GM Gert Mellak

SEO

How to Build a Team for Search Marketing in AI Times with Wil Reynolds

Gert Mellak revisits how to build a team for search marketing in ai times with wil reynolds as a practical advisory article for modern SEO, AI visibility, and business outcomes.

Artwork for 115 - How to Build a Team for Search Marketing in AI Times with Wil Reynolds

Listen to this episode on Apple Podcasts, Spotify, or YouTube. The full podcast archive is at /podcast.

What the episode covered

  • 115 - How to Build a Team for Search Marketing in AI Times with Wil Reynolds

  • 115 - How to Build a Team for Search Marketing in AI Times with Wil Reynolds

In today’s episode, Gert had an interesting conversation with fellow SEO expert, Wil Reynolds. Wil started Seer Interactive as a one-man operation but now, Seer is home to over 200 team members across Philadelphia, San Diego, and Remote.

Wil shared some valuable insights on the hiring process, promotions, building a team and creating a company culture based on values and collaboration between innovation and the teams. He believes that company owners need to celebrate not just people who get promoted, but also those who choose to stay in their roles, helping others to get promoted and really loving the work they are doing.

  • 01:12 Introduction of the episode and the guest

02:26 The future of search marketing, its disruptive changes and what Wil Reynolds thinks of it?

06:02 The rapid pace of change and why Wil see it as a massive reset?

08:05 How did Wil Reynolds grow his team to 200 people?

What I would emphasize today

Search has moved beyond one channel. Google still matters, but your buyers also use AI answers, YouTube, Reddit, reviews, communities, and comparison content before they decide.

That means the practical question is not just “does this rank?” The better question is: does this help the right buyer find enough proof to trust us?

How to apply this now

  1. Diagnose the actual visibility gap.
  2. Check whether the buyer researches this topic on Google, AI tools, YouTube, Reddit, or another platform.
  3. Create the asset that answers the question clearly.
  4. Add proof, examples, and structure so both humans and AI systems can understand it.
  5. Measure whether the work leads to better qualified demand, not just more traffic.

BEFORE YOU GO

Find out why AI recommends someone else.

Take the free workbook with you, or find me on social.

FREE WORKBOOK

The Get-Recommended Framework

Six steps, eighteen questions, about an hour of your time. The questions you cannot answer are the diagnosis. Free, and yours to keep.