Experiment
Portfolio Entity Architecture: SERP Co-Occurrence
A live test of whether a portfolio of independently-established specialist properties can collectively occupy more SERP real estate for a subject than a single optimized page on a single domain, now running a second phase: does strengthening a weak page inside a strong entity architecture add to the visibility already earned.
A page with weak content but strong entity context (named subdomain, topically coherent domain, interconnected network of sites) can rank in a competitive SERP, and its position will improve further when page-level content is strengthened to match the entity architecture around it. The observable signal is the number of distinct domains under the same ownership or influence appearing in the same SERP for the same query, and whether a rebuilt page holds, gains, or loses position relative to the baseline.
What this test is watching.
This test watches several Krisada-owned pages that appear together in the same Google results. We are checking whether improving one weak page helps it move up without making the related pages disappear.
How the test is structured.
The starting observation: an IBM FlashSystem search surfaced three glossary pages tied to Krisada's work, on three different domains with three different implementations: https://www.as400ibmsystems.com/glossary/ibm-flashsystem/, https://flashsystem.as400system.com/glossary/flashsystem/, and https://www.midlandinfosys.com/about/glossary-of-ibm-iseries-hardware-terms/ibm-flashsystems.
The as400ibmsystems.com entry had recently been substantially built out. The flashsystem.as400system.com entry was extremely thin at the time it was observed (two short blurbs and a CTA), sitting on a FlashSystem-named subdomain of an AS400-focused domain. The Midland Information Systems entry came from an older, unmaintained glossary and directory architecture, with content that also isn't unusually deep and that nobody has actively optimized.
Because content depth varied wildly across all three while all three still co-occurred in the same SERP, the glossary-schema explanation weakened, and entity-level context (subdomain naming, domain topical focus, cross-site interconnection) became the stronger working explanation.
Phase two, Entity Boot Camp, tests that explanation directly: the thin flashsystem.as400system.com page was rebuilt with substantially more content, semantic depth, structured data, and internal links, while the Midland property is being left untouched as a natural, unintervened control.
Control contamination note: the thin page was improved immediately after being noticed, before a baseline screenshot could be captured. Historical search and warehouse data are being used as the closest available proxy baseline instead.
What has happened so far.
The original three-domain co-occurrence remains a useful dated observation. It was one search snapshot, not proof that entity architecture caused the result.
September 11, 2026 checkpoint: the rebuilt FlashSystem glossary page did not produce a measurable page-level win. Before Entity Boot Camp, August 13 through 19, it received 16 impressions, 2 clicks, and a weighted average position of 38.4. During August 20 through 26, it received 7 impressions, zero clicks, and a weighted position of 40.9. In the latest fourteen-day window it received no impressions.
The control did not hold still either. Midland's older glossary page moved from 87 impressions and position 23.5 before the intervention to 71 impressions and position 22.0 in the first post period, then 17 impressions and position 20.7 in the latest fourteen days. AS400IBMSystems improved from position 26.1 to 21.5 immediately after the intervention, then lost most of its page-level exposure. BigTechReseller recorded its first two impressions for its FlashSystem glossary page at position 8.
A current web discovery check still surfaces multiple portfolio FlashSystem pages. Search Console confirms a broader footprint. But the August 22 AS400System family reset landed two days into the observation window and contaminated the test.
What would support or challenge it.
Status: inconclusive. Entity Boot Camp did not produce the lift the simple version of the hypothesis predicted.
The weak page became a stronger page. Its Google visibility got worse, then disappeared during a domain-family reclassification. That does not prove content depth hurt it, and it does not rescue the entity claim. The intervention and the algorithmic reset cannot be separated in this window.
The next version needs saved SERP captures, a fixed query set, page-level Search Console data, and a comparable page outside the AS400System family. The broader co-occurrence pattern is still worth watching. It is not a victory yet.
Keep Following the Tests
Move from this open thread back into the full experiment library.