This guide covers prompt examples for every major use case in Bïrch AI, along with tips for getting the most accurate and actionable answers. Whether you're analyzing performance, auditing automation rules, or using the Meta Ads MCP connection, you'll find ready-to-use prompts below.
Tips for getting the best results from Bïrch AI
Include only relevant information. Keep your prompts marketing-focused. Birch AI is built for marketing analysis — unrelated requests won't be processed.
Keep prompts specific, not complex. If you hit an error, it's usually because the request is too broad or vague. Break complex questions into smaller, focused prompts — Birch AI works best when given clear, direct instructions.
Working with mobile app campaigns. Birch AI doesn't support cohort-based analysis so far. If your question involves cohorts, you'll get a clear "not supported" response. For more complex requests involving cohorts, the chat may time out — so avoid cohort-related phrasing where possible.
Comparing multiple ad accounts. When requesting a cross-account analysis, always specify account names or use phrasing like "top N performers." For accounts with long names (3–4 words), 2 words are usually enough to match correctly.
Analyzing rules. Currently, Birch AI analyzes only active rules within a workspace. If you want to analyze a specific rule, provide the rule name, not the link so that Birch AI matches it correctly.
Review and edit your prompt before sending. Take a moment to review your message for clarity and completeness to reduce the need for follow-up messages.
How context & limits work
Bïrch AI remembers context from earlier messages in the conversation. Similar prompts may benefit from caching, which can improve response speed and reduce repeated processing.
With Business Context, you can save reusable business information once and use it across conversations without it counting toward your limits.
Bïrch AI prompts examples:
Understand performance (platform & metric-specific)
Account performance
What are the most likely causes of performance drops in my Meta ad account "Name"?
How is overall ROAS across all Meta accounts trending compared to last week?
Provide the Reach breakdown by country for the last 14 days for the ad account "Name".
Campaign, ad set & ad analysis
List ad sets with Cost Per Purchase under $30 for Meta ad account "Name"
What is the CPM trend over the last 30 days for ad account "Name"?
Which campaigns have the highest frequency, and is it affecting CTR?
Identify campaigns with possible creative fatigue based on declining CTR, rising CPM, and increasing frequency.
List my top 3 performing ads by CTR in ad account "Name"
Compare channels
Compare performance across platforms to find where your budget works hardest.
Compare spend, ROAS, and CPA across Meta, TikTok, Google, and Snapchat for Q1
Where are we scaling faster: Meta or TikTok?
Provide a revenue and ROAS report broken out by channel for the last 7 days
Which platform delivers the lowest CPC and CPM, and how do we protect that efficiency as costs rise?
Find growth opportunities
Identify where the budget can be reallocated or scaled more effectively.
Create a leadership-ready dashboard overviewing my monthly results
Where do we have the opportunity to scale the ad budget without sacrificing efficiency?
Identify high-performing ad sets that are underfunded and poor performers consuming budget to recommend immediate scaling and pausing decisions.
Where can I optimize my strategy to increase Meta ROAS?
What are my top-performing Meta campaigns in 2026?
Provide me a report of revenue and ROAS broken out by channel for the last 7 days.
Which ad accounts have the lowest CPCs and CPMs? How can I maintain overall efficiency as ad costs continue to rise?
Analyze automated rules
Review your automation strategy and improve existing rules.
Audit all my Meta rules and identify outdated, duplicated, risky, or conflicting automations
Suggest new rules to expand the current rules setup
Suggest 3 starter rules for a new ad account
Can you suggest the best CPA-based scaling strategy for lead generation campaigns?
Suggest ways to improve my rule "Name"
Can you check the conditions and potential rule conflicts for Rule "Name"? It seems to be prematurely evaluating and pausing ads, just hours after they are activated.
Meta MCP prompts examples:
Account health analysis:
"Pull my account's opportunity score and show me the recommendations. I want to see what Meta is flagging as the biggest gaps — budget, targeting, creative, bidding — and understand which ones are actually worth acting on versus noise."
"Show recent API/delivery issues for the ad account "Name" — AI chat may include Meta Opportunity Score to check your overall account health
Campaign/ad set/ad drill-downs:
"Compare Spend and Reach for Threads vs. Instagram vs. Facebook Feed for campaign "Name"
Ads Manager history:
"Access Meta Ads Manager history and show what recent changes were made to my campaigns?"
Meta Ads Library analysis:
“Find ads in the ecommerce space that use urgency or scarcity messaging — things like "limited time," "only X left," or countdown angles. US market, last 90 days. I'm looking for examples across different product categories to compare how different brands execute this tactic.”
”Pull a sample of ads from direct-to-consumer supplement brands targeting the US. I want to understand what claims, creative formats, and CTAs are most common right now. Give me a spread — include both small brands and larger players if possible.”
”Show me all active ads from [Competitor Name] running in the US right now. Focus on ads that have been active for more than 30 days — I want to see what's been working for them, not just what's new. Prioritize video and carousel formats.”
Industry/Auction benchmark comparisons:
"How is my account performing compared to other advertisers in my auction? I want to know where I'm above benchmark and where I'm losing ground — CPM, CTR, whatever is available."
"My campaign "Name" feels expensive lately. Compare its costs and delivery against auction benchmarks so I can tell if it's a me problem or a market problem."
"Compare ad account "Name" performance against industry benchmarks"
Anomalies detection:
"Has anything unusual happened in my campaigns in the last 7 days? Flag anything that looks off — sudden drops in delivery, weird CPM spikes, CTR anomalies. I want to catch problems before they get expensive."
"Run anomaly detection on my active campaigns and tell me if Meta flagged anything irregular on the delivery or auction side."
Examining the list of images/videos/custom audiences:
"List ad images/videos/custom audiences the ad account owns" — Meta MCP will provide image information, including hash: this info can be used when creating new ad creatives. By providing the specific ID/name of the entity, you can access additional information, such as video duration, description, file name, and audience size.
Meta Business Help Center search:
"How the learning phase behaves when you duplicate an ad set versus editing it in place — specifically, whether duplicating resets the learning phase even if the budget and audience stay the same?"
Note: the variety of use cases covered by Meta MCP may change as Meta rolls out new updates.