Google Ad Manager's forecasting feature helps web, app and video publishers periodically estimate how much traffic they can expect across a range of audience segments. While forecasts can accurately predict the future based on historical data they also need to reflect one-off events like the launch of a new app next month or a bombshell news story dropping next week if they're to be useful in helping publishers maximize their revenue without over-selling their inventory.
Conducted over more than a year, this project involved partnering with a dedicated researcher, a small engineering team dispersed across New York City and Chicago, product managers, and visual designers to plan and conduct several research studies, hold cross-disciplinary workshops, formulate and refine designs, prototype, build, and deliver two iterations of ’forecast adjustments'.
Successful closed beta, open beta and GA releases of ‘CSV upload’ and ‘visual editor’ adjustments with encouraging adoption and engagement rates and positive feedback from targeted publishers.