Media Kit Built From Audited Numbers, Not Claims

A media buyer spends around ninety seconds with a media kit before deciding whether to take the call. We build for those ninety seconds, and we build the numbers before we build the page. The measurement work comes first: Before any design, we reconcile three sources that usually disagree: your ad server's delivered impressions for the last twelve months, GA4 sessions for the same period, and what your CDN log shows once bot traffic is stripped. On most publisher sites the gap between the analytics figure and the ad server figure runs 15 to 40 percent, and every advertiser diligence process finds it. Publish the smaller number with a footnote explaining the difference. The larger one loses the deal in week three and the relationship with it. Audience evidence: Traffic volume sells nothing on its own. What an ecommerce brand or a fintech marketer wants to know is who reads you and whether those people hold budget. We field a first-party reader survey through your own list — job title, seniority, purchase authority, tools already in use, and the two or three category questions that matter for the advertisers you want. Sample size and fielding window get printed on the page. A survey of 380 readers stated honestly outperforms a claim about two million monthly visitors that nobody can verify. What is produced: A designed kit in Figma, delivered as an interactive PDF under 8MB — heavier files stall in corporate mail gateways and never reach the buyer — plus a web version on your own domain that you can update without us, and the editable source file. Inside: audience composition with every source cited, placement inventory with IAB display specs and newsletter dimensions, the rate card with its pricing logic visible, two or three case studies written as before and after with the advertiser's own reported result, and the detail a media agency asks for in its first email — file weights, viewability measurement, brand safety posture, cancellation terms. What we will not put in it: Follower counts summed across social platforms and presented as reach. Impression figures we cannot trace to an ad server line item. Stock photography. Awards nobody in the buying seat has heard of. A list of banner sizes with no reason attached to any of the prices. If the only differentiated thing about your inventory is that it is cheaper, a document will not fix that, and we will say so before you pay us. Who this is not for: Publishers who want the kit without giving us access to the analytics and the ad server. We do not write audience claims from a briefing document. Also not for anyone who needs it by Friday for a pitch — the survey alone takes two weeks in the field, and a kit produced faster than that is a design exercise with numbers pasted into it. Why we build category variants: Ecommerce and DTC, fintech and developer tools read a kit differently from each other. A DTC growth lead wants delivered clicks and a path to a first purchase. A fintech marketer wants brand safety, regulatory posture and audience seniority. A developer-tools buyer wants to know whether your readers write code or only read about it. The same audience data gets framed three ways, and we produce the variants rather than one document that half-persuades everybody. Handover: Revisions are taken as one consolidated set of comments per round. At handover you receive the Figma file, the raw survey data, and the reconciliation workbook showing how every number on the page was derived — so the next person who updates the kit does not have to guess.