Creator Vetting: Audience Evidence and Written Rejections

What we select on: Follower count and engagement rate settle nothing, and both can be bought for less than a week of your media budget. We shortlist on three inputs a purchased audience cannot fake: the creator's own platform analytics, screen recorded rather than sent as a slide, covering the last 90 days; the behavior of their comment section on posts nobody paid for; and their paid history read out of the ad libraries instead of the media kit. Saves and sends per reach tell us whether a post travels beyond the people already following. A like tells us almost nothing, which is why it is the number every media kit leads with. What the vetting checks: Audience geography against your shipping and licensing footprint, because a home brand serving the lower 48 should not pay for reach in markets it cannot fulfill. Age distribution, which decides more than most briefs assume. The shape of the growth curve, where a vertical week with no matching lift in views is a purchase. Comment authenticity, read by hand across a sample of recent posts rather than reduced to a vendor score. Category load, meaning how many paid posts ran in your category in the last six months and whether one of them was your closest competitor. And disclosure history, because a creator who has never run a visible paid partnership label is not going to start on your brief. Category rules we apply without being asked: In fintech we cut anyone carrying a promoted token, a prop firm affiliate link, or a history of return or income claims, and we flag creators whose audience skews under eighteen for any product ending in a credit application. In kids and family we take the age breakdown from the platform rather than the creator, and we treat a made for kids YouTube channel as a different product entirely: no comments, no remarketing, and a rights conversation that starts with whether a child appears on camera at all rather than with rate. In home and furniture we establish whether the creator owns or rents, because a renter cannot drill the wall your brief assumes, and that single question prevents most of the reshoots in the category. Tools, and where they stop: Modash, HypeAuditor and CreatorIQ are a first filter, not a verdict. Their audience figures are modelled, they cannot see private accounts, and they cannot see a whitelisted ad that ran from someone else's ad account. TikTok Creator Marketplace and Meta's ad library close part of that gap. The rest is manual: watching the posts end to end, reading the comments, and asking the creator to record their analytics on screen rather than export a PDF that anyone can edit before sending. The rejections are the deliverable: You get a shortlist, and you get a written reason for everyone we cut, because the cuts are where the money is. A team that knows precisely why the obvious name was rejected can hold that line when someone senior asks why the biggest account in the category is missing from the plan. Each shortlisted creator arrives with the evidence behind the recommendation, the risks we found and did not consider disqualifying, and the price band we expect them to sign at. What we refuse: We do not keep a roster and we take nothing from talent agencies or platforms. You are the only party paying us, and we put that in the engagement letter. We will not review a list you have already signed and hand back a blessing. We do not sort by follower tier, because nano, micro and macro are ad sales language and they predict nothing about whether a post sells a sofa. Who this is not for: Teams who need the list to arrive before the questions are answered, and programs where the casting was decided by a founder and the vetting exists to ratify it.