Turning YouTube Signals into Carbon Decisioning

Client: WPP & Indeed

YouTube is a proven performance driver for Indeed - but a channel that generates CO2 emissions roughly 20% higher than standard online video. Historically, the walled-garden nature of YouTube has made meaningful emissions optimisation almost impossible. Indeed gave The GoodNet the mandate to change that, with one non-negotiable condition: VCR must stay above 90%.

Using the rich data signals within The GoodNet's GoodIQ's carbon measurement platform, the team identified multiple data levers that could be applied as dynamic campaign optimisation levers. Rather than applying all signals simultaneously, levers were tested in isolation across two-week cycles against an Indeed-specific baseline, ensuring every finding was attributable and robust. We then modelled the results if the test variables were layered to predict what future campaigns could achieve. The results were clear. Device-level optimisation alone delivered a 6% carbon reduction. Adding geo signals brought that to 13%. Adding time-of-day signals all three levers combined projects a 27% reduction, and all with VCR maintained above 90% throughout. During the live optimisation window, over 4.4 tonnes of CO₂ were saved in real terms. The significance extends beyond the numbers. This campaign represents the most advanced application of GoodIQ carbon data in YouTube to date - transforming emissions from a post-campaign reporting metric into live buying signals that WPP's team could act on in real time.