What a bench actually is
A bench is the subset of creators who have carried three of your launches, understand what the product does without being told again, and have opinions about it. It is narrower and older than a roster.
A roster is assembled per campaign: whoever fits the brief, has the right audience, and is free that week. It works, and it is what most launches run on.
The names overlap with a roster. What differs is accumulated context, and context turns out to be worth a great deal.
The distinction matters most for companies shipping continuously. If you launch twice a year, a roster is correct and a bench is an expense with nothing to compound.
If you ship every two or three weeks, running each drop cold means starting from zero six times a quarter while a competitor with a bench starts from three launches of accumulated understanding.
Why familiarity compounds
Familiarity compounds because three separate things improve with each launch, for different reasons: the writing gets better, the opt-in rate rises, and the audience starts to recognise the pairing.
- The writing gets better. A creator who has covered your product three times writes about it with a fluency no brief produces. They know which part their audience found interesting last time, and they stop restating the basics, which is where most launch copy wastes its opening line.
- The opt-in rate rises. Creators choose whether to carry each launch, and a company they have covered successfully before is an easy yes. A company they have never heard of is a request to spend credibility on a stranger. That warming curve is one of the most under-discussed dynamics in this channel.
- The audience recognises the pairing. When the same credible account covers a company repeatedly, their followers stop reading it as a one-off and start reading it as a beat that person follows. That is a materially different signal from a single unexplained post.
The third one is the reason a bench beats a bigger roster for a company with cadence. Reach from a stranger is worth less than the fourth mention from someone whose judgment the reader has already decided to trust.
How one gets built
A bench is a by-product of running launches properly.
- Launch one runs wide. A broad roster selected on median post performance and audience overlap with your buyer. This is discovery, and it is the launch that tells you the most about who fits.
- Read the results per creator. Who wrote something that travelled, whose audience actually responded, who produced replies from people who look like customers. Campaign totals hide all of this.
- Invite the ones who fit back. By launch two you are re-engaging maybe a third of the first roster deliberately, and adding new names in the gaps where coverage was thin.
- Prune using your own judgement. Every client can remove any creator from their roster at any point, for any reason. Used consistently over a few launches, that right is the main tool shaping who ends up on the bench.
- Refresh at the edges. A bench that never changes becomes a bench reaching the same people repeatedly, which is the failure mode described below.
The selection criteria stay the same throughout: median post performance, genuine audience overlap with your buyer, and how little the selected creators overlap with each other. Creator amplification covers the criteria in full, and the creator-facing version covers what we look for from the other side.
Category exclusivity
Creators carrying your launch are restricted from working with direct competitors inside a defined window around your drop. This is a real protection and worth understanding precisely.
It is not full exclusivity, and it should not be. A creator who only ever posts about one company has stopped being a creator and become a channel, and their audience adjusts accordingly. The value you are buying is that these people have independent credibility, which requires them to be independent.
What the window does prevent is the specific failure of a creator carrying your launch on Tuesday and a competitor's on Thursday, which serves neither company and makes both look like buyers of the same inventory.
What it costs you
A bench costs you two things: repeated exposure to the same people performs worse each time, and a narrow bench sets a smaller ceiling.
Repeated exposure to the same people performs worse each time. Ranking systems predict engagement per impression, and an audience that has seen four posts about you from the same account responds less to the fifth. This is why a bench needs deliberate refreshing at the edges.
A narrow bench is a smaller ceiling. A launch aimed at genuine breadth needs names outside it, including creators whose audiences have never encountered you. The right structure for most companies with cadence is a stable core carrying the regular drops, widened deliberately for the launches that need to reach past the existing audience.
Both are manageable. Neither is a reason to run every launch cold, which is the alternative most companies default into without deciding to.
Questions we get asked
How is a bench different from a roster?
A roster is assembled per campaign from whoever fits the brief and is available. A bench is the subset who have carried your launches before, know the product, and no longer need the basics explained. The names may overlap; what differs is how much context has accumulated and how much better the writing gets as a result.
Do the creators work exclusively with us?
Not exclusively, because a creator who only ever posts about one company stops being credible to their own audience. What we do apply is category exclusivity: creators carrying your launch are restricted from working with direct competitors inside a defined window around your drop.
How long before a bench actually exists?
Three or four launches is the point where it starts behaving like one. The first launch is discovery, the second shows who wrote well and whose audience responded, and by the third the same names are opting in early and writing without needing the product explained again.