157 million clicks. The two-engine model that drives a news property at scale.

Strip out the two spike events in May and June, and the underlying baseline grew from approximately 430,000 daily clicks in November 2023 to 680,000 in July 2024 — a 58% lift over nine months. This is the quieter of the two stories, and the harder one to build.

Topical breadth at depth: the property covers every major topic category in its vertical with sufficient depth that Google treats it as the authoritative entity. This is not about keyword coverage — it is about the knowledge graph recognising the domain as the answer source. Crawl architecture for volume: at a publishing cadence of 50–200 articles per day, the property runs a strict URL taxonomy, paginated content canonicals, and a news sitemap pinged on every publish. Google's crawlers return at a rate that supports indexing within 3–8 minutes of publication for new articles on covered topics.

Headline CTR discipline: the 5.1% aggregate CTR is, in significant part, a headline product. News article titles serve double duty as the title tag and the display headline in Top Stories. The editorial standard enforces: primary keyword in the first four words, specificity over curiosity-bait, and a factual hook that signals completeness. A headline that makes the click unnecessary earns fewer clicks than a headline that makes the click inevitable.

May 19–20, 2024. India's 2024 General Election Phase 5. The property's election coverage section — built with structured article clusters for each phase, constituency-level schema markup, and live-results infrastructure prepared weeks in advance — indexed into Top Stories within 11 minutes of the phase's first significant search volume. Daily clicks on May 19 reached approximately 1.2 million. May closed at 24.5 million clicks.

June 8 produced a second event: post-election result analysis and political transition coverage. 580,000 incremental clicks in a single day. June closed at 23.8 million.

These events did not happen because the editorial team worked faster. They happened because the infrastructure was built before the moment arrived. Three technical prerequisites: sub-10-minute indexing via the Indexing API for articles matching the top-100 tracked entity list; pre-built template pages for predictable events (election phases, budget days, IPL match days) that are created, indexed, and internally linked days before the event fires; and entity schema on every article — NewsArticle with headline, datePublished, dateModified, author, and about properties mapped to the relevant entities.

The spike contribution was 16.7 million clicks — 10.6% of the 9-month total — from two event windows spanning a combined 8 days. The baseline contribution was 140.3 million clicks from the remaining 261 days.

The risk profile matters here. A property whose annual traffic budget is built around three event windows is one missed event away from a revenue shortfall. The correct model: build the baseline so conservatively that the spikes are upside, not budget.

The March 2024 traffic dip was a technical indexing regression, not an algorithm event. Around March 20, 2024, daily clicks dropped approximately 25% over four days. Root cause: an erroneous robots.txt update that restricted Googlebot from crawling a content subfolder containing 8,000 articles. The block was live for 72 hours before detection. Recovery took 12 days post-fix. The lesson: changes to crawl directives on a property at this publishing velocity require a 10-minute post-deploy crawl verification as a non-negotiable release step. It did not exist before March. It does now.

Video content indexing remained an unsolved problem. The property produces 40–60 video assets per week. None of them rank in Video Search results despite technically correct schema deployment. Current hypothesis: YouTube hosting is required for Video Rich Results on news properties at this query intent level, and self-hosted video infrastructure is not treated as equivalent.