Scarcity is easier to build than reputation

There are two ways to get out of the middle of a price band. You can become better known than the people around you, which takes a year and is not entirely within your control, or you can offer something they do not, which takes an afternoon.

The second one is underused, and it is underused for a boring reason: nobody looks at what is missing. New earners read the top listings to learn what to offer, which is how everyone ends up offering the same four things.

How to find one

The method is unglamorous and takes about an hour.

Open thirty or forty listings in your category and write down what each one sells, in a column. Do not evaluate anything. Count the results.

You will get a distribution with a very fat head: written assessment, scored assessment, video, and some combination of those. Then a thin tail of two or three unusual things, and then nothing.

The nothing is the interesting part, and the way to find it is to ask what a buyer might reasonably want that does not appear anywhere in your list. A short assessment at a much lower price point that fits an impulse purchase. A first-timer format aimed explicitly at people who have never commissioned this before, with the anxiety handled in the description. A structured self-assessment the buyer completes and you critique, which halves your assessment time and gives them something to keep. Assessment against a stated goal they supply, rather than against a general standard. A recurring subscription of periodic short assessments.

Some of those exist somewhere. The point of the exercise is not originality in the abstract - it is finding what is absent from the specific market you are listing in.

The other source is your own declines. Every request you turned down because it did not match anything you sell is a data point about unmet demand, and it is better data than the listing survey because someone was actually willing to pay for it. Keeping track of enquiries that did not convert turns that into something you can count rather than something you vaguely remember.

Test it before you commit to it

The failure mode is building a whole offer around a format that six people a year want.

Test cheaply, in this order.

List it and see. Adding a tier to an existing listing costs nothing and the market answers within a fortnight. This is the entire test for most formats, and people skip it in favour of thinking about it for a month.

Price it high on the first run. You are not optimising revenue yet, you are finding out whether anyone wants it at all, and a high price gives you room to come down. Coming down is easy; going up after you have anchored low is the problem described in the piece on price testing without annoying anyone.

Do the first three manually and badly. Do not build a template, a rubric or a workflow for something that might not survive the month. Produce three by hand, time them, and only then decide whether it deserves infrastructure.

Set a kill date. Six weeks, one line in your notes. Without it, a dead tier sits on your listing for a year confusing buyers about what you actually do.

Pricing with nothing to compare to

The awkward part of an unserved format is that there is no band to sit inside, which is normally where a price comes from.

Build it from below and from beside.

From below: your minutes per job times what your time is worth, plus the platform's deductions, which is the floor calculation in the piece on what a flat fee does to your minimum price. That gives you a number you must not go under, not a price.

From beside: what does the buyer's alternative cost them? If the closest substitute is your standard written assessment plus a follow-up, the niche format can price near that total even if it takes you less time - because you are pricing the outcome, not the labour.

Then add the scarcity component, which is real and which most people leave on the table. A format with no comparison set cannot be shopped against anyone. The buyer who wants it either buys from you or does not buy, and that is a genuinely different negotiating position from being the fourth cheapest written review on a page.

Expect the number to be wrong twice before it is right. That is fine, and it is much less costly than in your main tier, because a niche format has no established buyers to grandfather - the whole subject of why a crowded category compresses prices is the reason the escape is worth the fumbling.

The two real risks

Too small to fill a week. A format with genuine but thin demand is a good third tier and a terrible main business. Keep the standard offer running underneath it. The correct position is that the niche adds margin on top of a base that pays the bills, not that it replaces the base.

Absent for a reason. Sometimes nobody offers a thing because it is unpleasant, prohibited, or does not survive contact with the platform's rules. Check that before you build. Anything that requires holding more buyer material, or material of a different kind, changes your exposure, and the general treatment of what that means is on the privacy side rather than something to improvise. The rules themselves have moved recently: in the UK, Ofcom's illegal harms codes have required platforms to take specified safety measures against illegal content since 17 March 2025, so check the platform's current rules rather than last year's.

There is a third, softer one: a format nobody has seen has to be explained, and explanation costs conversion. If the tier needs a paragraph before a buyer understands what arrives, it will underperform an obvious tier at the same price. Name it in the buyer's language, not yours, and describe the deliverable in one sentence before anything else.

Two places are worth looking for gaps rather than at your competitors. What buyers say they are trying to find when they commission someone frequently describes something nobody is selling, and reading it as a list of unmet requests rather than as advice is the most productive way through it. And the automated tools define the floor: whatever they do well is not a niche, it is a commodity with a price of zero, so the gaps worth filling are always on the judgement side.

Once you have one worth keeping, Rate Cock's judge listings are where you can check whether it is genuinely absent in the market you are listing in, which is the only market whose answer matters.