Comparing shopping agents without a leaderboard
Four axes, and what each one changes
Agents in this market are more alike than the comparison pages suggest. Where they genuinely differ comes down to four things.
| Axis | What differs | How it shows up for a buyer |
|---|---|---|
| Fee shape | Where a minimum applies and what the share is calculated on | Small orders feel different from large ones at the same percentage |
| Route coverage | Which lines serve your destination, and how many | A destination with two routes has less room to trade cost against speed |
| Warehouse terms | How long storage is free and what happens after | Matters if you consolidate several sellers into one parcel |
| QC depth | How many photos are routine, and whether inspections are an extra | Matters most on categories where shape is the risk |
Four axes, and only one of them is the number comparison pages lead with. The other three are what people actually complain about afterwards.
Why we do not publish a league table
The case against a ranking is not modesty. It is that a single ordering would be wrong for most readers and right for almost none.
What a ranking would have to assume
- That every reader ships a parcel of the same weight to the same country.
- That fee levels, routes and warehouse terms stay fixed long enough for the ordering to survive a season.
- That speed and cost trade off the same way for everyone.
- That the weights it assigns to the four axes are the reader’s weights rather than the author’s.
What we publish instead
- The four axes, named, with what each one changes for a buyer.
- The arithmetic that turns weight and volume into a billable figure, so a reader can price their own parcel.
- Per-agent pages describing what that agent publishes, without ordering them.
- The disclosure that explains why links carry a parameter and what it does not buy.
The honest version of "which is best" is "which is best for a parcel shaped like yours", and that question has inputs only the reader has.
How to compare for your own parcel
Five criteria in the order they change the answer.
- Measure your parcel first
- Weight, longest side and a rough volume. Without these, every comparison is between numbers that do not apply to you.Where to look: A kitchen scale and a tape measure, or the volumetric tool.
- Fix the destination and list the routes
- Route coverage is a hard constraint. An agent with one line to your country is not really comparable with one that has four.Where to look: The route tables on each agent's own site.
- Apply the fee shape to your subtotal
- A minimum fee changes small orders dramatically and large orders barely. This is where the percentage comparison breaks down.Where to look: The fee structure page.
- Decide whether you will consolidate
- Consolidation turns three orders into one parcel and makes storage terms matter. If you never consolidate, that axis is worth zero to you.Where to look: The warehouse terms page of each agent.
- Weight the quality of documentation last
- It matters, but it is the axis least likely to change your total, and the one most likely to be written about at length by other sites.Where to look: This review and the agent pages on the index.
Done in this order, the comparison takes about twenty minutes and replaces a ranking you would otherwise have to take on faith.
Comparison questions
Four answers about comparing agents.
Is there an agent that is simply the best?
Not in a way that survives contact with a specific parcel. The four axes trade off against each other, and which one dominates depends on weight, destination and whether you consolidate.
Why does this site cover several agents if it links to one?
Because the comparison is useful even when the destination is one page. Outbound links carry a referral parameter to wherever they point, and the disclosure says so in full rather than in a footnote.
Do the other agent pages exist?
The index carries pages for the agents it tracks. This site links to the one whose mechanics it documents most closely; you can reach the others from the index itself.
How current is this comparison?
The axes are structural and change rarely; the numbers attached to them change often, which is why we publish the former and date the latter. The update log records edits.
The comparison only works with your numbers in it
Two inputs — weight and destination — decide more than any ranking. Both go into the estimator on the index, and both stay in your browser.