The advice problem
There is a genuine shift underway. Distribution has decentralised, individual people now hold audiences that used to belong to institutions, and a person with a laptop can build a livelihood that did not exist fifteen years ago. That much is true and worth taking seriously.
The problem is who explains it. Almost all creator advice is written by creators describing what worked for them, which is survivorship bias wearing a nice jacket.
It tells you what one person did. It cannot tell you what the market is actually buying, because the person writing it has never been on the buying side.
We are on the buying side. Doomers runs creator campaigns for AI companies, the launches, the funding announcements, the product drops.
That means we evaluate creators constantly, we say no far more often than yes, and we have a fairly unsentimental view of what makes someone worth paying. This is that view.
Median, not average, and never followers
Your follower count is close to the least predictive number on your profile.
We look at your median post performance, the middle of your recent posts. The mean is distorted by a single outlier, and almost every account has one.
An account whose posts typically do sixty thousand views is worth more to a campaign than one with ten times the followers whose posts typically do eight thousand, and the second account is usually more expensive because it prices on followers.
The practical advice is unglamorous: a stable median is worth more than a viral spike. One post at 2M and thirty at 4k reads to a buyer as luck.
Thirty posts between 40k and 90k reads as a reliable channel. We can plan a launch around the second. We cannot plan anything around the first.
Audience overlap beats audience size
The second thing we check is who is actually in your audience relative to the client's buyer. A creator with a large general-tech audience is the wrong carrier for an enterprise infrastructure launch, no matter how good the numbers look. A smaller account read by the specific engineers who evaluate that category is the right one.
This is why "grow as big as possible" is bad advice past a point. Beyond a certain size, audiences dilute, you accumulate people who followed you for one thing and tolerate the rest. Dilution shows up in the median, and it shows up in how well your posts convert for a client.
The creators we go back to repeatedly tend to have audiences with a recognisable centre of gravity. Not narrow. Centred.
Why range makes you more hireable
Here is where we part company with the standard advice, in both directions.
The usual guidance is to niche down hard. From the buy side, a creator who only ever posts about one narrow topic is useful for roughly one campaign a year, because that is how often a brief happens to match.
They are the equivalent of a very specific tool: excellent when you need exactly that, idle the rest of the time.
But the fashionable counter-advice, be a generalist, post about everything, the most profitable niche is you, overshoots. A creator with no discernible centre is unbriefable. We cannot predict what their audience will respond to, and neither can they.
What actually gets booked is range within a centre. Someone whose audience trusts them on AI tooling, who can credibly carry a developer product one month, a funding announcement the next, and a design tool the month after, because the through-line is their judgment rather than their subject.
That creator fits many more briefs than the specialist and remains predictable in a way the true generalist does not.
The useful test is not "what is my niche" but: what do people trust me to have an opinion about? That question has a broader answer than most creators assume, and a narrower one than the everything-is-content crowd would like.
The craft that actually matters
Three things separate creators we book repeatedly from ones we use once.
You write it, not us
We send a content brief. The creator writes the post. The creators worth paying take the brief and write something in their own voice that we could not have written, an angle we had not considered, an objection they know their audience will raise, a comparison that only someone inside that world would make.
A creator who returns the brief lightly reworded has added nothing, and it shows in the numbers.
You will defend it in the replies
A post opens a conversation. It does not close one. Replies, especially ones the author takes part in, sit high in X's ranking ordering, and they are also where a launch either becomes credible or falls apart.
Creators who post and disappear leave most of the value unclaimed.
You are reliable about time
Launches are timed. A creator posting four hours late is outside the window that made the campaign work at all. Reliability is boring and it is the single most common reason we stop working with someone talented.
What disqualifies you
- Refusing to tag paid partnerships. This one is absolute. Every compensated post in our campaigns carries X's paid partnership label. That is what the FTC's endorsement rules require of any material connection, and an undisclosed paid post is a liability that lands on the client. Creators who will not tag do not run in our launches. If an agency tells you to skip the label, that is information about the agency.
- Bought engagement. It is visible in the ratios and it is checked. Inflated numbers on an otherwise decent account are worse than modest honest ones, because they mean nothing you report can be trusted.
- An unstable median. Covered above, and it is the most common silent rejection.
- Audience that does not match the claim. If your bio says AI and your engagement comes from a general audience, we will see it in the reply composition.
- Taking every deal. A feed that reads as a rolling advertisement has spent the trust that made the audience valuable. The creators with the best long-term rates are the ones who decline most of what they are offered.
What raises your median
Since the median is the number that prices you, it is worth knowing what actually moves it. Since 2026 the For You feed has been ranked by Phoenix, a Grok-based transformer that replaced the older multi-service stack, and xAI publishes the architecture openly. Three things follow that are directly actionable:
- Ranking is relative to your own baseline. Consistency compounds and gaps are expensive, a quiet month lowers the bar everything afterwards is measured against.
- Replies outrank likes. Posts that make a specific, disputable claim generate conversation; agreeable posts collect likes and stop. Optimising for likes optimises for the weakest signal.
- Dwell time is predicted directly. A dense post someone reads beats a one-liner they scroll past.
Worth knowing what is not true: the numeric engagement weights are not public, so anyone selling you an exact multiplier is repeating a leak or guessing. The full breakdown is here.
How to be worth approaching
We find creators by watching who consistently produces good posts in the categories our clients operate in. There is no application form that substitutes for that.
What helps: a bio that makes your centre of gravity obvious in one line. A pinned post that is your best work rather than a promotion.
Visible history in the category, evidence that you have been paying attention for a while. And a reply history showing you engage with people rather than broadcasting at them.
What does not help: DMing agencies a media kit built around follower count and reach estimates. Those are the numbers we discount hardest, and leading with them signals you have not understood how you are being priced.
The honest summary is that being worth paying and being worth following turn out to be nearly the same problem. The creators we pay most are the ones who would still be doing this if nobody paid them, which is inconvenient advice, and true.
If you would rather be approached than wait, email creator@doomers.ai and we will assess you against exactly the criteria above and tell you honestly either way. Once somebody does approach you, the money conversation has its own rules.
What to charge covers how a buyer builds the number and the terms creators hand over free, what to agree before you post covers usage rights and the clauses worth refusing, and the trust budget covers the part that decides whether you still have an audience in a year.
Questions we get asked
What do brands actually look for when hiring creators on X?
Median post performance rather than follower count or average views, audience overlap with the client's actual buyer, and engagement authenticity. A stable median across recent posts is worth considerably more to a campaign than a single viral outlier, because a launch can be planned around the first and not the second.
Does follower count determine creator rates?
Not in this market. Rates are driven by typical post performance and audience quality. Accounts that price on follower count are routinely more expensive and less effective than smaller accounts whose posts reliably reach the right people.
Should creators niche down to get brand deals?
Neither extreme works well. A creator confined to one narrow topic fits roughly one brief a year, while a creator with no discernible centre is unbriefable because nobody can predict what their audience will respond to. What gets booked repeatedly is range within a centre, a recognisable area of trust, with enough breadth to carry several different campaigns.
Do paid posts on X have to be labelled?
Yes. Any creator receiving payment, product, equity or free access has a material connection that US FTC Endorsement Guides require to be disclosed, and X provides a paid partnership label for it. Every compensated post in a Doomers campaign carries that label, and creators who decline to tag are not used.