How to connect Metricool to an AI agent: what GTM AI agents can do with Metricool in Day AI
Updated
The short answer. To connect Metricool to an AI agent, give the agent Metricool's MCP server with your own Metricool login and keep scheduling behind an approval, so GTM AI agents, which carry AI for GTM into marketing, can plan social posts from what customers say on calls and check how they perform. In Day AI, you connect your own Metricool account from Settings > Connectors, with nothing for an Owner to set up, and your agent can read social and ads analytics, check the best times to post, schedule posts and research competitors from chat.

Imagine you run marketing at a construction-software company, and it's Friday at 10 with next week's posts still unplanned. You ask your Day AI agent what buyers kept bringing up on calls this month. It reads the team's recorded meetings you have access to and finds a theme: three prospects asked how other firms handle change orders on site. The agent then reads Metricool, sees that your how-to posts outperformed product announcements last month, and checks the best times to post for your LinkedIn audience. It drafts two short posts on change orders, written in plain language with no customer named, and proposes scheduling them for those times. You ask it to trim the first post, then approve both in the chat, and your week's posts now answer questions buyers are actually asking.
Where the usual answer goes wrong. Social plans usually start from a content calendar and last month's analytics. Analytics tell you which format worked; they can't tell you what your buyers are worried about this week. Day AI starts from the conversations your team has every day and uses Metricool for the timing, the scheduling and the read on what's working. If you already plan posts from a calendar and a performance report, keep both; the steps below add the one input they lack.
Listen, plan, publish: Metricool with Day AI
Posts that bring in pipeline come from three inputs in this order: what buyers say, what has worked, and when your audience is there.
- Listen: what buyers are asking. Sales and success calls are the richest source of topics a marketing team has. In Day AI, the agent reads the meetings and email you have access to and pulls out the questions buyers keep asking.
- Plan: what has worked and when. A topic needs a format and a time slot. In Day AI, the agent reads Metricool's social and ads analytics and best times to post, and researches competitors' posting when you ask.
- Publish: with a person's approval. Nothing goes on your company's feed without a person reading it first. In Day AI, scheduling a post in Metricool waits as an approval card in the chat until you approve it.
How to connect Metricool to Day AI
Metricool is in beta in Day AI. Each person connects their own account, and there's no workspace setup.
Step 1: Connect your Metricool account in Day AI
In Settings > Connectors, find Metricool, click Connect, and sign in with your Metricool account; your agent then uses that account's access.
Step 2: Set Metricool's tool permissions in Day AI
In the Metricool connector's Tool permissions, pick Always allow, Needs approval or Off per tool. Analytics and research tools start at Always allow and scheduling starts at Needs approval; keep it there so every post gets a human read.
Step 3: Ask your Day AI agent for post ideas from customer conversations
Ask "What did prospects ask about most on calls this month?" and then "Draft two LinkedIn posts on that topic for next week." Your agent reads the conversations in Day AI, checks Metricool for what's performed and when to post, and drafts the posts. Ask it to keep customer names and details out, and edit the drafts in plain words.
Step 4: Approve Metricool scheduling in the Day AI chat
When the agent proposes a scheduled post, an approval card appears in the chat naming Metricool and the action, with Approve and Deny. The post is scheduled only after you approve. If you'd rather keep a chat about strategy only, turn Metricool off for that chat from the + menu in the composer.
Step 5: Give a Day AI skill the weekly analytics read
Ask your agent to create a skill, such as a Monday report on last week's social and ads performance next to the themes from customer calls, and select Metricool in the skill's Connectors picker. The skill reads Metricool through its owner's connection, and only because it's selected there; it runs without approvals, so scheduling stays in chat. You can line Metricool up with every Day AI integration and connector your team uses.
Where a calendar-first social plan breaks
- Topics come from the marketing team's guesses. A calendar planned in a meeting room drifts from what buyers ask. In Day AI, the agent reads the team's calls and email for the questions buyers raise.
- Analytics are read once a month. By the time someone opens the report, the week's posts are already out. In Day AI, the agent reads Metricool's current analytics whenever you ask it to plan a post.
- Scheduling happens without a second look. A post written in a hurry goes live on the company page. In Day AI, every scheduled post waits for a person's approval in the chat.
- Competitor research lives in another tab. Someone checks what competitors posted, and it never reaches the plan. In Day AI, the agent researches competitors in Metricool in the same chat where it drafts your posts.