Illustrative shape, not measured data. The crossing point is not a fixed number of products; it moves with how well the catalogue is organised.
Every catalogue has a size past which browsing stops being a viable way to find anything, and search and filters have to carry the load instead. Most dropshipping stores cross it by importing products, then read the resulting flat conversion rate as a traffic problem. The symptom is specific and checkable: revenue concentrates in a handful of products, deep catalogue pages get almost no views, and the share of visitors using your search box climbs while a rising share of those searches return nothing.
The problem arrives by import
Almost nobody builds a 600-product store deliberately. It happens because importing is a button. A supplier app offers a category, the category has 300 items, and adding them costs nothing and takes a minute.
What it costs is not visible in the admin. Every product added past the point where a shopper can hold the store in their head makes the previous ones slightly harder to reach. You end up with a larger catalogue and the same handful of products in front of anyone.
Nothing breaks, which is why this one goes unnoticed. The store works. Every page loads. Conversion just settles at a number lower than it should be and stays there.
Five symptoms, all checkable this afternoon
You do not need a tool to check most of this. You need thirty minutes and a willingness to look at the unglamorous end of your own reports.
| Check | Healthy | Past the tipping point |
|---|---|---|
| Share of products with zero views this month | Small tail | Most of the catalogue |
| Revenue concentration | Top sellers lead | Nearly all revenue in a handful of items |
| Products per collection | Fits a few screens | Pagination nobody reaches page 3 of |
| Share of visitors using search | Steady | Climbing, because browsing stopped working |
| Searches returning nothing | Under 5% | 10 to 20%, or worse |
Row four is the counterintuitive one. A rising share of visitors using search is usually reported as a good thing, and in isolation it is: searchers convert better than browsers. But a climbing search share on an unchanged store often means navigation stopped working, not that shoppers got more decisive.
A search box is not a feature when browsing works. It is a symptom when browsing doesn't. The Dropshipping Playbook
The empty page is the whole problem
Two versions of it, and both end the visit.
Someone types the name of a product you stock, spelled slightly differently, or the word they use for it rather than the word your supplier used. They get a blank page. Industry compilations put the share of searches ending this way at 10 to 20% on a typical store, and report that a large majority of shoppers who hit an unsuccessful search simply go and buy elsewhere.
That visitor had the highest intent of anyone on your site that hour, and the store told them it does not sell the thing it sells.
Worse, because it is self-inflicted by the interface. A shopper picks a size, then a colour, and the page empties. Nothing warned them the combination was impossible, and nothing offers a way back except undoing their own choices one at a time.
Most stores have never seen a list of which filter combinations do this, because nothing records it.
Neither event appears in a Shopify report. Both appear as a session with a short duration and an exit, which reads as disinterest. Chapter 3.4 is about why that is structural.
Depth is not the same as size
The number of products is a weak predictor. What actually decides whether a catalogue is navigable is how many decisions stand between a landing page and a product page.
- One level deep is comfortable at almost any size, if the categories are the words shoppers use.
- Two levels is fine when each branch is meaningfully different from its siblings.
- Three or more is where community store audits keep flagging navigation "too complicated for anyone to actually find a product."
The other half is naming. A category called by your supplier's term rather than your customer's is a dead end that looks like a working menu, and it is invisible from the admin because you know what the word means.
The thirty-second test
One exercise covers the rest of this chapter, and it is the one from chapter 3.2 with the stopwatch pointed at a different thing.
- Name a specific product you sell to someone who has not seen the store
- Ask them to find it, on a phone, without using the search box
- Time it. Over thirty seconds, your navigation is the problem, not your traffic
- Now let them use search, and try the word they would naturally use rather than your product title
- Note whether that search returned the product, something else, or nothing
- Finally, apply two filters at once on a collection and see whether you can empty the page
If the page empties, write down the combination. That is a real leak with a real number attached to it, and it is one of the few in this level that can be fixed in an afternoon. The fixes themselves are chapter 4.8 onward.
Common questions
How many products is too many?
Should I just delete products?
Does a bigger catalogue help SEO?
Why does revenue concentrate in a few products?
One number pays for the rest
What a single order is worth decides what every other fix can afford. Chapter 4.1 is where that number moves.
Go to chapter 4.1