Welcome to Mythbusters. In this, our first episode, we dive into the complexities of attribution modelling in AdTech, highlighting the key challenges and exploring the myth that AI can solve all attribution problems and provide perfect attribution models. 

Javier is joined by Laurent Colard, SVP at VideoAmp and Co-founder and CEO at Elsy. Laurent has been a leading innovator in the field for more than 20 years during which time he has demonstrated a deep commitment to advancing data-driven solutions in media technology, with previous roles at prominent organisations, GroupM and Accenture. With a proven track record of strategic leadership and a keen understanding of the evolving media landscape, Laurent continues to play a pivotal role in shaping the future of advertising technology.

In a wide range conversation, Javier and Laurent interrogate the following themes:

Key Challenges to Attribution:

• Data Fragmentation: Gathering comprehensive data for attribution is hindered by the proliferation of digital platforms and media channels.
• Platform Silos: Different platforms measure conversions independently, leading to potential double counting and a lack of unified insights.
• Privacy and Data Access: Increasing privacy regulations and platform restrictions on data sharing complicate data collection and analysis for accurate attribution.

Key Points on Attribution Modelling:

• Historical Context: Attribution modelling has evolved from simple to complex models like multi-touch attribution (MTA) to adapt to the changing media landscape.
• Single-touch vs. Multi-touch Attribution: Recognising the need for a nuanced understanding of customer journeys, attribution models have shifted from single-touch to multi-touch.
• Incrementality and Accuracy: Determining the incremental impact of touchpoints is crucial for accurate attribution.

Debunking the AI Myth in Attribution Modelling and Future Directions:

• The Myth: AI is often seen as a solution to all attribution problems, but it faces limitations in data access and privacy concerns.
• Reality Check: While AI enhances efficiency, it cannot independently overcome fundamental attribution challenges.
• AI's Role: AI automates model development and removes biases but operates within data constraints.
• Future Directions: AI can aid in automation for better attribution modelling in the short term. In the long run, integration with composite measurement strategies and open-source Marketing Mix Modelling (MMM) promises a holistic and democratised approach to marketing analytics.

This podcast is brought to you by Fenestra. Confused by attribution models? Don't let AI myths hold you back. Fenestra, the programmatic automation platform, can help you maximise your marketing activity. Visit www.fenestra.io or contact us today at [email protected] to learn more.

We hope you enjoy the podcast and don’t forget to follow the show, it really helps us and, of course, you’ll get a nudge when our next episode drops. If you'd like to get in touch, you can do so at [email protected]

Hosted on Acast. See acast.com/privacy for more information.