Choose your niches as a set

When you build one store, you choose one niche. When you build a batch, you choose a set. This guide is how to pick niches that test different things, so the batch tells you something whichever way it goes.

The rules for a single good niche still apply to every store in the set: specific rather than broad, built to last, no brands or trademarks, and something you can stand to look at for a year. Those are in What makes a good niche and Deciding on your niche. This guide adds the rules for the set.

One or two related, the rest spread out

Start with one or two niches that sit near each other. If one store is about coffee, the next could be tea and herbal drinks, or home baking. Related stores can link to each other, share social accounts, and one audience is already interested in the other.

Then spread the rest. The point of a batch is that the stores do not all rise and fall together. If every niche is a summer sport in one country, one bad season tells you nothing about your niches and everything about the weather.

Spread across markets

Each niche card on the batch page shows how deep Amazon is for that niche in each country, for example "Amazon US 300+ products, UK 200+". Read those lines when you pick. A niche with a deep catalogue in your readers' country gives the store plenty to sell. A niche with a thin one will run short of offers however good the content is.

Put a store or two in a second country where your plan supports it. Different countries search at different times and for different things, so one store in the batch is usually busy while another is quiet.

Spread across seasons

Every niche has a calendar. Christmas gifts peak in November, garden kit in spring, ski gear in autumn. A set that mixes seasons gives you something to read all year, and it stops you from culling a winter store in July because it looks dead.

Make a note of which ones are seasonal when you build them, so you judge them against their own season and not against the batch average.

A long shot or two

Keep one or two slots for a niche nobody else is bothering with. Something small, odd or new, where the card shows a decent product depth but the ratings look modest. Most of these go nowhere, and that is fine: a long shot costs the same few minutes as a safe pick, and the one that lands tends to land without competition.

Keep long shots clear of fads. Face masks and fidget spinners both looked like markets for a season. Pick small and lasting.

Do not compete with yourself

Two stores in your own batch should never chase the same searches. "Robot vacuums" and "robot mops" are one store, not two. "Fishing kayaks" and "fishing kayak gear" are one store. If two niches on your list would share most of their products or most of their keywords, keep one and give the other slot to something new.

The simple test is one product class per store. If you can name the class in two or three words, and the second niche needs the same words, merge them.

A quick recipe for ten

A set that works for most members looks like this:

1. Two related niches (one you care about, one next to it).
2. Three or four in different product classes and, where you can, different countries.
3. Two seasonal, in different seasons.
4. One or two long shots.
5. At least two templates among them. A template is a proven niche with the products already chosen, and it costs fewer chips, so it is a cheap control to compare the rest against.

You do not need a spreadsheet. Tick the cards, watch the Your batch list on the right, and read it back as a set before you press Build. If it reads as ten versions of one idea, swap a few out.

When a niche is already yours

If you already have a store that is getting found, do not build its twin in the batch. Build next to it (the related slot above) or leave that niche alone. A second store in a niche you already rank for mostly splits your own traffic.

Now go and build the set: Build your first batch.