Predicting GDP With Google Trends: The OECD Weekly Tracker

Tech Blog Bhawna Krishnan

As we saw in the first blog post of the Dataiku AI Lab’s ML Research in Practice series, we'll once again observe how the Lab assists the machine learning (ML) community in academia and develops solutions to aid everyone on their data journey. Along with its research objectives, the Lab also applies its understanding to practical issues by working with Dataiku data scientists and architects to resolve enterprise data issues.

The Dataiku AI Lab met with Nicolas Woloszko of the Organization for Economic Co-operation and Development (OECD) Economics Department & NAEC Innovation LAB at the time, who walked us through predicting gross domestic product (GDP) with Google Trends in this particular edition of the series.

The NAEC Innovation LAB operates within the OECD with the aim of providing a space for researchers across the OECD to work together on specific projects that apply and experiment with new analytical tools and techniques so as to diversify and strengthen the OECD’s analytical tools.


The OECD Weekly Tracker of GDP growth offers a real-time, high-frequency gauge of economic activity using ML and Google Trends data. It covers a wide range of OECD and G20 countries. It applies an ML model to a panel of Google Trends data for 46 countries and gathers data about search behavior related to consumption, labor markets, housing, commerce, industrial activity, and economic uncertainty.

In the future, the study can be extended by studying regions based on ongoing work with CFE and including developing countries in the mix. There is also room to investigate indicators other than GDP, such as employment.

What’s up next in the series? We will delve into the business challenges around the issue of data distributional shifts over time in the context of textual data, where conventional approaches do not work.

You May Also Like

Machine Learning Research, in Practice: LMQL

Read More

From Chatbots to Agents: Augmenting LLMs With Tools

Read More

Tackling Imbalanced Learning With Generative Synthesizers

Read More