Before the timer starts
Write a one-sentence browsing mission. Include the product type, intended use, and one constraint. For example: “I need a lightweight everyday jacket with visible measurements.” This sentence becomes the filter for every row you open.
A brand can narrow style, but it does not explain use, fit, budget, evidence needs, or likely shipping trade-offs.
The five-step, 15-minute routine
Minutes 0–2: lock the category
Open one category only. Remove anything that does not match the browsing mission, even if it looks interesting. You can return later; keeping it now weakens the comparison.
Minutes 2–5: remove rows with basic gaps
Discard rows with a mismatched title and link, unusable images, no sizing for a fit-sensitive item, or an external page that no longer identifies the same product. Do not spend the whole session rescuing one weak lead.
Minutes 5–9: keep three comparable candidates
Three options are usually enough to expose meaningful differences without creating decision fatigue. If you have six, split them by use case and choose which use case matters today.
Minutes 9–12: record evidence and uncertainty
For each candidate, write one reason to keep it and one unanswered question. “Looks good” is not a reason. “Shows item-level measurements and clearer closure photos” is.
Minutes 12–15: choose the next action
Each candidate gets one status: compare now, research one missing detail, or remove. Do not use “save for later” as a fourth status unless you can name exactly what later work is required.
Use the same evidence test for all three
| Check | Useful signal | Reason to pause |
|---|---|---|
| Identity | Title, category, images, and destination agree | The row and destination describe different items |
| Photos | Angles answer category-specific questions | Repeated promotional views hide construction or scale |
| Fit or size | Measurements are tied to a visible option | Only generic sizing language is shown |
| Value | Price is compared with similar candidates | The lowest number is treated as proof |
| Delivery context | Known weight or an honest unknown is recorded | Bulky packaging is ignored |
Copy this shortlist note
A small, consistent note is more useful than saving screenshots with no context.
Product type:
Intended use:
Evidence that helps:
Important detail still missing:
Likely packed-weight concern:
Current destination checked on:
Decision: compare / research / removeKnow when to stop browsing
More options help only when they add a new, relevant trade-off. Stop opening links when the next item is merely another version of what you already have, when you can no longer remember why a row was saved, or when missing evidence is repeated across the category.
Once you have three comparable candidates, write one advantage and one uncertainty for each. You now have enough to check the missing details without reopening broad discovery.
Resolve one missing fact at a time
After the shortlist, resist reopening broad discovery. List the unanswered questions from most likely to change your decision to least. A missing size measurement may remove a candidate immediately; a different promotional photo probably will not. Start with the fact that could remove a candidate or change your preferred option.
Use one query or destination check for that fact, then update all three candidates before moving on. This keeps the comparison fair. If only one row receives extra research, it may look stronger simply because it has more notes.
If a required fact cannot be verified, write “unknown” rather than filling the gap with a guess. An honest missing field is easier to compare than a confident assumption. When the same required fact is missing from every candidate, pause the category and decide whether another source or product type would be more efficient.
Confirm identity, then fit or compatibility, then condition or construction, and finally price and shipping context. Stop as soon as a candidate fails something you require.
What to do after the shortlist
Use the seven-point checklist for the final comparison. If a candidate still lacks category context, return to the category directory. Continue to Findsindex only when you know what you are trying to confirm.
