Live Bearded - Olympus

Midas CANI — Sources & Knowledge Review

Meta Ads Report  •  March 18, 2026

Status: CANI Files Reviewed ✅

Files confirmed in team/midas/references/cani/. Here's what's there and what needs to happen next.


What's in the Files Right Now

sources.md — WHO TO FOLLOW

The sources file is solid. Here's the full practitioner list already loaded:

NameWhereWhy They Matter
Jon Loomerjonloomer.com / @jonloomerAdvanced Meta mechanics, Power Editor deep dives
Andrew FoxwellFoxwell Digital / @andrewfoxwellDTC Meta buying, scaling strategies
Nick ShackelfordStructured Social / @iamnickshackeComm Meta, creative strategy
Barry Hott@hott_barryDirect response creative, conversion optimization
Cody Plofker@codyplofkerDTC growth, Meta + email integration, publishes real account data
Zach Murray@zachemurrayDTC eComm, Meta performance
Taylor HolidayCommon Thread CollectiveMER frameworks, DTC Meta strategy — highly relevant to LB's situation
Andrew FarisAJF GrowthDTC operator, very current on Meta mechanics in 2025-2026

Taylor Holiday and Andrew Faris are NOT in the current sources.md. They need to be added. These are two of the most relevant voices for LB's specific situation (MER-first thinking, DTC scale, current Meta landscape).

Official Sources

SourceURLCadence
Meta Business Blogbusiness.facebook.com/business/newsDaily
Meta Ads Help Centerfacebook.com/business/helpAs needed
Meta Blueprintfacebookblueprint.comMonthly
Meta for Developers Blogdevelopers.facebook.com/blogWeekly

Research Cadence (Already Set)

FrequencyWhat
DailyTwitter/X (breaking insights, algorithm reactions), Meta Business Blog
WeeklyJon Loomer, Andrew Foxwell, Barry Hott, Instagram follows, LinkedIn
MonthlyFull CANI review — update best_practices.md, clear queue.md

What's Empty (Needs to Be Built)

FileStatusWhat Goes There
best_practices.md❌ EMPTYCurrent Meta SOPs — campaign structure, creative, bidding, audience, attribution. This is the core knowledge file and it's blank.
queue.md⚠️ STALE PLACEHOLDERSHas placeholder topics loaded but nothing LB-specific from real sessions yet. Needs active items.
semantic.md⚠️ PARTIALLY LOADEDAccount facts are pre-loaded. Audience patterns, creative patterns, seasonal patterns, bidding patterns are all empty — need to be built through onboarding and experience.
procedural.md⚠️ PARTIALLY LOADEDCore decision rules are pre-loaded (ROAS below floor, ROAS spike, creative fatigue). Learned rules section is empty — populates through experience.

The Gap That Matters Most Right Now

best_practices.md is empty. That's the file where current Meta knowledge lives — what's actually working in 2026, how the algorithm behaves now, what bidding strategies are effective, what creative formats are winning. Without it, I'm running on pre-loaded priors that may be stale (as today's session already surfaced).

Two ways to fill it fast:

  1. You tell me. You know what Meta looks like right now. A 15-minute conversation where you share what you're seeing, what's changed, what's working — I translate that directly into best_practices.md entries. Fastest path.
  2. Atlas builds a research pull. Pull current content from the practitioners in sources.md and distill into best practices. More thorough, takes longer.

Immediate Action Items

PriorityActionOwner
🔴 HIGHAdd Taylor Holiday + Andrew Faris to sources.mdAtlas or Midas on next update
🔴 HIGHFill best_practices.md — start with what Mink shares about current Meta landscapeMink → Midas
🟡 MEDIUMUpdate queue.md with active research questions from today's sessionMidas
🟡 MEDIUMBuild out semantic.md audience/creative/seasonal patterns during deep diveMidas (during onboarding)

Bottom Line

The CANI skeleton is there. The practitioner list is solid (needs 2 additions). The knowledge content is empty — best_practices.md especially. The fastest way to fix that is you talking to me right now. Tell me what you know about how Meta actually works in 2026 and I'll turn it into structured knowledge that every future session builds on.