# Phase II Preliminary Status Report Phase II begins where Phase I ended: the foundational search signal has already been established, and the research question now shifts from whether the corpus can be discovered to how broadly, quickly, and consistently that discovery can expand. --- ProbleMattic Semantic Corpus Discovery Experiment ## Abstract Phase II begins where Phase I ended: the foundational search signal has already been established, and the research question now shifts from whether the corpus can be discovered to how broadly, quickly, and consistently that discovery can expand. Early Phase II telemetry shows a meaningful change in search behavior. Standard search visibility has become more routine, crawl attention has increased, newly published material is being surfaced within days, and Google's Generative AI features are touching a wider range of page types and conceptual neighborhoods than during the earliest testing period. Generative AI visibility began as concentrated testing and developed into broader, more frequent exposure across a widening range of archive topics and page types. These observations are preliminary. They do not establish causal knowledge of Google's internal systems, stable site-wide semantic identity, or durable generative-search authority. They do, however, justify Phase II as a distinct expansion stage focused on recency, breadth, recurrence, and cross-neighborhood retrieval. ## 1. Purpose and Scope This report provides the preliminary Phase II status of the ProbleMattic Semantic Corpus Discovery Experiment and intentionally follows the structure and evidentiary discipline established in the Phase I Technical Report. Phase I asked whether a curiosity-driven, internally connected corpus could produce coherent search signals without being converted into a keyword-first publishing operation. The Phase I conclusion was affirmative but bounded: the signal existed across several independent forms, while broad, stable site-wide semantic recognition remained unproven. Phase II therefore begins from a different baseline. The experiment is no longer asking whether Google can find individual ProbleMattic pages or whether isolated concept neighborhoods can generate query fan-out. The new question is whether search systems will return to the corpus more frequently, process new material more quickly, surface a broader range of topics and page types, and begin to demonstrate recurring retrieval behavior outside the historically dominant neighborhoods. As in Phase I, this report treats Search Console observations as confirmatory telemetry rather than business-performance metrics. Specific impression totals, click totals, and other traffic counts are intentionally omitted. The goal is to document classes of behavior, changes in cadence, breadth of page participation, and the character of the emerging search pattern without overstating what aggregate telemetry can prove. ## 2. Phase I Handoff 2.1 What Phase I Established Phase I established that the corpus was producing more than isolated indexing events. Google repeatedly returned to coherent semantic neighborhoods, exposed multiple human-intent formulations within those neighborhoods, distributed visibility across related pages, recorded user visits on more than one page in the same concept family, surfaced a descendant page in generative search, and produced an interpretive-intent outlier outside the dominant topical clusters. The strongest Phase I conclusion was structural rather than volumetric. The corpus behaved as though related pages could be retrieved as parts of a larger conceptual environment. That result justified moving the experiment forward without changing the human-first editorial method. 2.2 What Phase I Did Not Establish Phase I did not prove that Google had formed a stable site-wide understanding of ProbleMattic, that every new page would be discovered rapidly, that generative-search participation would recur across unrelated topics, or that broader topic recognition would persist over time. Those unresolved questions became the natural research territory for Phase II. ## 3. Phase II Research Focus 3.1 Primary Research Question Once an internally connected corpus has demonstrated a coherent search signal, can that signal expand in breadth, recurrence, and processing speed while the editorial system remains curiosity-driven rather than query-manufactured? 3.2 Supporting Questions - Will search systems return to the corpus with increasing regularity rather than through isolated bursts? - Will newly published material become visible quickly enough to suggest a shorter publish-to-discovery and publish-to-testing cycle? - Will Generative AI features surface pages from unrelated conceptual neighborhoods rather than repeatedly selecting only the original baseline topics? - Will archive pages, collection pages, tag pages, and individual essays all participate in retrieval, indicating that search systems are interacting with more than one layer of the site architecture? - Will interpretive and adjacent-intent behavior recur often enough to move site-wide semantic identity from an early possibility toward a defensible pattern? - Can query telemetry continue to inform editorial observation without becoming the editorial assignment system? ## 4. Phase II Observation Framework 4.1 Recency and Processing Latency Phase II introduces recency as a first-class observation. During the earlier stage of the experiment, the central question was whether material would surface at all. The emerging Phase II question is how quickly newly published material enters observable search behavior. A recent essay entered Google's Generative AI reporting within days of publication, providing the clearest early indication that the distance between publication and search-system testing has shortened. This observation should not be treated as a precise measurement of crawl, indexing, interpretation, or selection latency because Search Console does not expose the full internal sequence. It is nevertheless valid external telemetry: newly published material moved from the site into an observable generative-search surface on a notably short timeline. 4.2 Crawl Attention Sitemap activity has also become more frequent. The observed cadence shifted from intermittent retrieval to daily attention during the opening period of Phase II. Crawl frequency alone is not a ranking signal and should not be interpreted as one. Its importance here is contextual. Increased crawl attention occurred alongside faster visibility of new material and broader participation across search surfaces, making it a useful supporting indicator of a more active relationship between Google and the corpus. 4.3 Recurring Standard Search Visibility Standard Search Console performance has become increasingly routine rather than episodic. Recent reporting windows have shown repeated visibility across consecutive periods, with familiar baseline topics continuing to appear while broader site activity develops around them. The significance of this pattern is not the volume of any individual window. It is the emergence of a repeatable baseline in which the corpus is being surfaced regularly enough that ordinary continuity now carries more evidentiary weight than an isolated spike. 4.4 Generative AI Breadth The most visible early Phase II development is the widening range of pages appearing in Google's Generative AI feature reporting. Early generative-search activity was concentrated in a small number of established neighborhoods. More recent activity has touched the home page, the Musings archive, collection pages, tag pages, and individual essays spanning operational analysis, behavioral interpretation, self-help, cultural criticism, music, everyday observation, and other unrelated subjects. This breadth matters because it reduces dependence on the original baseline neighborhoods. Generative AI visibility began as concentrated testing and developed into broader, more frequent exposure across a widening range of archive topics and page types. That statement is currently the clearest summary of the Phase II opening signal. 4.5 Fresh-Page Participation A newly published essay concerning internet character classes appeared in Generative AI reporting shortly after publication. This is especially useful evidence because it combines two Phase II research interests: recency and cross-topic breadth. The page did not belong to the oldest or most established search neighborhood, yet it was surfaced quickly enough to suggest that new material is no longer entering an entirely cold environment. The working interpretation is that fresh pages are joining a corpus Google already visits and tests rather than introducing themselves from scratch. This remains an inference from external behavior, not a claim about Google's internal representation, but it is consistent with the broader change in crawl cadence and retrieval recurrence. ## 5. Preliminary Phase II Evidence 5.1 Established Neighborhoods Remain Active The original Uncrustables and opportunism neighborhoods remain visible and continue to provide a stable baseline. Their persistence is useful because Phase II is not replacing the earlier signal; it is testing whether additional material can join it. A healthy expansion pattern should preserve established retrieval while adding new conceptual territory. 5.2 Generative Search Is Touching More of the Corpus Generative AI reporting now includes pages from multiple editorial and conceptual categories, including operational material, cultural defenses, behavior-oriented essays, self-help-adjacent pages, archive navigation, collection structures, and highly specific observational pieces. The appearance of structurally different pages is as important as the topical variety because it suggests that retrieval is not limited to one template or one content type. 5.3 Visibility Is Becoming More Frequent The temporal pattern of generative-search activity has changed from early concentrated bursts separated by quieter periods toward more frequent appearances across a broader sequence of days. The current evidence does not establish stable daily generative visibility, but it does indicate that participation is becoming less exceptional. 5.4 New Material Is Entering the System Quickly The early appearance of a recently published essay is the strongest preliminary recency signal in Phase II. The exact internal path is hidden, but the observable result is sufficient for the experiment: a fresh page became eligible for generative-search exposure within days rather than remaining invisible for an extended discovery period. 5.5 Search-System Attention Is Becoming Habitual Taken together, recurring standard-search visibility, increased sitemap attention, faster fresh-page participation, and broader generative-search exposure support a cautious but meaningful working description: Google is no longer behaving as though ProbleMattic is an unfamiliar site encountered occasionally. The corpus is receiving repeated attention across multiple observation layers. Phase II is designed to determine whether that emerging habit becomes durable. ## 6. Evidence Summary Test Area | Expected Phase II Signal | Observed Preliminary Evidence | Status Crawl attention | More regular return to the corpus | Sitemap retrieval shifted toward daily attention | Emerging Fresh-page discovery | New material surfaces on a shorter cycle | A newly published essay appeared in generative-search reporting within days | Preliminarily confirmed Generative-search breadth | Unrelated topics and page types begin to participate | Generative AI reporting spans archive, collection, tag, and essay pages across multiple conceptual territories | Preliminarily confirmed Visibility recurrence | Search activity becomes less episodic | Standard and generative surfaces show repeated activity across consecutive reporting periods | Emerging Baseline persistence | Original neighborhoods remain active while expansion occurs | Established topic families continue to surface during broader testing | Confirmed Cross-neighborhood retrieval | Search tests the corpus beyond dominant historical topics | Behavioral, operational, cultural, and miscellaneous pages are appearing beyond the original baselines | Emerging Site-wide semantic identity | The type of answer is recognized across unrelated subjects with stable recurrence | Breadth is increasing, but recurrence is not yet sufficient to establish durable site-wide identity | Not yet established ## 7. Interpretation The preliminary Phase II evidence suggests that the experiment has entered an expansion stage rather than merely repeating Phase I. The original neighborhoods remain useful anchors, but they no longer account for the full observed behavior. Search systems are touching a wider range of pages, fresh material is appearing quickly, and the temporal pattern is becoming more regular. The most defensible interpretation is not that Google now comprehensively understands the ProbleMattic brand or corpus. The available evidence still cannot support that claim. The defensible interpretation is that Google is interacting with the corpus more frequently and across a broader semantic and structural surface than during the earliest discovery period. This distinction preserves the evidentiary standard established in Phase I. The experiment does not need an inflated claim to be successful. A shorter observable path from publication to retrieval, combined with wider page participation and recurring attention, is already a meaningful Phase II development because those are exactly the classes of behavior the second phase was designed to observe. ## 8. Limitations and Threats to Validity - Google Search Console remains an aggregated and delayed observation layer and does not expose the internal sequence of crawling, indexing, interpretation, ranking, or generative selection. - The Generative AI feature view identifies participating pages but does not currently expose the triggering query, leaving the semantic path between user intent and surfaced page hidden. - More frequent sitemap retrieval cannot be treated as proof of ranking authority or direct evidence of semantic understanding. - A fresh page appearing quickly in generative search is important recency telemetry, but a small number of such cases cannot establish a stable publication-latency rule. - Generative-search participation may be influenced by product changes, experimental surfaces, query mix, algorithm updates, and reporting changes outside the control of the experiment. - The corpus continues to grow and change, so corpus size, internal linking, page age, editorial expansion, and search-system behavior remain confounded variables. - Cross-topic breadth is now visible, but durable site-wide semantic identity requires recurrence across unrelated neighborhoods over a longer observation period. - The experiment intentionally remains human-first and does not include keyword-manufactured control content, so causal comparison against a conventional SEO publishing strategy remains outside scope. ## 9. Preliminary Phase II Status Phase II has started strongly against its intended research questions. The opening telemetry contains positive indicators in each of the areas that matter most for the next stage: recurrence, crawl attention, fresh-page participation, generative-search breadth, and expansion beyond the original baseline neighborhoods. None of these indicators should be promoted prematurely into a final conclusion, but their convergence is sufficient to characterize the Phase II opening as materially different from the discovery conditions documented in Phase I. The key shift is behavioral. Phase I established that Google could find the corpus, recognize coherent neighborhoods, and test related pages. Early Phase II behavior suggests that Google is returning more often, touching more of the archive, and incorporating new material into observable search behavior on a shorter cycle. The experiment has therefore moved from signal existence toward signal expansion. ## 10. Phase II Measurement Priorities - Track whether fresh essays continue to appear in search and generative surfaces within short publication windows. - Track whether Generative AI participation continues to widen across unrelated conceptual neighborhoods and editorial collections. - Distinguish persistent page participation from one-time testing by observing recurrence over longer windows. - Watch for additional interpretive-intent queries outside established neighborhoods in standard Search Console reporting. - Monitor whether archive, collection, tag, and individual essay pages continue to participate as distinct retrieval layers. - Preserve the human-first editorial rule: search telemetry may identify live wires, but it must not become a production quota or keyword assignment engine. - Document platform limitations, especially the absence of query-level attribution within Generative AI reporting, so interpretive uncertainty is preserved in the final technical report. ## 11. Initial Thoughts Phase II begins with evidence that the corpus is entering a more active relationship with search systems. The original Phase I signal remains visible, but the important development is expansion: faster observable participation of newly published material, broader Generative AI exposure, more varied page types, and increasingly regular search-system attention. Generative AI visibility began as concentrated testing and developed into broader, more frequent exposure across a widening range of archive topics and page types. That pattern is not yet a final Phase II result. It is the preliminary condition from which the remainder of Phase II should be evaluated. The research question has therefore changed in a meaningful way. The experiment no longer needs to ask whether Google knows the address. The next question is how often it returns, how much of the building it explores, and whether newly added rooms become part of the route without requiring the editorial system to abandon the curiosity-driven method that produced the corpus in the first place. ## Appendix A. Phase II Working Definitions - Processing latency: The externally observable interval between publication and the first appearance of a page in search or generative-search telemetry. It does not imply direct knowledge of Google's internal crawl, indexing, or selection sequence. - Fresh-page participation: Search or generative-search activity involving recently published material rather than only long-established pages. - Retrieval recurrence: Repeated appearance of the corpus, a page, or a conceptual neighborhood across separate observation windows rather than one isolated event. - Generative-search breadth: The range of topics, page types, collections, and conceptual neighborhoods represented in Generative AI feature reporting. - Crawl attention: Observed frequency with which Google retrieves sitemap or site resources. It is contextual telemetry, not a ranking claim. - Cross-neighborhood retrieval: Search activity that reaches conceptually distinct areas of the corpus rather than remaining confined to established baseline topics. - Site-wide semantic identity: A hypothesized state in which search systems repeatedly retrieve ProbleMattic across unrelated topics because of a stable interpretive pattern or answer type. This remains unestablished. - Confirmatory telemetry: Observed search behavior that supports or weakens the experiment hypothesis without claiming direct access to search-engine internal reasoning. --- ProbleMattic is written and maintained by Matthew Kulcsar, a software engineer, project manager, technologist, platform builder, emergency-services-trained helper, grandfather, and lifelong collector of broken systems, odd behaviors, and useful nonsense.