LinkedIn Ads MCP Server
Social
Connect any AI assistant to LinkedIn Ads through one secure MCP connection to analyse, build and optimise B2B campaigns, with human approval on every spend.
What is the LinkedIn Ads MCP?
The LinkedIn Ads MCP Server is a secure connection that lets AI assistants — ChatGPT, Claude, Claude Code, Cursor or Codex — work directly inside your LinkedIn Ads account through OpenAds. Instead of exporting spreadsheets or clicking through Campaign Manager, you ask questions and describe outcomes in plain language. The AI researches audiences, analyses campaign performance, drafts new campaigns, models budget changes and writes reports — all grounded in your live account data through a single Model Context Protocol connection.
What makes it safe for a channel as expensive as LinkedIn is OpenAds’ separation of duties. Analysis, proposal, approval and execution are kept distinct. The AI can read everything and propose anything, but any action that spends money or changes your account — launching a campaign, shifting budget, editing bids — is held as a clear proposal until a human approves it. You keep the speed of an AI analyst with the control of a manual reviewer, which matters when B2B clicks routinely cost many times more than other social channels.
Key capabilities
Pull live campaign, campaign group and account metrics on demand.
Analyse cost per lead, cost per result and pipeline efficiency by segment.
Research job titles, functions, seniorities, industries and Matched Audiences.
Identify wasted spend across audiences, placements and creatives.
Draft Sponsored Content, Message Ads, Dynamic Ads and Lead Gen Form campaigns.
Model and pace budgets against monthly targets before committing.
Generate stakeholder-ready performance reports in seconds.
Hold every spend or write action for explicit human approval.
What you can do
Analyse performance
Ask for a read on any campaign, campaign group or the whole account across a date range. OpenAds retrieves impressions, clicks, CTR, spend, leads and cost per lead, then breaks results down by audience, job function, seniority, company size or ad format so you can see exactly where B2B budget is converting and where it is stalling.
Research keywords & audiences
LinkedIn targeting is built on professional attributes rather than search keywords. OpenAds helps you explore job titles, functions, seniorities, member skills, industries and company lists, estimate audience reach, and compare narrow account-based segments against broader awareness audiences before you spend anything.
Find wasted spend
Ask where budget is leaking. OpenAds surfaces audiences with high spend and no leads, overlapping segments competing against each other, underperforming creatives, and campaigns pacing far above their cost-per-lead target — then quantifies how much you could reclaim by pausing or reallocating.
Create campaigns
Describe an objective — for example, lead generation for a finance product aimed at CFOs in India — and OpenAds drafts a complete structure: objective, audience definition, placements, ad format, Lead Gen Form fields, budget and schedule. Everything is presented as a proposal for you to review and adjust before it is created.
Optimise & pace budget
OpenAds tracks spend against your monthly cap and flags campaigns pacing too fast or too slow. It proposes bid and daily-budget adjustments to hit your cost-per-lead goal, shows the projected impact of each change, and waits for approval before anything is applied.
Generate reports
Ask for a weekly or monthly summary and OpenAds compiles spend, leads, cost per lead and trend-over-time into a clear narrative, complete with the biggest movers and recommended next steps — ready to paste into a client update or board deck.
Supported campaign types
Campaign typePrimary useTypical workflow Sponsored Content (Single Image)Awareness and traffic in the feedDefine audience, draft creative, set budget, approve, launch Lead Gen FormsCapturing qualified B2B leads in-platformBuild form fields, target by seniority, approve, measure cost per lead Message AdsDirect outreach to inboxes for events or demosWrite message copy, pick audience, approve, track replies Document AdsDistributing gated whitepapers and guidesAttach document, target by function, approve, optimise CPL Dynamic AdsPersonalised follower and spotlight unitsSet template, define audience, approve, monitor engagement Video AdsThought leadership and product demandUpload creative, target by industry, approve, analyse view rate
Getting started
Sign in to OpenAds and open the integrations directory.
Select LinkedIn Ads and start the secure authorisation flow.
Grant access to the LinkedIn ad account you want to manage.
Confirm the account, currency and default spend approval settings.
Set your monthly budget cap and approval thresholds for write actions.
Once LinkedIn Ads is connected, link your AI assistant of choice — ChatGPT, Claude, Claude Code, Cursor or Codex — to OpenAds through the same MCP connection. From that point you can prompt in natural language from whichever tool you already use, and every proposal routes back through OpenAds' approval layer before anything executes.
First prompts to try
"Show me cost per lead by seniority for the last 30 days."
"Which audiences spent over ₹50,000 with no leads this month?"
"Draft a Lead Gen campaign targeting IT directors in India with a ₹2,00,000 budget."
"Where can I reclaim wasted spend across my active campaigns?"
"Compare CTR between my single-image and video ads."
"Propose budget shifts to keep me under ₹8,00,000 this month."
"Build a weekly performance report for my demand-gen campaign group."
"Estimate the audience size for CFOs at companies with 1,000+ employees."
A complete workflow example
Objective: reduce cost per lead on a demand-generation programme without losing volume.
User prompt: "My CPL has crept up this quarter. Find what's driving it and propose fixes to bring it back under ₹3,000 per lead."
What OpenAds analysed: ninety days of spend, leads and cost per lead across every campaign, broken down by audience, seniority and ad format, plus pacing against the ₹8,00,000 monthly cap.
Findings: two Sponsored Content audiences accounted for 40% of spend at a CPL above ₹6,500, a broad "all industries" segment overlapped heavily with a tighter account list, and one Lead Gen Form campaign delivered leads at ₹2,100 but was budget-constrained.
Proposed actions: pause the two high-CPL audiences, remove the overlapping broad segment, and reallocate roughly ₹1,50,000 into the efficient Lead Gen campaign — with a projected blended CPL of ₹2,750.
Approval step: each change was presented as a discrete, editable proposal. Nothing paused, nothing moved and no budget changed until the actions were explicitly approved. The user approved two of three and adjusted the reallocation amount before confirming.
Result: the approved changes executed in seconds, pacing returned within cap, and OpenAds scheduled a follow-up report to confirm the CPL trend the following week.
Best practices
Set a monthly budget cap so pacing alerts and proposals stay grounded in a real limit.
Start with read-only analysis prompts to build trust before approving write actions.
Keep audiences tightly defined and let OpenAds flag overlap between segments.
Review proposed changes on their projected cost-per-lead impact, not spend alone.
Schedule regular reports so trends are caught early rather than at month end.
Safety and control
OpenAds draws a firm line between read and write actions. Reading data — pulling metrics, researching audiences, building reports — happens freely because it changes nothing. Any write action that spends money or alters your account, such as launching a campaign, editing a bid, shifting budget or pausing an audience, is held at an approval checkpoint. You see exactly what will change and its projected impact, and you can edit, approve or reject each item. Nothing executes on LinkedIn Ads without your explicit sign-off.
OpenAds vs manual LinkedIn Ads management
AspectOpenAdsManual management Analysis speedLive answers in seconds from natural languageManual exports and pivot tables Wasted spend detectionProactively surfaced and quantifiedSpotted only on close review Campaign setupDrafted as an editable proposalBuilt field by field in Campaign Manager Control over spendExplicit approval on every write actionRelies on individual discipline ReportingGenerated on demand with next stepsCompiled manually each cycle
Limitations
Performance insight depends on conversion and Lead Gen data flowing back into LinkedIn; gaps in tracking limit analysis.
Creative assets such as images and video still need to be produced outside OpenAds before campaigns go live.
LinkedIn's minimum bids and audience-size floors apply, so some proposed budgets may need adjusting to be eligible.
Available fields and formats are bound by LinkedIn's Marketing API, which may lag the newest in-platform features.
Frequently asked questions
Can OpenAds spend money without my approval?
No. Every action that spends money or changes your account is held as a proposal until you approve it. Analysis and reporting are read-only and never touch your budget.
Which AI assistants can I use?
ChatGPT, Claude, Claude Code, Cursor and Codex all connect through the same OpenAds MCP connection, so you can work from whichever tool you already prefer.
Does it work with Lead Gen Form campaigns?
Yes. OpenAds can analyse Lead Gen performance, propose new form-based campaigns and optimise them against your cost-per-lead target, all within the approval workflow.
How does OpenAds keep my account secure?
Access is granted through LinkedIn's official authorisation flow, data moves over a single secure MCP connection, and write actions are gated behind approval checkpoints you control.
Can I still make changes manually?
Absolutely. OpenAds complements Campaign Manager rather than replacing it, so you can continue to edit anything directly in LinkedIn whenever you choose.
Connect it through one secure MCP connection.