More than a billion people around the world depend on AccuWeather’s forecasts and warnings for their travel, safety, and day-to-day planning. In 2019, the Pennsylvania-based media company found an opportunity to improve both its advertiser and user experience. The team noticed that their domain-level ad viewability was below that of their market cohort, hovering around 30% for mobile web in particular. AccuWeather gathered all hands on deck and set out to optimize its mobile site and inventory performance, with the goal of significantly improving ad viewability, site quality, and ad effectiveness.
A few months of laser-sharp focus led AccuWeather not only to faster loading times, but also to a revenue boost. Just three months after completing a final round of testing in April 2020, in the midst of a globally declining CPM (cost per thousand impressions) trend due to COVID-19, AccuWeather saw a 42% jump in CPM, along with 39% revenue growth and a 10% lift in render rates across all ad slots.
Gathering data to strategize for success
AccuWeather began by comparing the number of pages they delivered to their earnings and impressions. By digging into reporting data provided by Google Ad Manager, AccuWeather’s platform for managing and serving ads across all its sites, team members realized that their low mobile web viewability was likely the result of latency caused by tags on pages. The poor latency was slowing both ad and page load times. They needed a way to make these pages more lightweight, without affecting overall revenue performance.
Testing for impact
AccuWeather held a brainstorming session with its partners at Google, generating ideas on how to maximize the value of its existing content. Then, between Q4’19 and Q1’20, its teams ran various tests with multiple code clean-up strategies to reduce mobile web page load time, decrease latency, and improve ad viewability. In particular, AccuWeather tested the effectiveness of their premium top banner slots.
By merging behavioral data from Google Analytics with Ad Speed and Ad Unit data sets from Ad Manager reporting, Accuweather’s teams were able to identify the biggest areas of opportunity. They then tested each piece of code in the page-loading process to measure its impact, with the overarching goal of speeding up the requests sent to Ad Manager. Once the tests were complete, AccuWeather sorted out nonessential client-side processes or libraries by either moving them to the server, deprioritizing them, or removing them completely in order to decrease load times.
Though the tests continued into the early days of COVID-19, the possibility of a slimmer ad market only encouraged the team to strive toward greater improvements. “Our feeling was, as demand dried up, those who were still spending money would look for the highest performing opportunities, so we kicked off a project to improve all our core ad slots,” says Steve Mummey, VP of Programmatic Sales & Operations. That assumption turned out to be correct.
Focus on quality and monetization yields results across the board
Since completing its final testing sprint in April, AccuWeather has seen a 34% increase in domain-level viewability, 3x faster page load time, and a 50% decrease in time to first ad request. These product successes have translated to increased revenue, as shown by 42% CPM growth, 39% revenue growth, and a 10% increase in render rate across all ad slots.
In particular, viewability for top-tier 300x250 ads grew from 50% to 75%. “While the biggest improvements were on our ‘above the fold’ ad units, each slot on our core pages increased at least 10% points simply by being painted to the screen before the user scrolled by,” says Mummey.
As for next steps, AccuWeather is applying the learnings from mobile web across its platforms, generating greater value for users, advertisers, and its business - further demonstrating their unique ability to reach users in the right place at the right time.