What is MCP for Meta Ads?
MCP is the Model Context Protocol — an open standard that lets an AI client call external tools and read external data in a structured way. On its own an assistant like Claude or ChatGPT knows nothing about your Meta account. MCP is the mechanism that gives it access.
GoMarble MCP is an MCP server that connects supported AI clients to marketing data and workflows in GoMarble. For Meta Ads that means the assistant can reach your ad accounts, campaigns, ad sets, ads, creatives, audiences and pixels — and the performance data attached to each — rather than reasoning from whatever you paste into the chat.
The difference that matters for Meta specifically is structure. Meta performance questions are almost never answerable at one level: a CPA move shows up at account level, is caused at ad set or creative level, and is often explained by audience overlap or creative fatigue you cannot see from a summary. GoMarble MCP exposes those levels together, so the assistant can drill rather than guess.
How to connect Meta Ads to GoMarble MCP
Connecting Meta Ads to GoMarble MCP takes a few steps. Once connected, you work with your Meta data from the AI tool you already use.
- Connect your Meta Ads account to GoMarble.
- Connect GoMarble MCP to your preferred supported AI client.
- Authorize the access GoMarble requests.
- Choose the GoMarble workspace and accounts the client can use.
- Start asking questions or running supported workflows.
Connecting to Claude
The current steps, from the GoMarble MCP page:
- Open Settings, then Connectors.
- Choose Add custom connector.
- Enter https://apps.gomarble.ai/mcp-api/sse, then select Authorize.
- Choose the GoMarble workspace and accounts Claude can use.
Authorization uses GoMarble’s consent and workspace permissions, completed in the browser window your client opens.
Other supported clients
GoMarble MCP is a remote MCP server at https://apps.gomarble.ai/mcp-api/sse, currently supported in Claude, ChatGPT, Cursor, n8n, Hermes and OpenClaw, plus other clients that support compatible remote or local MCP servers. Setup differs slightly per client, so follow the canonical instructions rather than adapting the Claude steps.
View the complete GoMarble MCP setup guide →
What the free plan includes
- Up to $50,000 in cumulative ad spend
- One account per data source
- No edit actions — read and analysis only
What can you do with GoMarble MCP for Meta Ads?
Grouped by the job rather than by the tool that does it.
Analyze
- Analyse account, campaign, ad set and ad performance
- Compare time periods and spot trends
- Compare campaigns and ad sets against each other
- Break down spend, ROAS, CPA, CPM, CTR and frequency
- Pull account-level insights synchronously or as an async job for large ranges
Diagnose
- Investigate why ROAS or CPA moved
- Localise a change to the campaigns, ad sets or ads responsible
- Surface spend with weak conversion contribution
- Check audience overlap between ad sets competing for the same people
- Read the account change history to see what was edited before the shift
Creative intelligence
- Analyse creative performance at ad level
- Inspect creative details — copy, headlines, formats, assets
- Analyse a specific creative by ID or by URL
- Identify fatigue signals from CTR decay against rising frequency
- Find patterns shared by the ads that are working
Research
- Search brands in the ads library and find competitors
- Pull a competitor's running ads and brand-level analytics
- Discover ads by theme to see what a category is investing in
- Analyse an individual competitor ad
Recommend
- Surface campaigns and ad sets worth investigating
- Recommend where budget is constrained versus saturated
- Suggest creative directions grounded in what already performed
- Flag optimisation opportunities with the supporting numbers
Act
- Propose campaign, ad set, ad and creative updates
- Propose cloning a campaign, ad set, ad or creative
- Propose a new campaign structure
- Every write is a proposal that routes through your workspace's approval gates
GoMarble MCP capabilities for Meta Ads
| Capability | What GoMarble MCP can do |
|---|---|
| Performance analysis | Analyse Meta performance across spend, ROAS, CPA, CPM, CTR and frequency, at account, campaign, ad set or ad level |
| Campaign analysis | Compare campaigns and ad sets and identify what separates them |
| Creative analysis | Read creative details and performance, and analyse a creative by ID or URL |
| Performance diagnosis | Investigate the cause of a change and localise it to the responsible objects |
| Audience analysis | Read custom audiences and targeting, and check overlap between ad sets |
| Competitor research | Search the ads library, find competitors and pull their running ads |
| Change history | Read what was edited in the account and when |
| Reporting | Generate account and campaign summaries on request |
| Actions | Propose updates, clones and new structures, subject to approval gates |
What data can GoMarble MCP access?
The data available for Meta Ads, grouped by type.
Account data
Ad accounts · Account details and settings · Account-level insights · Historical performance · Account activity and change log
Campaign structure
Campaigns · Ad sets · Ads · Budgets · Objectives and bid strategy · Delivery status
Performance metrics
Spend · Revenue and ROAS · CPA and conversions · CPM, CPC and CTR · Frequency and reach · Placement breakdowns
Creative data
Ad creatives · Copy and headlines · Images and videos · Formats · Creative-level performance
Audience data
Custom audiences · Targeting definitions · Audience-level performance · Targeting search
Commerce and tracking
Pixels · Product catalogs · Product sets
Page data
Pages · Page insights · Posts · Post insights · Post comments
Competitor data
Brand search · Competitor discovery · Competitor ads · Brand analytics
GoMarble MCP workflows for Meta Ads
Jobs marketers actually want done, rather than features. Each one is a single natural-language request.
Find out why ROAS dropped
“Why did my Meta ROAS drop over the last 7 days?”
What GoMarble MCP does: Pulls account insights for the period and the one before it, decomposes the change across spend, CPM, CTR and conversion rate, then breaks down by campaign and ad set to find where the movement concentrated.
What you get: The contributing metrics ranked by impact and the specific campaigns or ad sets responsible, rather than a restatement that ROAS fell.
Find wasted spend
“Which ad sets are spending without generating conversions?”
What GoMarble MCP does: Reads ad set level spend and conversion data across the account and isolates line items with meaningful spend and weak conversion contribution.
What you get: A list sized in money rather than count, with the ad sets and their parent campaigns named.
Check for creative fatigue
“Which of my ads look like they are fatiguing?”
What GoMarble MCP does: Reads ad-level CTR, frequency, spend and creative age over time and looks for CTR decline against rising frequency at stable spend.
What you get: The ads showing fatigue signals, with the metrics behind the call so you can judge whether it is fatigue or audience saturation.
Diagnose a CPA increase
“What caused my CPA to increase this week?”
What GoMarble MCP does: Compares the period against the prior one, separates the cost side from the conversion side, then drills to campaign, ad set, placement and creative level.
What you get: The mechanism behind the increase — auction cost, click-through, conversion rate or mix — and where it originated.
Compare your top campaigns
“Compare my top five campaigns this month and explain what separates the winners.”
What GoMarble MCP does: Pulls campaign-level performance, normalises for spend, and inspects the ad sets and creatives inside each.
What you get: A like-for-like comparison and the structural or creative differences that explain the gap.
Check audience overlap
“Are any of my ad sets competing for the same audience?”
What GoMarble MCP does: Reads targeting definitions and custom audiences across ad sets and compares them for overlap, alongside CPM trends.
What you get: The ad sets bidding against each other, which is waste that never appears as a bad line item.
Research what competitors are running
“What ads is [competitor] running right now, and what patterns do they share?”
What GoMarble MCP does: Searches the ads library for the brand, pulls their live ads and brand analytics, and analyses the set for shared hooks, formats and angles.
What you get: Their running creative with duration as the signal, plus the pattern across it rather than a single ad.
Generate a performance summary
“Give me a summary of my Meta account this week.”
What GoMarble MCP does: Pulls account and campaign performance for the period, compares against the prior one, and surfaces the notable movements.
What you get: A summary with the changes that matter and their causes attached.
Example GoMarble MCP prompts for Meta Ads
Things real marketers ask, grouped by intent. Paste any of them once connected.
Performance analysis
- “How did my Meta account perform last month compared to the month before?”
- “Which campaigns drove the most conversions in the last 14 days?”
- “Show me ROAS by campaign for this month.”
- “What is my blended CPA across the account right now?”
Diagnosis
- “Why did my ROAS drop last week?”
- “What caused CPA to increase in my prospecting campaigns?”
- “Which ad sets are responsible for the spend increase?”
- “Did anything change in the account before performance shifted?”
Creative analysis
- “Which creatives are driving the strongest conversion rates?”
- “Which ads are showing signs of creative fatigue?”
- “What do my best-performing ads have in common?”
- “Analyse this ad and tell me what it is doing well.”
Budget and spend
- “Where does my budget allocation look inefficient?”
- “Which campaigns are budget-constrained?”
- “Which ad sets are spending without converting?”
Audience analysis
- “Which audiences are performing best?”
- “Are any of my ad sets overlapping on targeting?”
- “How is frequency trending on my top campaigns?”
Competitor research
- “Who are my main competitors advertising on Meta?”
- “What ads has [competitor] been running longest?”
- “What creative angles is my category using right now?”
Reporting
- “Summarise my Meta performance for this week.”
- “Build me a campaign-level breakdown I can send to a client.”
Actions
- “Propose pausing the ad sets with no conversions in the last 14 days.”
- “Propose shifting budget from my saturated campaigns to the constrained ones.”
Why use GoMarble MCP for Meta Ads?
An API gives AI access to data. GoMarble MCP gives it the marketing context and workflows needed to work with that data.
Structure, not just rows
The Meta API returns whatever object you ask for. GoMarble MCP exposes account, campaign, ad set, ad and creative together, which is what lets an assistant drill from a symptom to a cause instead of reporting a number.
Creative is joined to performance
Creative details and creative-level results arrive together, so a question about hooks or formats can actually be answered. A raw connection gives you asset IDs.
Competitor context is included
The ads library is part of the same surface, so outside research and your own performance can be reasoned about in one conversation.
Writes are governed
Every change is a proposal that routes through your workspace's approval gates, rather than a direct API call with whatever permissions the token carries.
Cross-platform by default
Meta rarely explains itself alone. The same connection reaches Google, TikTok, GA4, Shopify and Klaviyo, so a Meta question can be answered with the rest of the funnel in view.
GoMarble MCP for cross-platform marketing analysis
Marketers rarely decide anything from one platform. GoMarble MCP can give AI a view of the wider marketing system rather than forcing it to reason from one disconnected source at a time.
| Combination | What becomes answerable |
|---|---|
| Meta Ads + Google Ads | Compare channels on a consistent basis, see where budget should move between them, and catch the case where cutting Meta prospecting raises Google brand-search CPA a fortnight later. |
| Meta Ads + GA4 | Join ad-side delivery to on-site behaviour and conversions, which is how you tell a creative problem from a landing-page problem. |
| Meta Ads + Shopify | Put Meta spend against store revenue, orders and product-level performance rather than platform-reported conversions alone. |
| Meta Ads + Klaviyo | See how paid acquisition feeds email and retention, and whether the customers a campaign brings are worth having. |
| Meta + TikTok + LinkedIn + Bing | Compare paid social and search side by side when budget decisions span more than one platform. |
How does GoMarble MCP work?
| Step | What happens |
|---|---|
| 1 | You ask a question in natural language |
| 2 | The AI client works out what information it needs |
| 3 | GoMarble MCP retrieves the relevant marketing data and tools |
| 4 | GoMarble supplies the marketing context around it |
| 5 | The AI analyses the result |
| 6 | You get an answer, a recommendation, a report — or a proposed action |
Write access is plan-dependent. On the read-only plan you can connect a client and analyse supported accounts without making changes. Where write access is enabled, changes are proposed rather than applied directly — every write tool in the server is a propose call, and consequential actions route through the permissions and approval gates configured for your workspace.
Who is GoMarble MCP for?
Meta media buyers
Investigate an account without assembling six breakdown views by hand, and get to the responsible ad set faster.
Creative strategists
Ask which hooks and formats are working, with the performance data attached rather than inferred.
Performance leads
Diagnose a ROAS or CPA move in one conversation instead of a morning of exports.
Agencies running Meta accounts
Standardise the same investigation across client accounts and run it from the client's own context.
Ecommerce founders
Understand what Meta is doing to revenue without learning Ads Manager's reporting model.
GoMarble MCP vs traditional Meta Ads workflows
| Traditional workflow | With GoMarble MCP |
|---|---|
| Open Ads Manager and build a breakdown | Ask the question from your AI client |
| Export to a spreadsheet to compare periods | Compare periods in the conversation |
| Check campaign, then ad set, then ad, then creative | Ask for the cause and let it drill the levels |
| Open the ads library in another tab for competitor work | Research competitors in the same thread |
| Rebuild the same weekly analysis by hand | Re-run the same prompt or workflow |
| Apply changes manually, hoping the diagnosis was right | Propose the change with the analysis attached, then approve it |
Frequently asked questions
What is MCP for Meta Ads?
What is GoMarble MCP?
Can GoMarble MCP analyse Meta Ads campaigns?
Can GoMarble MCP analyse Meta creatives?
Can GoMarble MCP explain why Meta performance changed?
Can GoMarble MCP identify wasted Meta spend?
Can GoMarble MCP make changes to my Meta account?
Is GoMarble MCP read-only?
What are the free plan limits?
Which AI assistants work with GoMarble MCP?
How do I connect Meta Ads to GoMarble MCP?
Can GoMarble MCP work across multiple marketing platforms?
How is this different from using the Meta Marketing API directly?
Does it support competitor research on Meta?
Why GoMarble MCP?
The work you stop doing:
- Stop assembling six breakdown views to answer one question
- Stop exporting to a spreadsheet to compare two periods
- Stop re-explaining your account structure to an assistant every session
- Stop rebuilding the same weekly investigation by hand
- Stop switching tabs between Ads Manager, the ads library and your store data
Connect your marketing stack once. Give AI the context it needs. Ask in plain English, and let it analyse, explain, recommend and — where enabled — propose the change.
Your Meta Ads data is already there. Now give AI access to it.
Connect GoMarble MCP to your marketing stack and start analysing Meta Ads from the AI tool you already use.