Better measurement, deeper sites, working tools September 7 to 12, 2026 This week we made the portfolio easier to measure, gave several sites more useful material, and connected working tools to clearer business offers. The rundown draws on 92 portfolio work-log records. Those include follow-ups and corrections, so the stories below group related work instead of treating every event as a separate accomplishment. First, make the evidence dependable A portfolio this size needs a measurement system we can question. This week we found several places where the numbers looked settled until we followed them back to the requests behind them. We repaired crawler counting and classification, recovered 35 days of historical measurements, and fixed a nightly cleanup that was deleting eligible Applebot and Sogou page records. The last repair brought the page detail and the dashboard into agreement across all 130 cached sites in that verification. Those 130 sites were the scope of the warehouse check, which also includes properties outside the portfolio. The practical result is a better starting point for decisions. We can distinguish successful content requests from other traffic, keep crawler purposes separate, and show a missing historical day as unavailable. The Good Bots query column now uses whole observed requests to associated landing pages. Several queries can share the same page, so those rows cannot be added together as a portfolio total. None of that tells us what an AI system said about us. It tells us what reached the site and gives us a cleaner way to investigate what happened next. Give the healthcare sites something useful to be found for The largest content expansion happened across the healthcare group. AIMedicineToday went from 29 sitemap URLs to 181. That included 63 glossary terms, 41 device records, 21 physician tools, eight new articles, and useful content behind 18 previously thin hubs. AIHealthcareNow gained a market map covering 27 companies, seven buyer-category briefs, and 16 company profiles. Where AIMedicineNow already had deeper clinical coverage, the sites connected readers to that work. AIMedicineNow also published seven vendor stories during the week. AgeLifeForward added ten clinic profiles and four buyer guides around tests, scans, costs, and evidence. The common thread was usefulness: give someone comparing a service or trying to understand a term enough substance to make a better next decision. The new pages have shipped. Their search results are still something to observe. Publishing 152 additional URLs on one site is a build milestone; the next question is which of those pages earn useful visibility. Connect a useful inspection to a clear next step QuickRankAI was rebuilt around a working 38-check website inspection engine. AISymantix gained its 15-check visibility checker, and we aligned their basic text measurement so the same page would not receive conflicting word counts simply because the tools extracted its text differently. AISymantix then gained a clearer paid-audit offer, a request form, private inquiry storage, and source tracking. A visitor can run the free page check, understand what it covers, and request the deeper audit when that is useful. This gives us something concrete to measure from here: checker use, audit requests, and qualified inquiries. We do not have a historical conversion rate to announce. We have the flow and the measurement in place. The IBM sites became better buying references Across nine Power subdomains, we added CPU reference pages and built or replaced power-and-cooling references. The work used generation-specific specifications and identified the questions that still lack confirmed answers, especially around Power 12. AS400System gained a generation-spanning server hub and a charted comparison of server requirements. BigTech Reseller expanded from one IBM provider profile to eight, while related generation sites gained relevant links into its directory and quote flow. Together, those changes connect a technical question to a useful comparison and then to a commercial next step. A big crawler spike sent us in the wrong direction at first BellyUp supplied the most interesting correction of the week. A single-day GPTBot surge reached 43,309 requests. Following the URLs showed that a small venue inventory had generated a very large space of filter combinations: 40,404 distinct faceted URLs around 31 venues. The crawler activity was real. The interpretation needed work. Much of the fetching went into near-duplicate filter pages, while the machine-readable venue information received very little attention. Crawl controls and discovery fixes were shipped, and the investigation became a three-part RealSEOLife article series, field notes, a proof entry, and a follow-up experiment. BellyUp also gained a broader discovery framework across Atlanta and five category subdomains: 34 categories, 20 published category pages, a venue website scanner, an Atlanta discovery report, and a restaurant claim flow. Fourteen category gaps remain openly identified. The next useful result is whether requests move toward the actual venue information. Some of the most useful work covered the whole portfolio A configuration exposure sweep protected 122 config directories and tightened permissions on 34 world-writable directories. The sitemap schedule gained 17 live sites that had been omitted, and four generators were repaired after they had been writing to the wrong location. Three later federation fixes added 100 previously unlisted records to machine catalogs. These are the kinds of failures that let a page exist while making the supporting information harder to find or maintain. Fixing them gives the work already on the sites a more dependable way to stay discoverable. Keep the findings connected to the work RealSEOLife refreshed its existing experiments, added new proof and controlled tests, completed a broad voice cleanup, and added audio narration to a proof entry. Its weekly research routine completed its first verified end-to-end publication. Signal Architect Group expanded eight research articles and connected BellyUp to Project Everywhere. Other focused builds included a nine-provider Tampa managed IT comparison on BestLocalX, stronger service pages and articles on PharmDigitalMarketing, and deeper encryption, privileged-access, and vulnerability coverage on the IBM i Security site. AgeBetterToday gained four sourced reference-range tools. NaturalHerbLibrary added interaction coverage and researched draft monographs that still require human review. There was also a useful limit to preserve. The five-site AISymantix study found that crawler activity rose on all five sites while impressions rose on only three. That did not establish a reliable search-visibility forecast. We published the finding with the counterexamples because those are part of the result. At publication, the latest available Google data ended on September 9. Across the 115 active public properties included in this snapshot, the preceding 28 days recorded 237,141 Google impressions and 450 clicks, a click-through rate of 0.19 percent. Average position was 58.9, weighted by impressions. The work log covers September 7 to 12 and has a broader portfolio scope. These are dated baseline values for later review; this week's new releases still need their own observation window. Watch whether the new healthcare and IBM resources begin earning impressions for the questions they were built to answer, and whether AISymantix checker visitors become qualified audit inquiries. Revisit BellyUp after enough new crawler data arrives to see whether requests shift toward useful pages. A deployed crawl control is a completed fix; the resulting behavior needs another measurement. The ten-site content rotation has its roster, tooling, and public experiment prepared. Activation was still pending in the records reviewed for this edition. NaturalHerbLibrary's new draft monographs also remain in the human-review queue. Each future edition will keep its own evidence and revisit earlier questions when there is something new to say. A longer conversation can take one of these lead stories and follow the decisions behind it.