Skip to main content

Getting started with Bïrch AI and Meta MCP

Learn what Bïrch AI can do, how it works, current limitations, and best practices for getting accurate, actionable answers from your advertising data

Written by Alsu

Managing advertising performance often means jumping between dashboards, comparing metrics, reviewing automation rules, and manually looking for optimization opportunities.

Bïrch AI helps you do all of that with simple questions. Whether you need to understand why ROAS dropped, find your best-performing campaigns, audit automation rules, or generate new automations, Bïrch AI turns your workspace data into actionable recommendations in seconds.

Unlike general-purpose AI tools, Bïrch AI is built specifically for performance marketers and understands advertising platforms, marketing metrics, and automation workflows out of the box.

Important note: Bïrch AI is available exclusively for PRO subscribers.

Accessing Bïrch AI

To open Bïrch AI:

  1. Click Bïrch AI in the top-right corner of the platform.

  2. The chat window will open with suggested prompts to help you get started.

  3. Ask a question about your advertising performance, automation setup, or workspace data.

What can Bïrch AI do?

Performance analysis

Bïrch AI can analyze performance data across:

  • Meta Ads

  • TikTok Ads

  • Google Ads

  • Snapchat Ads

  • Facebook Pages

  • Google Analytics

It can help you:

  • Understand performance trends

  • Compare results across platforms

  • Investigate performance changes

  • Review ROAS, CPA, CPC, CPM, CTR, and revenue metrics

  • Identify optimization and scaling opportunities

Automation analysis

Bïrch AI can review your active Meta automation rules and help you:

  • Identify overlapping or conflicting rules

  • Audit your automation strategy

  • Recommend improvements

  • Suggest additional automations

  • Explain existing rule logic

Rule generation

Describe your goal and Bïrch AI can generate automation rules directly in Bïrch. For example, you can ask Bïrch AI to:

  • Create a scaling rule

  • Generate a CPA protection rule

  • Suggest a starter automation strategy for a new account

  • Build a creative testing workflow

Generated rules are created as drafts, allowing you to review and edit them before publishing.

Get the most out of Bïrch AI

Bïrch AI works out of the box — but two optional connections give it significantly more context, leading to sharper and more relevant insights.

Connect Meta Ads MCP

Meta recently launched Meta Ads AI Connectors, an official integration that lets AI tools like Bïrch AI connect securely to your Meta ad account via the Meta Ads MCP (Model Context Protocol) server. Once connected, Bïrch AI has read-only access to your live campaign performance — so it can analyze what's actually happening in your account, not just give generic advice.

With Meta MCP connected, Bïrch AI gains access to:

  • Campaign, ad set, and ad-level performance metrics (spend, ROAS, CPA, CTR, impressions, and more) and their trends over time

  • Changes made to campaigns, ad sets, and ads (mirrors Ads Manager history)

  • Budget and delivery status across active and paused campaigns

  • Meta Opportunity Score

  • Meta Business Help Center

  • An overview of the advertiser's business context and marketing funnel to guide optimization goals.

  • Alerts about deviations in ad performance: unusual patterns, spikes, or drops.

  • Insights on which ads performed stronger in the auction and which factors (bid, ad quality) to optimize.

  • Ad account industry benchmarks and campaigns, ad set performance against aggregated benchmarks from similar advertisers, optionally filtered by spend tier /goal.

  • Meta Ad Library for ads matching given criteria (public data)

Bïrch AI uses this data in read-only mode. It cannot create, edit, pause, or delete any campaigns on your behalf.

👉 For prompt examples, see our Bïrch AI Prompting Guide.

How to connect Meta Ads MCP

Setup requires no developer credentials or API tokens.

  1. Go to Settings → Platforms in Bïrch and click Meta Ads.

  2. Click on Profiles

  3. Next to your personal profile, click "Connect MCP" (Note: MCP connection is only available for personal profiles, not business profiles):

  1. You'll be redirected to Meta, where you grant MCP access to Bïrch using the standard OAuth flow — the same login you use for Ads Manager.

Your Bïrch rules and automations continue to work exactly as before — the Meta MCP connection simply gives Bïrch AI more context to give you better answers, faster.

Add Business Context

Business Context lets you save reusable information about your business — your goals, funnel, KPIs, or anything Bïrch AI should always know — so you don't have to repeat it in every conversation. It doesn't count toward your usage limits.

Best practices

To get the most accurate results:

  • Be specific about what you want to analyze.

  • Mention ad account names when working with multiple accounts.

  • Ask focused questions instead of combining many requests into one prompt.

  • Use rule names instead of rule URLs when analyzing automations.

  • Ask follow-up questions to build on previous responses.

Current limitations

While Bïrch AI supports a wide range of performance and automation analysis tasks, there are a few limitations to keep in mind.

Large accounts

In accounts with very large numbers of campaigns, ad sets, or ads, Bïrch AI may ask you to narrow the scope of your request to ensure accurate results.

Automation analysis

Currently, Bïrch AI can analyze Meta Ads automation rules only. Rule analysis is limited to active rules connected to ad accounts and does not include rule activity logs or folder information.

Not currently supported

The following features are not currently supported:

  • Automation rule analysis for TikTok, Google Ads, or Snapchat

  • Cohort analysis

  • Google Sheets-based custom metrics

  • Custom metric formulas

  • Top/Bottom ranking requests

  • Simultaneous conversations across multiple browser tabs

Privacy & Security

Bïrch AI uses data available within your Bïrch workspace, including advertising performance data and automation rule configurations. Your data is processed only to answer your requests and is never used to train external AI models. For more information, please refer to our Privacy Policy.

FAQ

Which large language model (LLM) does Bïrch AI use?

Bïrch AI is powered by Google's Gemini 3.0 Flash model, chosen for its speed and efficiency in processing complex campaign data and delivering precise, actionable insights. The Bïrch team is continuously evaluating additional state-of-the-art models to expand capabilities and ensure you always get the best possible experience.

What kind of data does Bïrch use to power its AI?

Bïrch AI is powered by your connected ad account data - including campaign performance metrics, spend breakdowns, and automated rule configurations - all organized to deliver fast, accurate insights and smarter decision-making for your paid media strategy.

Does Bïrch share my data with LLM providers?

Bïrch AI is built with your data privacy and security as a top priority. LLMs are engaged solely to process your queries and never use your data for model training or service improvement. All data handling fully complies with our Privacy Policy and strict enterprise privacy standards.

How is it different from Meta Ads MCP?

Meta Ads MCP connects general-purpose AI tools like Claude or ChatGPT to your ad account - but it requires setup and has no built-in paid media expertise. Bïrch AI is purpose-built for performance marketers. It understands your rules, flags logic errors, and works across your entire workspace out of the box. No setup, no coding - just actionable insights from day one.

Can I use Bïrch AI outside of Bïrch?

No, Bïrch AI is exclusively available within the Bïrch platform. It is deeply integrated with your connected ad accounts and workspace data, which means it can only operate within the Bïrch environment to deliver accurate insights and rule suggestions.

Did this answer your question?