How to Frame Experiments as Exploration

After my conference talk yesterday, marketers said privately, “We’re overwhelmed." Yesterday, I gave my first conference talk on Google's new "AI Mode" - Google's AI mode uses vector embeddings and deep personalization, changing how marketers approach organic visibility. - Traditional keyword targeting and rank tracking are becoming obsolete. After my talk, some attendees vented their frustrations dealing with so much change all at once. E.g. new AI tools, less traffic, AI search platforms. My advice: Get comfortable operating in two distinct mindsets: 1) Creative exploration — Curiosity-driven brainstorming to uncover ideas for new marketing strategies and applications of AI tools. 2) Focused execution — Ruthlessly narrowing down to the most promising ideas and refining them. Trying to explore creatively and execute strategically simultaneously leads to confusion, inefficiency, and burnout. My approach: Separate your innovation process into two distinct phases. Phase 1: Exploration -> Time for open-ended experimentation with new AI tools and marketing strategies. Phase 2: Focus -> Time to identify the most promising ideas, prioritize aggressively, and commit resources to execution. In my work, I toggle between these modes multiple times per week. This works for me as long as I'm clear about which mode I'm in. Otherwise, I might dive too deep into the weeds in exploration mode or get distracted by a tempting new idea in focus mode. Marketers who are clear on when it's time to explore and when it's time to focus can avoid frustration and innovate as the ground shifts beneath us.

Most people will say that being hypothesis-based is the primary way to conduct an experiment… but not me. I think it misses the big opportunity. There are two totally different, yet completely effective ways to conduct an experiment: - Hypothesis Based  and - Discovery-Based When I was a Ph.D. student, I would say that discovery-based experimentation played a bigger part in my life because you’re actually trying to accomplish new feats. Let’s lay out the differences within the scope of a marketer: Say you’re unsure what kind of message is going to attract what type of person. The discovery-based way to conduct this test is to put out a bunch of hooks of value and see who engages with what. You determine which demographics of people end up engaging the most based on the types of value put out. The key here is you’re not forming an opinion beforehand. You’re letting the experiment guide a theory. Contrarily, hypothesis-based experiments would start by stating a theory that you are going to test. In this instance, you might say “Parents love this supplement to stay healthy during flu season,” and then by putting out hooks of value, you can either confirm or deny, generally, that hypothesis. This is where A/B tests generally live. With these differences in mind, I want you to consider this going forward: How do you drive discovery? Discovery-based experiments are where the magic happens because you *clears throat* discover. You find new answers you may have never deemed possible.  There are 3 things you need to do to drive effective discover-based experimentation: 1) Ask Open Questions 2) Have Very Good Data Instrumentation  3) Be Completely Unopinionated I broke down these three points a bit further in today’s Whiteboard Wednesday video below. 👇 Go check it out and let me know your take on experimentation in the replies!