Starting out
Do not compete on the axis they already won
Tenure is unbeatable and irrelevant if you are the only person offering your thing.
Guides on Starting out: A rubric is a production tool before it is a quality claim, The brief is rarely the request, Getting the first booking, in order
The people ranked above you have three advantages you cannot take: they started earlier, they have more completed jobs, and the platform's sort order rewards both. Do not compete on those axes; compete by being the only person offering a specific thing.
Competing on tenure is a plan to lose slowly, and the position actually available to a new entrant is not "better" - it is "the only one". That sounds like a slogan and it is really an arithmetic point: on a crowded axis you are one of forty, and on an axis nobody else has claimed you are one of one.
The three axes you should stop looking at
Tenure. It accrues at one day per day for everybody, including the people ahead of you. The gap never closes by effort.
Review count. Same problem, compounded, because reviews arrive faster to people who already have them. Salganik, Dodds and Watts showed the mechanism in a 2006 experiment with 14,341 participants: when people could see what others had chosen, social influence increased both inequality and unpredictability of success. This is the loop described in the cold-start piece, and the only useful response to it is to stop treating the count as the thing you are competing on.
Price. The one axis where you can win instantly and where winning costs you the market. Undercutting invites the response covered in what happens when a market gets crowded: the established earner drops for a fortnight, absorbs it out of a cash buffer you do not have, and you are gone.
Each of these is a contest against someone with more of the resource than you. None of them is where new entrants actually break through.
What an unserved angle looks like
An unserved angle is a request buyers make that nobody in the visible listings answers directly. It is almost never a new kind of work. It is usually a specific combination of format, constraint and audience that everyone else has left implicit.
Concrete shapes it takes:
| Axis | Crowded version | Claimable version |
|---|---|---|
| Format | Written assessment | Written assessment in a fixed one-page structure, delivered as a file |
| Turnaround | "Fast" | Same evening, stated cut-off time |
| Language | English | A second language, natively |
| Depth | A score and a paragraph | Six named axes, a justification for each |
| Buyer type | Anyone | People submitting for the first time and nervous about it |
None of those require skill nobody else has. They require a decision nobody else has bothered to make in public.
The honest test: could a buyer search the listings and find three other people who say the same sentence you say? If yes, you have described a category, not a position.
Finding yours without guessing
Read forty listings in your category and write down the first line of each. You will find the same four or five sentences repeated with different word orders. The gaps are visible within an hour, and they are usually not subtle - whole formats, whole turnaround promises and whole buyer types go unmentioned by everybody.
Then check the gap is a gap and not a graveyard. Some angles are empty because nobody wants them. Cross-reference against where enquiries actually originate, which the piece on demand sources sets out, and against what buyers describe wanting when they explain a commission in their own words - the buyer-side account of choosing a judge is the clearest version of that, and it names several things the listings rarely mention.
The other filter is automation. Any angle whose entire value is a fast generic verdict is already occupied by software that costs nothing, and the technical account of what an automated score reliably produces marks that line precisely. Positioning into it is positioning into a free product. The same logic applies to anything a buyer can self-serve from a free consumer scoring tool in ninety seconds.
Naming it so it survives a scan
An unclaimed position is worthless if the buyer does not register it in the two seconds they spend on your first line.
Name the thing, not the quality of the thing. "Detailed and thoughtful" is a claim every listing makes and no buyer can verify. "Six-axis written assessment, one page, same evening before 9pm" is a claim that is either true or false, and a buyer can tell which without asking you.
Put it in the first line, in the buyer's noun rather than yours. The first-line post goes further into what survives a scan, and it matters more here than in most contexts, because a position nobody notices is functionally the same as not having one.
Then repeat it identically everywhere. The position is a memory device, and a memory device that changes wording between your listing, your reply and your delivery is not one.
Refusing the head-on comparison
Buyers will sometimes put you next to someone with four hundred reviews and ask why they should pick you. The losing answer is any version of "I try harder" or "I am cheaper".
The answer that works is to move the comparison. "They do general assessments; I only do the six-axis written format, same evening." That is not a claim to be better. It reframes the choice as a question of what the buyer wants rather than who is more established, and a buyer who wants the specific thing is no longer comparing.
This is also why the narrowest possible offer is the correct one while you are unknown, and why widening it later is safe. Specificity is credibility when you have no record, as the piece on picking a first format works through, and it converts into range once the record exists.
What to do when someone copies you
They will, and it takes about six weeks. The position was never the moat.
What accumulates behind a claimed position is the thing that is actually hard to copy: reviews that all describe the same specific deliverable, in the same words, from buyers who came for that thing. Twelve reviews saying "exactly the six-axis breakdown described" is a stronger asset than forty saying "great, thanks", and the copier starts at zero on it.
By then the next move is usually not a new angle but a deeper one - the formats nobody is offering covers where that goes, and the specificity worth being pedantic about, if precision is what you are selling, is set out in the measurement conventions rather than invented per job.
If you want to see how a listed position reads to somebody browsing rather than to the person who wrote it, the judges directory on Rate Cock is a live page of them, sorted the way buyers actually see them, and reading forty of those first lines in a row is a faster education than any amount of thinking about your own.