agentsclimarketplace

Microsoft campaign setup and optimization

Skill scumunna/programmatic-skills/skills/microsoft-campaign-setup-and-optimization

Build, QA, and optimize Microsoft Advertising campaigns across Search, Performance Max, Shopping, Audience, conversion tracking, UET, asset groups, search themes, product feeds, budgets, imports, brand controls, learning, and reporting. Use when the user asks how to improve Microsoft Ads or Bing campaigns beyond UET or PMax-only setup.From its SKILL.md

Install
npx -y skills add scumunna/programmatic-skills --skill microsoft-campaign-setup-and-optimization

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

8.8 KB, ~1.8k tokens by cl100k_base, as published. Nobody here has run it

Microsoft campaign setup and optimization

Help an operator build and improve Microsoft Advertising campaigns across Search, Performance Max, Shopping, Audience, and feed-led retail. Microsoft Ads often starts from a Google import, but the account still needs Microsoft-specific tracking, feed, budget, search, and placement checks. This skill supports campaign operators. It does not replace them or approve live account changes.

For UET, Conversions API, offline conversions, and consent, use microsoft-uet-and-conversions-api. For Performance Max and Shopping object detail, use microsoft-performance-max-and-shopping.

When to use this skill

  • "Set up Microsoft Ads for this account."
  • "Import Google campaigns into Microsoft and QA them."
  • "My Microsoft campaign is not spending."
  • "Should I use Search, Performance Max, Shopping, or Audience?"
  • "How do I optimize Microsoft PMax or Shopping after launch?"
  • "What should I check in UET before scaling?"
  • "How do I use search themes, audience signals, product feeds, and brand controls?"
  • "How do I read Microsoft results next to Google?"

Boundaries with sibling skills:

  • UET, conversion goals, Conversions API, and offline conversions: microsoft-uet-and-conversions-api.
  • PMax and Shopping asset groups, feeds, listing groups, and bid strategy: microsoft-performance-max-and-shopping.
  • Google Ads equivalent: google-ads-performance-max, google-ads-shopping-and-feed, and search skills.
  • Cross-platform reconciliation: cross-platform-conversion-reconciliation.
  • Search Ads 360 engine management: sa360-account-structure-and-engines.

Quick reference

GoalCampaign pathOperator priorities
Capture existing demandSearchKeywords, match types, negatives, RSA assets, UET conversions
Retail catalog scalePerformance Max or ShoppingMerchant Center feed, product scope, asset groups, ROAS target
Product-group bid controlShoppingProduct partitions, priorities, enhanced CPC or manual control
Broad automated conversion growthPerformance MaxConversion goal, asset groups, search themes, audience signals, final URL expansion
Native/display reachAudience campaignsContent targeting, audience segments, creative, placement quality
Google migrationImport then QABudgets, goals, feed mapping, tracking, negatives, brand controls

Core process

  1. Confirm tracking first. UET, conversion goals, revenue values, consent handling, and offline conversion imports must be correct before conversion bidding. If UET is not healthy, stop and use microsoft-uet-and-conversions-api.
  2. Choose campaign type by control need. Search gives keyword control. Shopping gives product-group control. PMax gives automation across Microsoft inventory. Audience gives native and display reach. Do not use PMax when the client needs keyword-level or product-group bid control.
  3. QA any Google import. Imports are useful, not trustworthy by default. Re-check budgets, bid targets, conversion goals, Merchant Center store, product feed, URL tracking, locations, ad schedules, negatives, and policy status before unpausing.
  4. Build the feed foundation for retail. Microsoft Merchant Center store, catalog freshness, product IDs, availability, price, image, GTIN or MPN where required, and landing URL quality decide Shopping and retail PMax scale.
  5. Build search with explicit intent. Use exact and phrase for proven terms, broad only with conversion tracking and negatives in place, and keep brand separate when budget and reporting require it.
  6. Build PMax with strong inputs. Asset groups, search themes, audience signals, product listing groups, final URL expansion, page feeds, brand lists, and conversion value decide quality. Microsoft AI controls much of the delivery.
  7. Set budgets and targets sanely. A target CPA or ROAS that is too tight at launch blocks delivery. Let campaigns gather data, then add targets based on actual conversion economics.
  8. Set brand and content boundaries. Protect brand Search from PMax cannibalization, set URL expansion intentionally, and use content controls when placements or sensitive categories matter.
  9. Launch with a QA gate. Check tracking, budgets, dates, bids, asset approvals, feed disapprovals, policy, URLs, and naming before spend.
  10. Optimize one lever at a time. Creative, search terms, negatives, product groups, feed fixes, target changes, and budgets should be logged and read on a stable window.
  11. Reconcile with the source of truth. Microsoft-attributed conversions will not match Google, GA4, CRM, or the warehouse exactly. Explain the method and reconcile before client reporting.

Decision rules and thresholds

  • Do not scale without UET confidence. Conversion bidding depends on a working UET or offline conversion feed.
  • Search stays separate when query control matters. PMax can extend reach, but it does not replace a strong Search structure for brand and high-intent terms.
  • Shopping stays separate when product control matters. Use feed-based Shopping when product partitions, priorities, and per-group bids are required.
  • Imports need verification. Never assume a Google budget, target, URL, negative, or conversion goal imported into the right Microsoft equivalent.
  • Feed fixes often beat bid changes. Disapprovals, stale prices, poor images, missing identifiers, and broken URLs cap Shopping and retail PMax before bidding gets a chance.
  • Final URL expansion is a business decision. Leave it on only when the full domain is a valid paid destination, or use opt-out and page feeds.
  • Step targets gradually. Tight CPA or ROAS targets at launch can make a campaign look broken when it is simply starved.

Optimization menu

SymptomFirst checksSafer first move
No spendStatus, budget, bid target, UET, policy, feed, audience sizeRemove excessive target or fix policy/feed
High CPAQuery quality, product feed, landing page, conversion goalAdd negatives, fix feed/page, then adjust target
PMax cannibalizes brandBrand list, Search structure, query/category insightApply brand controls and keep Search separate
Shopping low volumeFeed disapprovals, product partitions, bid, priorityFix feed and product scope before raising bids
Import underperformsGoals, budgets, tracking templates, location, negativesRe-QA import mapping
Audience low qualityPlacement/content report, audience segment, creativeTighten content or refresh creative

Templates and examples

Google import QA:

Imported from: Google Search and PMax.
Before unpause: UET goal selected, budgets converted, locations verified, negatives imported, tracking templates working, Merchant Center store linked, product feed approved, brand controls set.
Decision: Keep brand Search separate, launch Microsoft PMax at lower test budget, read after learning.

Retail setup:

Goal: Grow incremental ecommerce revenue.
Tracking: UET purchase event with revenue value.
Feed: Microsoft Merchant Center catalog approved and refreshed.
Campaigns: Shopping for best-seller bid control, PMax for automated expansion.
Guardrails: ROAS, revenue, new customer share, impression share, product disapprovals.

Common pitfalls

  • Importing from Google and unpausing without QA.
  • Running PMax with no UET conversion goal.
  • Letting PMax absorb brand Search and then reporting inflated growth.
  • Ignoring Merchant Center disapprovals and trying to bid around a feed problem.
  • Setting a ROAS target before the account has enough conversion value history.
  • Leaving final URL expansion on for pages that should not receive paid traffic.
  • Reading Microsoft performance as a direct copy of Google performance.
  • Making budget, bid, feed, and creative changes together, then losing causality.

Sources

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,861. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.