Path Interactive is a performance-driven digital marketing agency with offices in New York City and Nashville. The team provides deep expertise in a full suite of results-oriented marketing services including PPC management, SEO, social media and display/digital media advertising. Clients include well-known brands such as the WWE, Jackson Hewitt, and Scholastic.
“We’ve saved an enormous amount of time by trusting Marin to do the daily optimization of bid assignments for our Sponsored Product campaigns. Finding the right amount of volume while keeping a consistent KPI can be a challenge, but the algorithm accomplishes this quite clearly. Marin bidding has allowed us to focus on strategic testing and growth initiatives, which makes us really excited heading into 2019.”
– Aurora Estrada, Strategist, Digital Media | Path Interactive
To get started, the team assessed the historical performance of the branded Sponsored Product keywords and set a target to improve efficiency. Path enlisted the support of Marin Software to apply machine learning to the bidding initiative. Marin Software provides enterprise marketing software for advertisers and agencies to integrate, align, and amplify their digital advertising spend across the web and mobile devices. Marin Software offers a unified SaaS ad management platform for search, social, and eCommerce advertising.
Marin’s algorithm quickly identified areas to rightsize spend across the portfolio, achieving this with thousands of bid changes each day to zero in on the right bids, yet still critically maintain a share of voice.
Path additionally leveraged boost functionality to account for seasonal volatility and budget pacing challenges across the entire folder of keywords. This proved to be an easy and beneficial way to layer human control on top of machine learning and proactively control the algorithm.
Marin bidding on Sponsored Products launched and proved immediately impactful. Looking at a period over the period comparison, revenue was 63% higher in the three week period after bidding activation compared to the three week period prior.
Positive and negative boosts were applied towards the end of the month to adjust for spend pacing. These inputs allowed the algorithm to focus on the overall ACoS target, yet still guide volume to make sure performance KPIs were attained.