The IBM FlashCore Module Campaign: One Dataset, Seven Sites, Seven Different Jobs
Same Part Numbers, Seven Search Intents: What Happened When One IBM Dataset Went to Seven Sites at Once
Two things stood out, and neither was in the original brief.
What was measured.
The short version.
Krisada supplied a full IBM FlashSystem FlashCore Module part-number and feature-code dataset, 17 SKUs across four capacity tiers, and asked for a cross-site content strategy across the Big Tech constellation. The brief was deliberate: publish the same underlying facts on all seven sites, each framed for that site's actual commercial role, rather than picking one site to own the topic. Round one shipped products, articles, glossary terms, and FAQs across all seven properties and caught a production CSS bug live on two of them along the way. Round two, eleven days later, came back and fixed a factual gap the first round didn't know it had: IBM had shipped a fifth FlashCore Module generation in the meantime, and three sites were still telling readers there were only four.
What was the system?
This wasn't a single article syndicated seven times. Each of the seven Big Tech constellation sites plays a different commercial role: AS400System.com is a procurement catalog, the Power11 sites are spec-and-dataset authorities built on a precision-over-marketing promise, AS400IBMSystems.com does foundational education, AS400IBMSystem.com runs an upgrade/ROI funnel, AS400Software.com sells only the licensing layer and has a strict no-hardware rule, and BigTechReseller.com is a multi-vendor reseller discovery site. The same 17-SKU dataset had to become seven different documents, each answering a different question a different buyer was actually asking, without contradicting itself anywhere or breaking any site's own brand rules.
What changed?
Round one (2026-08-06) built out all seven sites against that constraint: four hardware products with full part-number tables and a buyer's guide on AS400System.com; a directory listing with a 15-row spec dataset on Power11.AS400System.com; a flagship dataset-download article (CSV, JSON, and Dataset JSON-LD) on Power11.AS400IBMSystem.com, matching that site's existing hero-content pattern; a foundational-education article and dedicated FAQ page on AS400IBMSystems.com; an upgrade/ROI article framing the FCM3-vs-FCM4 decision on AS400IBMSystem.com; a new storage-management category on AS400Software.com holding only software licensing entries and deliberately zero hardware listings; and a product listing plus a vendor-comparison article against Dell and HPE on BigTechReseller.com, in that site's existing comparison-table format.
Round two (2026-08-17) started as a technical question, not a content brief: which IBM FlashSystem models actually support which FlashCore Module generation. IBM's own documentation pages blocked automated fetch, so the compatibility mapping got cross-validated across the IBM FCM Product Guide Redbook, IBM support and announcement pages, and trade press until multiple independent sources agreed, turning up a fifth FCM generation (FCM5, GA March 2026) that the first round had no way to know about. Midway through, Krisada supplied a competitive-analysis list built from real Midland Infosys warehouse and GSC data, identifying specific model pages the portfolio had never built despite proven demand. The second round expanded flashsystem.as400system.com into the primary FCM hub (new listings for models 5030, 5200, 9100, 9110, and 9150, new Support/Parts/Drives hub pages, a dedicated Dell PowerStore comparison section, conversational-question headings matching how the target queries are actually phrased), added a verified FCM5 catalog entry to AS400System.com, added VIOS/NPIV host-attachment content to AS400Software.com, added a confidence-labeled roadmap page to Power12.AS400System.com, and added a rep-facing sales-objection article to BigTechReseller.com.
What did the evidence show?
Two things stood out, and neither was in the original brief.
First, round one's real value wasn't the seven pieces of content. It was a production bug the content work exposed. BigTechReseller.com's glossary and article templates referenced CSS classes with zero matching stylesheet rules, and in one case, a fully built, correctly styled component sat completely unused in the CSS while the template used different made-up class names instead. "More Glossary Terms" and related-item lists were rendering as unstyled run-on text with headings sitting directly on top of list content, on every page using those templates, not just the new FlashCore ones. A follow-up portfolio-wide audit, run with parallel research agents checking one site each, found the same pattern on AS400IBMSystems.com and confirmed the other five sites were structurally clean. That's a bug that had nothing to do with FlashCore Modules and would have kept shipping unstyled pages indefinitely if the content campaign hadn't gone looking at the actual templates first.
Second, round two proved the redundancy strategy compounds instead of decaying. IBM's own field had moved (FCM5 launched between the two rounds), and the first round's "four generations" claim was now wrong in three places at once, on AS400System.com's buyer guide and glossary and on BigTechReseller.com's glossary. A single-site strategy would have had one page to fix. Seven sites meant seven places the same error could have lived, and it did, at three of them. The fix was the same discipline applied twice: verify every generation and feature code against primary IBM sources and independent trade press before publishing, not once at launch and never again.
The competitive-analysis data changed round two's shape entirely. Instead of guessing which FlashSystem models mattered, the campaign built exactly the models Midland Infosys's own warehouse and GSC data showed were ranking gaps or proven-demand terms, deliberately building on the portfolio's own separate properties rather than touching or linking Midland at all, since Midland is an existing Krisada client explicitly excluded from the buildout.
What can be reused?
Publishing the same facts on seven sites isn't redundant if each site is answering a different question. It's coverage. But redundancy across seven properties means an error, or a fact that goes stale, also has seven places to hide. The campaign's second round wasn't a content refresh. It was a fact audit that happened to also add content, and finding it required checking IBM's own field state again, not assuming the dataset from eleven days earlier was still complete.
Keep Browsing the Case Library
Move from this evidence file back into the full proof system.