording. Reproduction status: not tested. Error, affected versions, frequency and root cause: unknown. 2. Feature request: choose columns Evidence: F2, “Please let me choose export columns.” Count: 1 message. This is a request, not proof the current product cannot select columns. The desired columns and current workaround are unknown. 3. Positive observation: CSV usefulness Evidence: F4, “The CSV download is useful, but I cannot tell whether hidden rows are included.” Count: 1 message. It expresses usefulness alongside uncertainty; it is not a satisfaction score. 4. Clarity question: hidden-row scope Evidence: the same F4. Count: 1 message already included above. It does not establish whether hidden rows are exported or omitted. Proposed first investigation — not performed Clarify the select-all failure because it may block finishing an export, not because this sample proves it is the highest-priority issue. Ask for permitted reproduction steps, product/browser version, what “all rows” selects, expected output and the exact non-sensitive error. Use a small synthetic table with known rows/columns. Compare a single-row export with select-all while keeping settings unchanged. Record the actual file/error and count rows and columns in any CSV. A successful download alone is not proof that all expected data arrived. If no reproduction is available, retain “reported, not reproduced”. Do not replace a database or promise a fix from these comments. Neutral follow-up questions — not sent F2: which columns do you need to include or exclude, and how do you handle that today? F4: when rows are hidden or filtered, which rows do you expect in the file? What did the interface lead you to expect? Sample owner: how were these comments collected, over what period, and can duplicates be checked without attaching personal identifiers? Repair exercise Wrong: “Half our users hit a database bug, so column selection is the top priority.” Repair: “Two of four supplied messages repeat the same export-failure wording. User count, root cause and prevalence are unknown. Column selection is a separate request in F2. Investigate the failure and clarify the requests before a prioritisation decision.” Keep the four original labels with the brief. A neat theme name should make a comment easier to find, not make its uncertainty disappear.
Check F1–F4 all appear and their wording is unchanged. Count four messages, three distinct wordings and five theme memberships; respondent count must remain unknown. Keep F4’s positive and uncertain parts together when quoting it. Does any sentence claim a tested cause, affected-user percentage, severity or priority that the sample cannot support? Rewrite the incorrect conclusion in your own draft. Choose one next investigation and explain the information it would add. Record it as proposed until someone actually does it. If a later test contradicts the comments, keep both the original report and the dated observation instead of silently deleting either. Tick the checks only after revising; they do not validate a bug or approve a roadmap.
Copy the reviewed brief, original labels and unanswered questions before leaving; edited worksheet text is not saved here. Use the PRD-draft Task below only for a requirement supported by clarified evidence, keeping it proposed until reviewed. Use roadmap priority to compare it with a stated objective and other work; message count alone is not the decision rule. No follow-up is sent by AhaDo. Completion here means a bounded investigation brief is ready, not that a defect was reproduced, a requirement approved or a fix delivered.
Saved for this browser session. Sign in keeps it attached to you on this device.
Editorial workflow with a synthetic example. No measured user outcome or provider performance claim.
Point out a missing step or an unsafe claim. Your note is saved in your workspace for review.
Other products from our team
DictoGo AI dictionary with offline lookup, native examples and an AI speaking coach
DictoBoard An iOS keyboard that shows bilingual definitions as you type
CS Keyboard Save frequent replies once and send them with one tap
Contrack WhatsApp attribution for TikTok Ads
Inboxto Anonymous email addresses that keep spam out of your real inbox
Phonics Joy Master English phonics in 5 minutes a day
Math Joy Fun arithmetic practice for kids
GroomGain Revenue clarity for independent pet-grooming salons
Authloom A local-first 2FA authenticator with encrypted recovery you control
Skillstore Security-audited skills for Claude, Codex and Claude Code
Flicklish Turn videos you choose into short English lessons with subtitles and an AI Coach