The Digivisory experiment

Search is shifting from a list of links to a direct answer. If AI answer engines are going to summarize and cite content instead of sending traffic to it, the old SEO playbook stops being the whole game. I wanted to understand the new rules by running them, not reading about them.

Stand up a real publishing property and treat it as a laboratory. Test how structure, schema, and content design affect whether AI systems discover, trust, and cite a source — and measure what actually moves the needle.

Digivisory is a live content property built to probe answer-engine optimization and generative-engine optimization. It is deliberately separate from this site: Digivisory is the experiment, and jeffjellis.com is the record of the work.

I publish and instrument, then watch how AI answer engines and traditional search respond. The property is structured for machine readability from the ground up, so I can isolate what earns citations and surfacing versus what does not.

The experiment is ongoing. As results accumulate, this page will carry the findings and case-study data. For now it stands as a working test bed rather than a finished study.

The lessons feed directly back into the memory and publishing work. Understanding how machines read and cite content is the same muscle as building systems that remember and surface the right context at the right time.