When 270 Articles Produce 65 Indexed Pages
Case Study Summary
This case study examines a young SaaS company in the relationship advice space that implemented programmatic AI-generated content. Despite publishing over 270 blog articles, Google indexed only approximately 65 pages—of which 23 were tag pages—meaning roughly three quarters of their content portfolio was effectively ignored by search engines.
Content Inflation
Content inflation occurs when published content grows faster than the unique information contributed. The study identifies several patterns that allow categorizing content inflation:
- Repetitive structures, including openings and headers
- Commodity content (especially titles)
- FAQ farms
The site exhibited commodity content produced without visible authority or demonstrated engagement with the topic. The repetitive structure of generic titles, with no cited real persons or actual authors—particularly concerning given the medical/psychology domain—contributed to the indexation problem.
Portfolio Perspective
A single competitor review page drove 30% of organic clicks. The entire competitor-focused cluster (reviews, advice, brand-vs-competitor content) accounted for roughly half of all non-branded traffic. No other portfolio—including the broader blog, service pages, or informational content—could compare with the competitor-focused portfolio.
The core problem: the website accumulated content assets while simultaneously losing portfolio quality outside the competitor-driven content cluster. All clusters except the competitor-focused ones failed to demonstrate incremental value, creating a negative feedback loop where "bad" signals from poor-performing pages negatively impacted performance of all pages within the segment. When older pages "sunk," newer ones shared the same fate.
Recommendations
The analysis produced two key recommendations:
- Cut down the competitor-related cluster by consolidating to 1 page per competitor, redirecting or consolidating all other competitor content. This was deemed low-risk for traffic loss
- Secure indexation recovery of native article clusters by stopping new content publication and focusing on trimming low-performing content through redirecting, removing, and consolidating articles, while manually revising blog titles
Signals Google May Detect
The study identified potential signals Google may detect that contribute to poor indexation:
- Weak uniqueness
- Thin differentiation
- Low perceived usefulness
- Lack of demonstrated expertise
- Low engagement
Why This Matters in the AI Era
Generative AI search systems require information gain, unique observations, and differentiated sources. When ten pages say the same thing, the eleventh adds little value.
Portfolio Utilization Rate
The study proposes a new metric: Portfolio Utilization Rate—calculated as indexed pages divided by total published pages. In this case: 65 indexed / 270 published = 24% effective utilization. This metric is critical to monitor because of the high risk of larger deindexation as AI content strategies backfire.
Core Advice
De-commoditize content by moving away from generic content toward demonstrating what the company's app can do versus other similar tools. This demonstrates maker-specific content value and results in better indexation and more keywords per page.