Roku, Samsung and LG Buying with Verified Supply Paths

In the connected television accounts we have audited since 2023, between a tenth and a quarter of the spend labelled CTV landed in inventory whose app bundle we could not resolve against an app-ads.txt file. That is why this engagement opens with a supply path review rather than a media plan. Buying the three television makers direct: Roku, Samsung Ads and LG Ad Solutions each sell their own inventory, and each also lets it leak into the open exchange where it arrives resold, marked up and occasionally impersonated. We buy operating system inventory from the operating system: Roku through its own manager or a direct insertion order, with the exchange used only where a demand-side platform is genuinely warranted; Samsung through its own console and platform; LG through Ad Solutions. Where a reseller is the only route into a format, we tell you what the extra hop costs before you approve it. Roku and Turkey: Roku does not operate in Turkey. A Turkish brand buying Roku is buying the United States, Canada, the United Kingdom, Mexico and Brazil — export markets and diaspora, not the home audience. We say that before the contract rather than after the first invoice. If Turkey is the whole target, the plan is built on Samsung and LG home screen inventory and the local streaming applications instead. Automatic content recognition, and what it can honestly do: All three makers run content recognition on their sets, and it is the only measurement in this part of the market that sees linear and streaming in one panel. It supports reach and frequency across the two and incremental reach against a television schedule. It does not support the conquesting story that gets told about it: segment definitions and refresh windows differ by maker, and none of them will show you the underlying event log. We use it for reach planning and linear overlap. We do not present it as proof that we reached a competitor's viewers. Frequency, across three stacks that do not speak to each other: There is no shared identifier across Roku, Samsung and LG, so a unified cap is not available and anyone offering one is modelling it. We cap per platform, set the caps from a reach curve rather than habit, and reconcile duplication afterwards through a measurement partner. On a four-week flight our working ceiling is around three exposures per household per week per platform, moved on evidence rather than on a request to spend faster. What we refuse: Home screen and first screen takeovers as an opening purchase. They are the most expensive unit each maker sells and the hardest to attribute, and we buy them once there is a baseline to compare against. Open exchange connected television without verified bundles. And annual guaranteed-rate commitments for a brand with no measured baseline in the channel. Crypto and energy: Crypto and web3 clients go through a policy check before a plan is written. Several of these platforms restrict financial products by market, and Turkey carries its own limits following the central bank's 2021 ban on using crypto assets for payments. Energy clients — retail supply, solar installation, charging networks — face a different constraint: the buy is regional and geographic granularity varies by maker, so we plan around postcode-level delivery being unavailable rather than promising it and explaining later. Who this is not for: Brands who want the lowest connected television CPM available. It is available, it is on the open exchange, and we will not buy it. Also not for teams who need the plan inside a week; the supply review alone takes longer than that and it is the part that earns the fee. What you receive: A supply path map naming every seller between your budget and the impression, the buy live on the platforms we recommended, a per-platform reach and frequency report, and a written note on what the flight did not establish.