Compare evidence side by side, not memory against memory.
Open only two candidates from the same product category. Give each one the same five checks: item match, photo coverage, measurement clarity, source-page health, and cost context. Keep a row only when it wins for a reason you can name. If neither row answers the basic questions, drop both.
The winner is not the row with the most attractive thumbnail. It is the row that leaves fewer important questions unanswered.
Minute 0–1: make the comparison fair
Before looking closely, confirm that the candidates are actually comparable. A lightweight hoodie and a lined jacket cannot share the same fit or shipping assumptions. A shoulder bag and a rigid suitcase may both appear under “bags,” but the useful dimensions and packing risks are different.
- Both rows describe the same narrow product type.
- You are comparing the same intended size or option.
- You know which three details could change the decision.
- You have no more than two rows open.
If these conditions are not true, pause and return to the category guide.
Minute 1–7: run five shared checks
| Check | Question for both rows | Useful note |
|---|---|---|
| 1. Item match | Does the destination clearly match the row, thumbnail, category, and intended option? | Match / uncertain / mismatch |
| 2. Photo coverage | Do the images show the category views needed for this decision? | Name the strongest and missing view |
| 3. Measurements | Are the relevant dimensions shown, explained, and tied to the option? | Supported / typed only / absent |
| 4. Source health | Is the destination usable, specific, and consistent rather than redirected or stale? | Direct / broad / changed |
| 5. Cost context | Does the apparent value survive likely weight, packing, and missing-information risk? | Low / medium / high uncertainty |
Do not invent a numerical score unless it helps you. Plain words often expose the difference more honestly.
Minute 7–8: write two sentences
Candidate A
“The photos match the black zip jacket and show the lining, closure, chest, and length. Packed weight is uncertain, but the source page and measurements are usable.”
Candidate B
“The price is lower, but the album mixes two jackets, the size is typed without a measurement photo, and the destination does not identify the option clearly.”
Once both sentences exist, the stronger research candidate is usually obvious. This does not verify Candidate A; it explains why A deserves the next check and B does not.
Minute 8–9: use a tie-breaker that matters
If the rows remain close, choose one decision-changing question. Good tie-breakers include:
- Which row better supports the exact size or option?
- Which photo set includes the view most likely to reveal a problem?
- Which destination page is easier to identify and revisit?
- Which row has a more believable packed-weight range?
- Which row requires fewer assumptions to explain?
Popularity, a polished thumbnail, or a dramatic “best quality” label is not a useful tie-breaker.
Minute 9–10: close something
A comparison is unfinished while every tab remains open. Choose one outcome:
Keep one
Save the better-supported row and attach the sentence that explains why it survived.
Hold one
Use this only when one specific missing fact could reasonably change the result. Name that fact.
Drop one
Close the weaker row when it adds uncertainty without adding useful evidence.
Drop both
If neither row passes the basic item, photo, measurement, or source checks, widening the search is better than choosing by default.
Why the cheaper row may not be the better shortlist row
A lower displayed price can be meaningful, but only after the option, item, and likely parcel are comparable. A cheap row with missing measurements may create more fit risk. A rigid package or heavier option may change later cost. A broad album may require extra time just to identify the product.
Price should break a tie between similarly supported rows—not erase the evidence gap between them.
A note template you can reuse
Compared: [Row A] vs [Row B]. Same category/option: yes / no. Winner: [row]. Reason: better support for [photos / measurements / source / weight]. Remaining uncertainty: [one item]. Next action: keep / hold / drop.
Keep the note shorter than the research. If you need a paragraph to defend a weak row, that is useful information.