Bending Spoons: Using data science to take apps from start up to stand out

About Bending Spoons

Bending Spoons is responsible for over 20 popular apps and focuses on delivering an excellent user experience with a data-driven approach.

Industries: Technology
Location: Italy

Bending Spoons implements BigQuery on Google Cloud Platform to perform hundreds of thousands of queries daily and optimize its app monetization and targeting, increasing downloads.

Google Cloud Results

  • Analyzes the app market using BigQuery, helping to identify opportunities for growth
  • Collects and aggregates hundreds of thousands of daily events, providing complex data insights with queries by segment
  • Supports quick company growth with minimal upfront financial investment, with easy scaling as the company expands

Millions of data points analyzed every day

In a marketplace where hundreds of apps often offer similar functionalities, developers have to find ways to make their product stand out. Milan-based app developer Bending Spoons takes a data-driven approach to its choices, both in identifying the best apps to develop, and in terms of app optimizations. "By focusing on data science, Bending Spoons has grown a successful portfolio of over 20 apps including our most popular app, 30 Day Fitness," says Marco Meneghelli, Head of the Data Science and Analytics Team at Bending Spoons. "In total, our apps have been downloaded 120 million times."

"We decide which apps to develop based on our understanding of potential market reach and the monetization possibilities an app offers," Marco explains. "Our business model is built around auto-renewable subscriptions and in-app purchases, with a little advertising revenue. It's really important for us to understand the behavior of our users and be able to optimize the user experience." Once the app has been developed, Bending Spoons also improves its app store position using tools developed by the data science team.

"Every day, we collect information from millions of active users of our apps. The goal is to scale up our data science to an industrial level in order to make the most of the data we gather and make strategic business decisions."

Marco Meneghelli, Head of the Data Science and Analytics Team, Bending Spoons

In order to make data-driven choices using its internal optimization tools and grow the business, Bending Spoons needed to analyze large volumes of data very quickly. It looked for a way to implement a powerful data storage and analysis system that didn't require specialist technical support. BigQuery on Google Cloud Platform (GCP) was the perfect fit.

"Every day, we collect information from millions of active users of our apps," says Marco. "The goal is to scale up our data science to an industrial level in order to make the most of the data we gather and take strategic business decisions."

Powering big data tools

Growing a start-up quickly involves making the right choices for the business. When Bending Spoons decided to establish a data science team three years ago, it was looking for data insights, but didn't have a large tech team to manage its system. "We wanted to set up fast and grow quickly," says Marco. "We didn't want to invest too much in operations and infrastructure, but we knew we would also need the capacity to expand rapidly once the business grew."

To do that, it chose to implement BigQuery to provide the analytics power for its internal data tools. "Once our users opt-in, we begin to analyze usage data and in-app events using our internal Pico tool," explains Marco. "Events include things like installation, starting a session, beginning a new workout if it's a fitness app, or whatever other action it might be. That data alone amounts to tens of gigabytes a day, and then we also collect data points from advertising platforms and app stores. We use BigQuery on a daily basis to perform hundreds of thousands of queries with this data. The typical size of one of our queries varies from tens to hundreds of megabytes, which we would not be able to run without BigQuery."

"BigQuery was a great choice for us at the entry-level. It is easy to set up data storage and automated queries, so we were able to develop complex projects like our tool without too much system operations knowledge. We have been able to learn on the go."

Marco Meneghelli, Head of the Data Science and Analytics Team, Bending Spoons

"Thanks to BigQuery, we are able to aggregate the raw data by segment or type of user, for example, and then carry out very complex queries across these segments: to look at American users using a specific operating system on a particular phone, for example," Marco explains. "We can then feed that into algorithms to describe and predict user behavior."

"BigQuery was a great choice for us at the entry-level," says Marco. "It is easy to set up data storage and automated queries, so we were able to develop complex projects like our tool without too much system operations knowledge. We have been able to learn on the go."

"As well as using GCP to host some of our apps with Compute Engine and App Engine, we have started using Kubernetes Engine to run certain data tasks," says Marco. "We love its scaling capabilities and ease of deployment."

Providing transparency on the app market

"Another internal tool connected to BigQuery is called Crystal. It collects hundreds of millions of data points from our own apps as well as a lot of external sources, every day, and then feeds them to a number of models we've created, doing most of the computational heavy lifting for us," Marco explains. "This ultimately results in a set of analyses about the app market that help us have an insightful view of trends and competition, track the performance of our own apps, and identify opportunities for growth."

"We're extremely proud of the work we've done and continue doing on Crystal, and I must say we might not have been able to build such a powerful technology so quickly and effectively without BigQuery," he adds.

"The initial infrastructure was straightforward to set up, which enabled us to bootstrap from the beginning. Our goal now is to be one of the biggest app publishers in the world. To do that, we will continue systemizing our data collection processes with the help of Google Cloud Platform."

Marco Meneghelli, Head of the Data Science and Analytics Team, Bending Spoons

Bootstrapping: From start-up to top ten

With the help of Google Cloud Platform, Bending Spoons has been able to grow the business without spending on up-front investment in infrastructure or technical support. "Thanks to the capacity of BigQuery to aggregate data, it is possible to extract a relevant metric from our raw data quickly and easily," says Marco. "For example, ascertaining the value of a download takes up to a minute, rather than the hours it would have taken on a standard computer."

"We are using this capability in many ways. For example, testing out different creatives for our advertising campaigns to understand which is more likely to attract different segments of target users," he continues. "In one instance, the download value of the best creative was a hundred times that of the worst, so we were able to move our budget to the idea that would have the most impact."

Now, Bending Spoons plans to focus on improving systems operation routines, and is looking at starting to use Cloud Pub/Sub to stream its events data, as well as Cloud Dataflow.

"We have really grown: in the past few months we have been in the top ten global app publishers in terms of downloads, excluding games," says Marco. "The initial infrastructure was straightforward to set up, which enabled us to bootstrap from the beginning. Our goal now is to be one of the biggest app publishers in the world. To do that, we will continue systemizing our data collection processes with the help of Google Cloud Platform."

About Bending Spoons

Bending Spoons is responsible for over 20 popular apps and focuses on delivering an excellent user experience with a data-driven approach.

Industries: Technology
Location: Italy