Summary
Problems
- In a freelance agent website, many pages gradually got unindexed from Google's search results unexpectedly
- The tech team had been trying several fixes, but none of them achieved meaningful results
- The marketers raised concerns about the diminished search performance
Actions
- Investigated from multiple angles - local code behavior, production deployment, and Google Search Console, stepping into the marketers’ toolset to bridge gaps in insight
- Worked around limited debugging capabilities by deploying careful production changes to test hypotheses, as the crawler's behavior couldn't be replicated locally and logging provided only truncated Search Console outputs
- Iteratively deployed small, low-risk changes to validate each assumption about the crawler’s behavior
- After testing fixes almost daily, I narrowed the issue down to an unexpected cause — complex client-side JavaScript behaving unreliably in Google’s unique crawler environment
- We removed JavaScript entirely for the crawlers and confirmed that there was no longer unexpected behavior
Results
- After the JavaScript removal, I requested Google for re-indexing of the affected pages, and all of them got indexed again within a few days
- I carefully monitored the indexing status for a few weeks and promptly addressed any questions from the marketers to help rebuild their confidence after the months-long issue
- The index rate recovered from previous 60~80% to a stable over 99% and the marketers reported that the search performance got back to the previous level
- Early recovery indicators were strong: impressions climbed more than 20% above the suppressed May–August baseline, and visits continued rising, reaching nearly 85% above that level by late November
- The issue was shared company-wide as an important case study so that other teams could fix similar issues more efficiently or even avoid before happening in the future
My Strengths Demonstrated In This Work
-
Focusing on fixing the business impact:
I stuck to the business objective instead of staying on the technical responsibility by keeping communication with the marketers to minimize their concerns from the long-lived issue -
Doing Everything Needed:
I went through manual and repetitive tasks such as checking indexing status for hundreds of pages and requesting re-indexing for each of them on Google's Search Console to ensure the issue was fully resolved instead of waiting for others to do it or crawlers to slowly re-crawl the pages naturally because it impacted the business by taking longer -
Persistent problem-solving:
I kept trying various approaches with over 50 tasks during 2 months to find the definitive fix without giving up, even though many earlier attempts produced limited progress -
Analytical thinking:
Despite limited visibility into the crawler environment, I systematically narrowed down the root cause to client-side JavaScript execution through logical analysis of each trial