Caroline Gilbert — Director of Content & Editorial at Angi — joins Ross Hudgens to explore how one of the largest home-services brands is using AI to scale, transform, and optimize content operations.
They discuss how Caroline’s team doubled Angi’s content library to 12,000+ articles, how AI helps maintain top-funnel visibility amid shifting search behaviors, and the nuanced balance between human expertise and generative automation.
Caroline breaks down the three pillars of Angi’s AI framework — generation, transformation, and operations — and how they’ve automated everything from brief creation to CMS ingestion while keeping a strong editorial voice. Plus: lessons learned from early AI missteps, real examples of disclosure testing, and how AI is reshaping both content creation and search behavior itself.
Show Notes
0:08 – From cost guides to “everything home”: expanding Angi’s library
1:00 – Moving up-funnel: education, troubleshooting, light DIY, and when to hire
2:14 – Top-funnel ROI drops as AI overviews change search behavior
2:56 – Why Angi adopts gen-AI: maintain helpful content without heavy human lift
3:27 – Staying visible where intent is early but brand value matters
4:45 – Content becomes commodity: deciding where humans add unique value
5:55 – Evergreen “how-to” steps vs. human nuance (risks, gotchas, pro POV)
7:18 – How AI is used: full refreshes, component injections, data-driven sizing
9:01 – The real trade-off: human time vs. technology time
9:26 – Scale context: 5k → 12k pages (growth mostly human-led)
9:41 – Three pillars: generation, transformation, operations
10:38 – Ops wins: briefs, automated edits/QA, CMS ingestion, minutes → hours
12:21 – Designers focus on UX polish when ingestion is 98% done
12:58 – Where AI struggled: risky DIY (HVAC/roofing), brand POV, nuance
15:02 – Licensing content is messy (state/county rules): tighter human review
17:19 – Design: generate/transform/streamline + “vibe coding” prototypes
20:18 – Pilot first: small cohorts, guardrails, measure engagement/conversions
21:22 – Disclosure tests: placement/tone matter; users split ~50/50 on trust
25:53 – “Tools, time, or talent”: when to push pro hire messaging
27:39 – Differentiators: pro interviews + Angi’s original homeowner cost data
28:07 – AI helps editors analyze like analysts; humans shape the story
29:19 – Looking ahead: AI shifting search from answers to utilities/experiences
