The Athletic - The New York Times

Project Overview

The Athletic is a subscription sports publication owned by The New York Times, serving a very large catalogue of articles, videos and entity pages. Work here was not about building something new: it was about making an enormous existing platform legible to search engines and fast in the hands of real readers on real phones. That means structured data that actually produces rich results, an indexing strategy that stops the site competing with itself, and a class of React hydration defect that only ever showed up under production traffic. Because the product is the client's, there is no public code and no live URL to demonstrate - what is described here is the work, not the codebase.

Challenges

• Making search engines correctly understand a very large catalogue of stories, videos and entities spanning many different page types.
• Consolidating ranking signals across duplicate and broken URLs accumulated over years of migrations.
• Diagnosing SSR hydration mismatches that produced layout shift and inconsistent server output across multiple shared page templates.
• Reproducing defects that appeared only in production, under real traffic, on real devices.
• Improving Core Web Vitals on article and video templates without regressing the editorial and subscription features built on top of them.

Solutions

• Implemented Schema.org structured data (JSON-LD) across article, video and entity page types, and verified in Google Search Console that stories, videos and key entities surfaced correctly in search and rich results.
• Rebuilt the crawling and indexing strategy - XML sitemaps, canonical and metadata handling, semantic markup and migration redirect maps - consolidating ranking signals and clearing duplicate and broken URLs out of coverage reports.
• Diagnosed and eliminated Next.js and React SSR hydration mismatches, a recurring class of defect spanning multiple shared page templates.
• Improved Core Web Vitals across article and video templates, resolving the LCP and CLS regressions that showed up in mobile field data.
• Instrumented Datadog RUM with custom context and Session Replay, turning hard-to-reproduce hydration and runtime errors into alertable events with full session context.

Key Features

• Schema.org / JSON-LD Coverage - Article, video and entity page types marked up and validated against real search results.
• Indexing Strategy - Sitemaps, canonicals, metadata and redirect maps rebuilt as one coherent system.
• Hydration Debugging - A recurring SSR defect class traced across shared templates and fixed at the source.
• Core Web Vitals - LCP and CLS regressions resolved on the templates that carry the most traffic.
• Production Observability - Datadog RUM and Session Replay wired with custom context for alerting on real user sessions.