AdSense to Ad Manager and header bidding migration

A publisher earning four thousand euros a month from AdSense will not double it by adding header bidding. On a clean editorial site doing two to eight million monthly pageviews, the realistic first-year lift is 25 to 60 percent, and most of it arrives before a wrapper is installed at all: from consent that produces a string buyers accept, from a layout whose slots can be seen, and from supply files that let a buyer identify you. The wrapper is the last fifteen points, not the first forty. Sequence, and why it is not negotiable: Consent, then ads.txt and sellers.json, then layout, then Ad Manager, then header bidding. Shops sell the wrapper first because it is the visible product and it demonstrates activity in week one. Installed on a broken base it produces a small ambiguous lift, gets blamed for the site slowing down, and is removed within two quarters. We have been hired to repair that order more often than we have been hired to run it correctly the first time. Which Ad Manager, and what happens to AdSense: Ad Manager small business is free and carries a monthly impression ceiling. Past it you are on 360, paying a CPM-based serving fee that changes the arithmetic on low-value inventory and makes some geographies unprofitable to serve at all. Ad Exchange access needs a policy review that crypto content and parts of the beauty category fail more often than publishers expect, and we will tell you before you apply whether your archive is the problem, because a declined application is not quickly reversed. AdSense does not get switched off. It stays in the auction as backfill, and on long-tail geographies it still wins impressions no exchange bids on. Demand partners, and the MCM question: Direct seats with Magnite, PubMatic, Index Exchange, Sharethrough, TripleLift and Criteo each carry a traffic threshold and a review, and a Turkish domain serving mostly Turkish readers will be declined by several of them regardless of content quality. That is the point at which a network partner under MCM Manage Inventory is the honest answer, and it costs a revenue share you will not renegotiate afterwards. We map which doors are open to you before you spend a month knocking on closed ones. We hold no partnership with any of these companies and receive nothing if you sign with them. The money, plainly: Terms run from net 30 to net 90 depending on the partner, so the first quarter of a migration is a cash-flow trough even while revenue rises. For a Turkish entity receiving euro payments there is an invoicing and currency question that belongs to your accountant, and we will describe the mechanics without advising on them. Your effective take after wrapper fees, serving fees and any MCM share is the only number worth tracking, and it is not the number on the Ad Manager dashboard. What we do not take: Revenue share, referral fees, or ongoing management. This engagement ends. You receive the Prebid configuration in your repository, the Ad Manager setup written up as a runbook with the line item logic explained, and a working session with whoever will own it. If nobody at your company will own it, do not begin. A header bidding stack without an owner degrades inside a year and the degradation is invisible until it is large. Too early to do this if: Under roughly 500,000 monthly pageviews, where the exchanges will not seat you and the fee stack consumes the difference. Traffic that is bought, from any source, including the sources people describe as content discovery. Archives that are machine-generated, which pass a crawler and fail a demand partner's review, and that review happens after you have done all the work. We also decline sites earning from a single sponsor that want programmatic as insurance. That is a sales problem wearing an ad ops costume.