Blog

Search Blog

App Annie News

Interview: A Look into App Annie’s Data Science Methodology

App Annie

App Annie’s Chief Data Scientist, Paul Stolorz, breaks down what makes App Annie’s approach to data science unique and what’s coming next from his team.

Data is central to everything we do here at App Annie. In order to deliver our industry-leading mobile performance insights, we have a world-class data science team that takes data directly from top app publishers and other sources and applies machine learning and AI algorithms to generate market estimates. 

Leading data science for us is Paul Stolorz, App Annie’s Chief Data Scientist. Paul brings over 25 years of experience, leading data science at App Annie for five years, and previously at NASA, Google, and Netflix. We sat down with Paul to shed some light on the data science methodologies that make App Annie’s mobile data platform a trusted source for over 1 million registered users around the globe. 

App Annie’s Chief Data Scientist, Paul Stolorz

App Annie: Can you explain our approach to data science?

Paul Stolorz: While data is key to our work here at App Annie, what comes first is privacy. We are on the forefront of keeping consumer data private, including aggregating it and never selling it to third parties. As regulations tighten in Europe and now California, we have always made sure that our data was collected and used in the right ways from day one. 

To say the least, data science is fundamental at App Annie, instead of an afterthought. The hallmark of our approach is to focus on innovation. We’re constantly striving for newer and better ways to analyze mobile data and make that data actionable for our customers. We emphasize and heavily invest in cutting-edge predictive modeling techniques with a world-class team and capabilities. We currently have 10 data science PhDs on our team! 

AA: How do we carry out our data science methodology? 

PS: We take actual data from leading app publishers, usage data from consenting consumers around the world, and publicly available app market data that we collect in-house. We then feed the relevant data into our machine learning algorithms that my team builds and hones. 

We leverage our big data platform and team, which ensures efficient, powerful, and scalable solutions. Our data science includes extremely rigorous model testing and evaluation to ensure we avoid overfitting. Because a regression curve that can accurately predict downloads for many apps is much more useful to the mobile market than simply memorizing the downloads for just a small number of apps. Next, we pay detailed attention to benchmarking of our results against ground truth to quantify and improve model accuracy.

AA: How is App Annie’s approach to data science unique?

PS:We work closely with our clients to provide practical solutions and insights to business problems. We are here to solve larger problems, not just provide data. Data needs to tell a story to be useful, not simply be a static asset. So our algorithms are always being updated to make sure we can be the best partner possible for all of our clients. 

We’re also unique in our ability to leverage data at scale. By combining industry-leading data science with our best-of-breed big data expertise, we’re able to develop and deploy solutions that find the best answers by using sophisticated computational techniques tuned to huge data sets. At the same time, our collective expertise enables us to support both rapid product development cycles and fast, efficient delivery of results to our customers.

Finally, a key aspect that makes us unique is that data science at App Annie is deeply embedded within the product development process. Our data scientists work extremely closely with product managers, UX designers, and big data engineers on a daily basis. At the end of the day, our goal is to delight our customers with powerful easy-to-use products that help them make quality data-based decisions. The best way to do this is to put the complexity of our data science under the hood, while providing the power that drives those products.

AA: What can the mobile market look forward to next from App Annie’s data science team? 

PS: We have an enormous amount of new products to release going forward. Doubling down on the innovation theme, look out for exciting new products that use our Natural Language Processing (NLP) expertise in fantastic ways. We’ll also have new products tailored initially to the needs of our gaming customers, which are a key audience for us. We also have ambitious plans to expand those offerings to the broader mobile market over time.

In addition to entirely new products, we won’t rest in our quest to enhance our existing products. 

While our clients enjoy the accuracy of our data, we are always working to improve. Clients can look forward to new ways to experience and interact with our data by adding features that help them in their daily lives. Even though we love technology, we know It’s all about impact, not just the technology itself. 

AA: Thanks for your time, Paul!

PS: My pleasure.

 

To learn more about App Annie’s approach to data science, get in touch with our Sales Team. 

January 6, 2020

App Annie News

Related blog posts