A tech enterprise speeds up manual tasks with Marvin to gain time for more in-depth research
Criteo reinvented its repository into a centralized, self-serve hub by switching to Marvin
By replacing Dovetail with Marvin, Criteo unified fragmented data and created a powerful resource for stakeholders to derive actionable insights.
A centralized, self-serve repository
With Marvin, Criteo solved the issue of having a repository that had low adoption among the various stakeholders and consumers of research.
Now, Criteo can use raw and synthesized data to inform their business decisions. They are more closely aligned to using research to keep innovating their platform.
With features such as automated Slack summaries and shareable clips for non-account stakeholders, collaboration has improved significantly. In Marvin, they’ve found the centralized, AI-forward and self-serve repository they needed.

Why did Criteo need a new repository?
Criteo wanted an AI-forward repository that would be the single source of truth for stakeholders to search and extract insights. They did not want teams to work in silos across different tools.
Criteo onboarded Dovetail in 2021. The user research team moved data for one of their platforms from Confluence pages to Dovetail. They also implemented a tagging nomenclature.
However, they soon realized the solution would not scale for 5 platforms and more stakeholders.
As a large adtech platform, Criteo has multiple teams gathering customer feedback, including designers, product managers and product marketing. The user research team conducted an internal study to understand existing challenges.
They learnt that 80% of the stakeholders were dissatisfied with the existing research retrieval process. No one was going into Dovetail because the platform is oriented towards conducting research and not consuming it. Stakeholders were still using Confluence pages to combine their research.
They drew up clear and comprehensive requirements for a new tool:
- Centralize client feedback across different teams — sales, product marketing and design, and product managers
- Derive insights from existing data — support tickets, CSAT surveys and sales data
- Create a frictionless platform that would allow stakeholders to self-serve
- Contribute to identify and build innovative product features to give Criteo an edge over competitors
The tools they considered:
- Product Board: they soon realized it was project-management oriented
- Chat GPT
- Marvin
Criteo conducted a detailed proof of concept with Marvin. They tested over 20 research projects across 5 weeks. Each team of stakeholders was represented, and 28 members were involved in this comprehensive trial.
They imported and uploaded various types of projects and data, including interviews, CSAT surveys and 1000+ Intercom conversations.
How Marvin emerged the winner
User-friendly design
The team found that Marvin’s AI integrated seamlessly in the retrieval as well as analysis of data.
A range of features
Marvin made every step of the workflow easier, from transcription to translation, syncing discussion guides, analysis and easy sharing.
Integrations
Criteo did not want to disrupt how teams were already working. They appreciated that they could connect Marvin to apps they were using and import data from various sources.
More insights
The stakeholders could now leverage the raw data too. Dovetail’s AI did not search for insights among data that had not been tagged. As a result, not all the research could be used effectively.
Want to explore Marvin for yourself? Set up time with us to find out more.