Before the first search

Write a three-line product brief

A narrow brief protects the shortlist from trend words and attractive thumbnails. Write the object, the decision-changing detail, and the reason you would reject it.

Object

Midweight zip hoodie for cool evenings.

Must prove

Chest and body length must match a hoodie I already wear.

Stop condition

No usable measurements, unclear material, or photos that hide the zip and cuffs.

The brief does not need to be perfect. It only needs to make weak candidates easier to notice.

Use two passes instead of one endless search

  1. Discovery pass. Collect no more than five candidates from the same category. At this stage, record only the link, visible option, approximate price, and one reason it might fit the brief.
  2. Decision pass. Reopen those candidates side by side. Add measurements, photo coverage, source clarity, weight context, and the risk that still needs checking.

Separating discovery from judgment prevents every new result from resetting the comparison.

Give every surviving row four decision notes

Reason saved

The concrete advantage: clearer measurements, better photo coverage, a useful option, or lower package bulk.

Evidence seen

The visible facts that support the reason. Name the measurements, angles, materials, or included parts.

Risk remaining

The unanswered point that could still change the decision. Keep it specific.

Next action

Compare one alternative, request one QC check, verify the live option, replace, or remove.

A worked hoodie comparison

CandidateWhat the row provesDecision
A · lowest item priceFront photo only; generic size chart; material unclearRemove: price cannot replace missing fit evidence
B · middle priceChest and length shown; cuffs and zip visible; fabric weight statedShortlist: request one flat measurement check
C · highest item priceStrong photos but measurements conflict between imagesResearch: clarify sizing before comparison

Candidate B wins the next action, not the purchase. It has enough evidence to justify one more check.

Use statuses that describe what happens next

Lead

Interesting but not compared.

Research

Missing one or more decision facts.

Shortlist

Survived a fair comparison.

QC-ready

The exact verification request is written.

Remove

Weak, stale, replaced, or outside the brief.

A status should change behavior. If “saved” means everything, it helps with nothing.

Run a ten-minute recheck before promoting a row

Reopen the live destination instead of relying on the saved thumbnail. Confirm the product name, selected option, visible price, measurement source, photo set, and any weight note. Then read the reason you originally saved the row. If that reason is no longer supported, lower the status even when the page still loads.

This recheck catches a common spreadsheet failure: the link survives while the evidence changes. A listing may switch options, replace images, remove a size chart, or show a different package. Record the date and the one fact you revalidated; “link works” is not enough.

Promotion rule

Move a row from Research to Shortlist only when another person could read your four notes and understand both the advantage and the remaining risk.

Know when to stop browsing

Stop adding candidates when you have three comparable rows and a new result does not improve the evidence, cost range, or risk profile. More tabs create work without necessarily creating information.

Stopping rule

If the next result cannot beat the current shortlist on one named dimension, return to the decision pass instead of opening another tab.