No one wants to overpay for impressions or let whales get away because you didn’t bid enough to win the auction. Bidding the right price for the right impression demands next-level machine learning to analyze all 350 billion opportunities we see each day – a volume higher than the queries handled by the world’s largest consumer search, streaming, and social platforms.
When there are an average of 22 different places to buy the exact same impression, all served simultaneously, it’s obvious that not all ad supply paths are equal. Fed by 4 million bid requests per second, Bid Optimizer identifies the minimum bid price required to win the impressions that drive ROI.
Moloco starts learning as soon as first-party data is ingested and combines eight simultaneous model inferences per bid request, based on both historical data and shared postbacks from MMPs.
First- and second-price auctions require very different bidding strategies. Moloco has developed models for a wide variety of scenarios to accurately predict the right price to bid, regardless of which network and which bidding strategies are in place.
It’s every marketer's dream: more budget to spend when performance is strong, but without sacrifice. Weekly Budget Optimizer delivers performance results by optimizing your budget over the course of a week, both time of day and day of week. Rather than meet daily spending requirements, Moloco learns from your app’s historical data to identify when users are most active and inclined to convert, and the Budget Optimizer ensures your budget is always oriented towards ROI.
Installs occur when people have more time to explore and enjoy their apps. In fact, iOS installs are 25% higher, and Android installs are 6% higher on weekends than on weekdays. Weekly Budget Optimizer uses machine learning to reserve additional spend for periods of highest return, where a daily approach would have exhausted the option of increased budget allocation.
Every ad impression is an opportunity to build your business. Moloco will spend your entire weekly budget wisely, without ever going over. Built-in safety nets prevent early depletion or lowered spending on weekends.
Moloco’s results speak for themselves. Advertisers enjoy about a 10% lift in performance by optimizing weekly rather than daily. And e-commerce apps averaged a 29% performance improvement on re-engagement campaign budgets managed for weekly optimization.
Are you sending the right message to the right user? With Moloco, you don’t need to make assumptions or rely on your gut. Use data and insights to optimize your creatives so they predictably and reliably drive growth and performance.
By uploading a range of creatives, you can optimize which messages, ad sizes, and formats deliver the best results and focus your budget on the combinations that drive the highest success rates for each specific campaign and variation.
Look at the creative groups’ performance for the late-funnel metrics that are important to your business objectives. This may be IPM, CTR, CVR, or CPI. Remove those creatives that are not performing well and automatically invest more in the groups that are.
Moloco Studio is powered by multi-disciplinary experts from around the globe who are focused on combining design, technology, and a surplus of data insights to ensure we outperform even your most demanding performance goals.
Moloco’s machine learning engine lets you unlock the full value of your first-party data on the open internet, driving campaign performance at a demonstrably accelerated rate. Our machine learning is powered by a Deep Neural Network (DNN) that optimizes immediately, iteratively, and in real time. It’s how we find your highest-value users at scale.
Moloco is uniquely suited for the challenges today's marketers face with limited data. Our machine learning engine optimizes based on all the signals available, ensuring you’ll see stellar campaign results without relying on third-party data that may be erroneous, irrelevant, or subject to privacy and policy changes.
Our machine-scale automation doesn’t rely on limited audience segments. We look at every one of the 350 billion bid requests we see each day and calculate the probability of them representing a likely converter for your campaign. It’s how we scale, and how we avoid diminishing returns.
Moloco starts learning as soon as first-party data is ingested, building its targeting criteria even before you bid on your first impression. By combining eight simultaneous model inferences per bid request – from user conversion likelihood to price optimization and fraud prevention – you’re guaranteed rapid and sustained performance.
Most platforms adapt slowly to change and rely narrowly on signals from past performance. Moloco trains on a much broader set of context-relevant signals and actually improves as policies change over time. In fact, Moloco generated 40% better ROAS at up to 50% lower cost on iOS’s LAT than on non-LAT impressions.
Want to learn more? Dr. Sechan Oh explains machine learning and deep neural networks in this blog article.