Doing Data-Driven Content Maintenance

on YouTube ↗33m

" summary

An automatic script runs against Looker to flag high‑traffic articles not updated in 18 months, items get filed in Jira, and edits split into opportunistic versus tech‑debt. Success is defined by comparing three months before and after (ignore the month of change) against a median threshold of 30%. Outcomes show 53% successful maintenance generating about 1.2 million additional monthly visitors, tech‑debt edits perform better than opportunistic ones, and the program consumes roughly 25% of content team time.

AI-written summary · about

" chapters (15)

1. 0:00Introduction
2. 1:26Data-driven content maintenance
3. 3:54Content library workflows
4. 6:00KPIs and maintenance types
5. 9:58Quincy's qualitative insights
6. 12:22Measuring maintenance efforts
7. 14:27Data views and outcomes
8. 15:30Tech debt results
9. 16:11Opportunistic maintenance insights
10. 17:14Piece-by-piece outcomes
11. 19:42Program view and examples
12. 20:24Key takeaways
13. 23:12Analysts and writers collaboration
14. 24:16Revisiting old content
15. 25:38Q&A

" related talks

• 30mData Driven DevEx at LinkedInMax Kanat-Alexander · IDPCON NYC 2024
• 41mLong-term Documentation MaintenanceKenzie Woodbridge · Craft Conference 2021
• 31mHow We Deprecated 500+ Articles: Cleaning Up Documentation at ScaleAileen Mary · Write the Docs Portland 2026
• 20mPrioritizing Technical Debt as if Time and Money MattersAdam Tornhill · GOTO Chicago 2020
• 38mThe Power of Two: How Support and Docs Teams Transform Technical Content QualityYanjie Niu · Write the Docs Portland 2025
• 39m7 Secrets of Maintainable CodebasesAdam Tornhill · GOTO Stockholm 2016
• 19mLeveraging Analytics for Technical Efficiency and Enhanced User ExperienceRenisha Fernandes, Muskan Prajapati · SREcon23 Asia/Pacific
• 27mA content manager's guide to crowdsourcing the docsBecky Todd · Write the Docs Prague 2017