Podcast Media Plans, With the Shows We Turned Down

A finished plan from us is nine to fourteen shows, and it arrives with the sixty or so we rejected and why each one failed. Clients read the rejection file twice. It is where the judgement is. Where the shortlist comes from: Not the network's top sellers, and not a podcast ranking chart. We start with ad-occurrence data from Magellan AI and Podscribe to see who has been running on a show and for how long. An advertiser renewing on the same show across three flights is telling you something no download figure will. Then audience composition from Edison Podcast Metrics and the publisher's own listener survey, read against your customer definition rather than against a generic age band. Then we ask for the things networks do not volunteer. A twelve-month download curve per episode, not a show average. The split between the first seven days and the first thirty. Ad slots sold per forty minutes of content. Whether the sponsor read sits baked into the episode file or is served dynamically. A back-catalogue show delivering 55 to 70 percent of its downloads inside seven days is behaving normally. A show at 95 percent has no catalogue value and should be priced as a flight, not as a partnership. A show where one episode carries more than 40 percent of a month is a guest-driven spike, and it will not repeat for you. The disqualifiers: No measurement certified against IAB Tech Lab v2.2, no place on the plan. Downloads quoted without a stated window, same answer. More than six ad slots in a forty-minute episode, because your read will be the fourth thing that listener sits through. Blind network buys where shows are named after the contract is signed. Shows whose host has read for a direct competitor inside ninety days. Shows that will not say whether their download counts include automatic downloads from apps nobody opened. Sector notes: Education and edtech: the plan is built around enrolment windows, not the media calendar. Two shows running for eight weeks ahead of an intake beats twelve shows running all year on the same money, and we will say that before you brief an annual plan, not after. Local services: podcast geo-targeting resolves at IP level and degrades badly on mobile networks. Below a catchment of roughly 400,000 adults we would rather point you at streaming audio with postcode-level targeting than sell you a podcast plan that spills most of its impressions outside your service area. Beauty: the category is thick with promo-code buys that cannibalise each other. We check code overlap across the shortlist before anything gets priced, because two shows sharing an audience and a discount code will each claim the same conversion. This is not for you if: You have already chosen the shows and want them booked. That is trafficking, and it is cheaper elsewhere. You need the plan in four days, because the download-curve requests alone take a week to come back and a plan built without them is a guess with a logo on it. You want the planning fee bundled into a buying commission. We price this separately so the recommendation is not an argument for our own execution fee. The deliverable: A show-by-show document: what each show is doing in the plan, the certified download figure and its window, ad load, a host-read versus dynamic recommendation, the opening rate against our view of a fair rate, and the specific reason to be there. A rejection appendix. A flighting grid with the reasoning behind the sequence. A one-page brief that your buyer, or ours, can execute without a call.