Skip to main content

Documentation Index

Fetch the complete documentation index at: https://adadvisor.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

The AdAdvisor playbooks skill ships five named multi-week playbooks for the moments where a senior Meta (Facebook) media buyer follows a defined cadence over weeks. Some situations need a sequence, not a one-shot workflow. The skill auto-loads when you mention the situation, walks you through the relevant phase, and tracks where you are in the timeline.

​
When should I use the playbooks skill?

Triggers on phrases like:
  • “BFCM”, “Black Friday”, “Cyber Monday”, “holiday campaign”
  • “Launch a new product”, “product launch”, “drop”
  • “Lead gen account”, “lead form”
  • “Banned”, “account banned”, “account restricted”
  • “Recover account”, “pixel got banned”
  • “Scale from X to Y”, “1k to 10k”, “ramp plan”

​
What are the five AdAdvisor playbooks?

Synthesizes Ray Jang (Pilothouse), Andrew Foxwell, Common Thread Collective, Triple Whale, and Motion’s 2024 to 2025 BFCM postmortems.Why BFCM is structurally different: CPM spikes 50% to 80%, conversion rate spikes 30% to 100% (urgency), creative fatigues in 2 to 4 days instead of 7 to 10, and CPM inflation is unavoidable. The win comes from being prepared.Timeline:
  • T minus 8 to 6 weeks (early October): audience warm-up. Seed retargeting pools, surface BFCM-specific creative concepts at low spend.
  • T minus 4 weeks: structural prep. Build BFCM landing pages, verify pixel, build BFCM-specific custom audiences, pre-build retargeting ad sets (paused).
  • T minus 2 weeks: structural freeze. No new pixels, no new campaigns, no audience swaps. Lock in learning.
  • T minus 1 week: final creative QA, budgets +30% to 50% for pre-Thanksgiving warm-up.
  • BFCM weekend (Thursday to Monday): ramp budget 200% to 400% vs baseline. Hour-by-hour pacing plan. Expect CPM +50% to 80%. Use bid caps if margins are tight.
  • Cyber Week: last-chance creative, urgency messaging, abandoned-cart layer.
  • Post-BFCM (Dec 1 to 15): cool down. Reduce budgets 30% to 40%. Reactivate new customers with welcome series.
Budget split during peak: 50% to 70% evergreen plus 30% to 50% BFCM-specific (Ray Jang playbook).
The sequence for launching a product Meta has never seen perform.
  • Pre-launch (T minus 2 weeks): TOF video views and engagement to seed retargeting pool with people who have shown interest.
  • Launch week (Day 1 to 7): ASC plus manual prospecting CBO side-by-side (with exclusions). 5 to 8 distinct concept angles: problem-aware, solution-aware, product-feature, founder-story, comparison, UGC testimonial.
  • Validation (Day 7 to 14): identify true winners (3+ days stable CPA). Tight kill rules. At launch you do not have benchmark data, so kill at 3x target CPA with zero conversions.
  • Scale (Day 14+): standard scaling. 20% vertical, expand horizontally.
Bid strategy: Lowest Cost for first 7 days (max signal collection), then layer Cost Cap once CPA stabilizes.
End-to-end lead-gen flow.
  1. list_pages to confirm the Page has both ADVERTISE and MANAGE_LEADS tasks.
  2. create_lead_form with mandatory privacy policy URL. follow_up_action_url is also required (undocumented, but Meta rejects with subcode 1892085 otherwise).
  3. create_campaign with objective='OUTCOME_LEADS'.
  4. create_adset with optimization_goal='LEAD_GENERATION' and destination_type='ON_AD'.
  5. create_creative with lead_gen_form_id, CTA = SIGN_UP, APPLY_NOW, GET_QUOTE, LEARN_MORE, SUBSCRIBE, or DOWNLOAD.
  6. create_ad. No conversion_domain needed for lead-form ads.
  7. Activate leaf-up.
Quality filter: require 2 to 3 form fields beyond email to filter junk. CRM-fired CAPI optimization (firing a “Qualified Lead” event from your CRM) is the highest-quality optimization signal.
  1. Don’t appeal repeatedly. Each rejected appeal makes the next harder. Submit ONE appeal through Account Quality with documentation.
  2. While the appeal pends, warm up a backup ad account (always have one pre-built in Business Manager). Move pixel to backup. Restart with low budgets (25to25 to 50 a day).
  3. New-account discipline: 50% normal budget for 7 days. No flagged creative. Conservative claims. No policy-edge content.
  4. Identity-confirm everything: domain verification, business verification, payment method match.
Pixel-specific recovery: if pixel events stop firing, verify CAPI is still firing via adadvisor:get_pixel_health. Campaigns can survive a pixel outage if CAPI is healthy. Then debug client-side. GTM or theme updates often break it.
Three stages (Disruptive Digital, Pawliw, Theriot synthesis).
  • Stage 1 (1kto1k to 2.5k, 2 to 4 weeks): vertical 20% on confirmed winners. Add 1 to 2 broader ad sets. Lock 3 to 5 creatives.
  • Stage 2 (2.5kto2.5k to 5k, 2 to 6 weeks): horizontal scaling (LAL expansions, geo). ASC plus manual with exclusions. Weekly creative production. Track MER.
  • Stage 3 (5kto5k to 10k, 4 to 8 weeks): bid-cap diversification (Tichenor method). Target ROAS (Faris method). 10% to 15% brand/awareness layer. 4+ new concepts per week.
Most operators stall at 5kto5k to 10k because they protect a misleading ROAS instead of optimizing for marginal profit on marginal spend. The skill makes the trade-off explicit at every stage transition: CPA inflation of 10% to 20% is expected. The question is whether marginal revenue exceeds it.

​
How does the playbook auto-pick work?

The skill listens for the trigger phrase and walks you through the relevant phase only. It does not dump the entire playbook in one response. At each milestone it checks in and confirms before advancing.

​
What anti-patterns does the playbooks skill prevent?

The lead time is 6 to 8 weeks for audience warm-up and creative production. If it is too late, the skill does a compressed version honestly rather than pretending the timeline still works.
5 to 8 distinct angles is the minimum for a launch test. The skill enforces this before approving the launch.
Each rejected appeal makes the next harder. The skill enforces “one careful appeal with documentation, then move to backup”.
6 to 12 weeks per the stage plan is the realistic timeline. Faster requires accepting CPA inflation and tightened kill rules. The skill makes the trade-off explicit.
Optimization on Meta’s “Lead Submitted” event alone leaks quality. The skill upgrades to a CRM-fired “Qualified Lead” event when the CRM is connected.
Pacing, creative refresh cadence, bid strategy, exclusion logic, post-BFCM cool-down. All part of the play. The skill enforces the full sequence.

​
How do I install the playbooks skill?

Claude Code

Ships with the full plugin install.

Claude.ai & Desktop

Download adadvisor-playbooks.zip and upload via Settings, Capabilities, Skills.

​
Frequently asked questions

6 to 8 weeks before BFCM (early to mid October). The audience warm-up phase (TOF video views and engagement to seed retargeting pools) needs 4 to 6 weeks to deliver enough signal. By T minus 2 weeks you should freeze structural changes and lock in learning. Starting later means a compressed playbook and less retargeting fuel.
Pre-launch (T minus 2 weeks): seed retargeting with TOF video views. Week 1: ASC plus manual prospecting CBO with exclusions, 5 to 8 distinct creative angles. Day 7 to 14: tight kill rules (3x target CPA with zero conversions kills), identify winners. Day 14+: standard scaling. Use Lowest Cost for the first 7 days, then layer Cost Cap once CPA stabilizes.
Submit ONE appeal through Account Quality with documentation. Do not appeal repeatedly. Each rejected appeal makes the next harder. While the appeal pends, switch to a backup ad account (Business Manager makes this easy if you pre-built one). Restart at 50% normal budget for 7 days with no flagged creative. The playbooks skill walks you through the full recovery.
Confirm your Page has the ADVERTISE and MANAGE_LEADS tasks. Create the lead form with a privacy policy URL and the (undocumented but required) follow_up_action_url. Create the campaign with objective='OUTCOME_LEADS'. Create the ad set with optimization_goal='LEAD_GENERATION' and destination_type='ON_AD'. Create the creative with lead_gen_form_id and one of the allowed CTAs (SIGN_UP, APPLY_NOW, GET_QUOTE, LEARN_MORE, SUBSCRIBE, DOWNLOAD). Activate leaf-up.
6 to 12 weeks across three stages. Stage 1 (1kto1k to 2.5k) is 2 to 4 weeks of vertical scaling. Stage 2 (2.5kto2.5k to 5k) is 2 to 6 weeks of horizontal expansion plus ASC. Stage 3 (5kto5k to 10k) is 4 to 8 weeks of bid-cap diversification and Target ROAS portfolio. Faster timelines require accepting CPA inflation and tightened kill rules.
Auction competition. Every advertiser is bidding harder during BFCM, and Meta’s auction is a second-price auction so CPMs rise across the board. The BFCM playbook expects 50% to 80% CPM inflation during peak. The compensating factor is that conversion rates also spike 30% to 100% due to urgency, so total ROAS can hold or improve if you prepared the audience warm-up phase correctly.

Launch

Building block for the product-launch and lead-gen playbooks.

Scale

Building block for the 1kto1k to 10k playbook.

Audit

Run before starting any playbook to confirm the account state.
Last modified on May 13, 2026