351 pages in one channel: what that number is worth
Three things this piece settles
- Every channel exposes 351 pages of 24 rows. The reading repeats because one catalogue is filtered per agent, so channels overlap and their rows do not add up.
- Six of eleven channels publish a self-description, and none of the six describes scale.
- Switching channels means rebuilding the list; the column set, not the rows, is the cost.
Which stop this touches
8 stops · open a stop to read its windows · “n open” = actions still available
- Stop 1: Pick the channel — 3 actions still open. A channel is a filtered view of one agent, not a shop. Learn what its row count proves and what it cannot.
- Stop 2: Hand the link over — 3 actions still open. A link carries an identity, not a product. Keep the identity intact and the order stays traceable.
- Stop 3: Pay for the item — 2 actions still open. Item payment and freight payment are two separate moments. Most closed doors close at the first one.
- Stop 4: The warehouse clock — 3 actions still open. Arrival starts a clock. Verify the window on your own account before you plan around it.
- Stop 5: QC photos — 3 actions still open. A photo set is evidence of coverage, not a verdict on quality. Ask for angles, not for opinions.
- Stop 6: Merge and pack — 3 actions still open. Merging trades a second freight charge for a bigger volumetric weight. Run both sides.
- Stop 7: Freight and customs — 2 actions still open. Freight is charged on chargeable weight. Customs is decided on declaration fields you fill in.
- Stop 8: Checking the parcel — 2 actions still open. Delivery starts the claim clock. Photograph the unopened parcel before you open it.
The same four fields on all eleven channels
Compare channels on four columns and nothing more. The columns are what the channel says about itself, how many list pages it exposes, how many rows sit on one page, and the product of those last two. Four columns is enough to rank eleven channels. More columns usually means more gaps rather than more information. Rank first and read later.
The page count is the only column that is uniform. Each of the eleven channels reported 351 pages on its list page when the index was captured on 2026-10-06. That uniformity is itself the finding worth reporting. A column that never varies cannot separate anything. It can only set the scale of the choice.
Rows per page is 24 across all eleven channels. The derived column is therefore identical everywhere as well. That gives 8,424 rows on any one channel. The reading repeats across all eleven, which tells you the list is filtered rather than partitioned. The derived column is the one worth remembering. It sets the size of the problem you are choosing between.
The self-description column is where the table breaks. Six channels publish one and five publish nothing at all. A blank is not a small value but an absent one. The column cannot be averaged or ranked in any useful way. It can only be counted, and counting gives six.
A comparison resting on the self-description column compares six marketing lines against five blanks. That is a weak basis for a decision. It is, however, a strong basis for a question. Ask what the five silent channels are choosing not to say. The silence is more informative than the copy.
Where the eleven channels actually differ
Strip out the shared numbers and the visible differences are thin. joyagoo calls itself a one-click shopping agent, and acbuy publishes four words and stops there. Those two descriptions differ more in length than in content. Neither one names a capability you could test from outside. Both are aimed at a search result rather than a buyer.
Three channels name marketplace coverage in their first line, listing Taobao, 1688 and Weidian together. Two of those three attach a savings or deals claim to the same sentence. Those claims are written for a search result rather than a specification sheet. They tell you what the channel wants to be found for. That is useful to know and useless to compare.
One channel opens with a superlative about itself and a low-cost claim. That is the entire self-description captured for it in the snapshot. Nothing in that text is checkable from outside the channel. Treat it as positioning rather than as information. Positioning answers a question you did not ask.
None of the six describes a category focus, a parcel limit or a warehouse location. The columns you would want for selection are absent from every one of them. That absence is consistent across all six rather than patchy. Consistent absence is far easier to plan around than patchy detail. You stop looking for something that was never there.
Read that absence as a fact about the source rather than a fault in it. These pages were built to route traffic toward an agent. Routing does not require a specification, so none was written. Expecting one is a category error on your side. Look for the specification somewhere else.
What 8,424 rows per channel actually means
The number 8,424 is a claim about a list, not about a warehouse. It counts rows on pages, and each row is one listing rather than one product. Variants hide inside rows, so the product count is larger than the row count. Nothing in the number says how many rows are live. Treat it as an upper bound on what you could look at.
Divide it by what you can actually read in a session. Eight pages at 24 rows is 192 rows in one sitting. That is 2.28 percent of one channel and 0.21 percent of the whole index. Reading all of it is not a plan anybody finishes. The number sets the scale rather than the target.
Treat the row count as a denominator rather than a destination. Your job is to make the denominator irrelevant by narrowing early. A category filter turns 8,424 rows into a much smaller number. Choosing a channel for a category focus you cannot verify is the opposite move. Narrow by category before you narrow by channel.
Keep the export file as the working set rather than the channel list. An exported list is finite, countable and stable across sessions. A channel list is none of those three things. The export is where the 8,424 rows become your rows. Everything after that happens in the file.
Re-measure the page count before reusing it. The figure is dated 2026-10-06 and describes exactly one day. A channel that adds or removes pages changes the arithmetic quietly. The derived total changes with it. Re-read the pagination text rather than trusting a stored number.
Choosing by order volume and category
For a single small parcel, treat the eleven as interchangeable at the catalogue level. Choose on the page you will actually operate rather than on the description. The operational page is what you will look at twenty times. The description you will read once and forget. Optimise for the page you keep open.
For repeated hauls, weight the export path first. Build the list once, export it, and carry the file rather than the browser tab. A file survives a channel change and a tab does not. That single habit removes most of the switching cost described below. Habit beats tooling here.
For one heavy category, check the category counts before you check the channel name. Our own 192-row sample of joyagoo is 44 shirts-tees, 36 hoodies-sweaters and 27 pants-shorts. Those three categories are 107 of 192 rows, or 55.7 percent. A sample tells you what a channel carries, and a name does not. Count first and name second.
For multi-variant listings, count how many rows force a variant choice. Twenty-six of those 192 rows, 13.5 percent, declare more than one style. Each one costs a decision before it can be packed. Variant density predicts effort better than catalogue size does. The size of the list is not the size of the job.
Where the catalogue matters less than the paperwork, pick on the fields. Choose the channel that lets you set the fields you will need later. Fields get used under time pressure and row counts do not. The field list is short and worth reading once. Read it before you read the descriptions.
What switching channels costs you
Assume nothing transfers between channels. A channel is a view onto listings, and the saved rows live in your own export. Treat the export as the asset and the channel as a lens. Lenses are replaceable and assets are not. That single distinction settles most switching decisions.
Re-registering a row means re-reading the title, the image and the variant decision. That is the same work as the first pass over that row. There is no partial credit for having seen it before. The work scales with the list rather than with the change. A large list makes a small switch expensive.
Keep the export column set identical across channels so two lists can be compared. A column renamed in one export breaks the comparison silently and without an error. Silent breakage is worse than a failed import. Name the columns once and reuse the names everywhere. Consistency in the header row is what makes the join work.
Budget the switch as a full pass over the list rather than as a migration. The rows do not move between channels; you rebuild the list from the same source. A migration implies transfer and a rebuild does not. Choosing the wrong word sets the wrong expectation about effort. The word matters because the work is real.
If the switch is prompted by a missing field rather than a missing row, check your own tooling first. Tooling is cheaper to change than a habit is. A habit costs weeks and a script costs an afternoon. Move the habit only when tooling genuinely cannot close the gap. Most gaps close in the file rather than in the channel.
The data point behind this piece
The eleven agent channels each expose 351 list pages of 24 rows, which is 8,424 rows on every channel. That is one reading repeated eleven times rather than eleven separate pools; the site’s own sitemap lists 11,393 unique product URLs on the same date.
Basis: 2026-10-06 snapshot: pagination text read from the first list page of each of the eleven channels on the source site, multiplied by the 24 rows per page counted in the same snapshot.
Six of the eleven channels publish a self-description and five publish none, so any comparison table built on self-description is empty for 45 percent of its rows.
Basis: 2026-10-06 snapshot of the eleven channel hero subtitles; presence or absence of text recorded per channel, five channels recorded as no self-description published.
Our own 192-row sample covers 8 of 351 pages, or 2.28 percent of one channel, and the eleven-channel figure is not a larger pool, because the channels overlap. It also shows three categories holding 107 of 192 rows, or 55.7 percent.
Basis: Sample size and category counts from src/data/finds.ts (192 rows, 8 pages) divided by the 8,424 rows reported on the channel page in the same 2026-10-06 snapshot; the 11,393 unique product URLs come from the source site’s own sitemap.
This piece answers searches about “joyagoo agent comparison”. It is part of the rules and scope cluster.
Where to go next
Links to Kakox are tracked so the destination can attribute the visit. Nothing on this site is sold by us.
Reviewed 2026-10-06 · Weekly review · next review weekly review