Host: Mark Fidelman
Guest: Aksel Bedikyan, marketing research and analytics practitioner based in Montreal, Canada

How is AI changing marketing analytics—and what remains valuable when AI can write code, analyze data, and automate reports? Mark Fidelman talks with Axel Benikian about AI agents, the changing role of data science, and practical ways marketers can get more insight from the data they already have.

Axel explains how his team uses AI to automate recurring analysis and presentation work, while emphasizing that useful insights still depend on access to good data. His advice to businesses: map your existing data first, use AI to explore what it can tell you, and turn to custom research when important questions can't be answered with public or existing data.

• AI is broadening access to data science. Coding may become less central to some roles, while more people use AI to build apps, models, and tools for business questions.
• AI can automate repetitive analytics work. Aksel's team uses Claude to help generate analyses and recurring PowerPoint reports, including brand trackers.
• Start with the data you already have. Map your internal data and connect it to the questions you want to answer before investing in new data collection.
• AI has limits when it comes to proprietary information. LLMs can help analyze available data, but they don't automatically have access to a company's private customer or business data.
• Custom research still matters. Interviews, focus groups, ethnography, and surveys can fill gaps when existing data can't answer new questions.
• Learn by doing. Aksel recommends experimenting with AI tools and sharing what you learn with others.

• 00:27 —Aksel's background in aerospace engineering, consulting, and market research
• 02:36 — Will AI replace data scientists?
• 05:25 — AI agents: risks to jobs and opportunities for innovation
• 07:05 — How Aksel's team applies AI to marketing analytics
• 09:54 — Automating recurring analyses and PowerPoint reports
• 11:46 — How marketers can make better use of existing data
• 15:13 — When AI needs to be supplemented with custom research
• 15:47 — Staying curious and learning through hands-on experimentation

• Mark Fidelman's book, Agent Shock: agentize.com/book
• Aksel Bedikyan: OG Insights · Find Axel on LinkedIn