Your social data, in every AI tool you use
The Measure MCP server connects your workspace to Claude, ChatGPT, Perplexity — any client that speaks the Model Context Protocol. Ask real questions where you already work, and get answers grounded in your live numbers.
40+ tools, included on every plan — trial included.
One server. Every AI tool your team uses.
MCP — the Model Context Protocol — is the open standard AI tools use to reach live data. Connect the Measure MCP server once, and the same questions work everywhere: in Claude, in ChatGPT, in Perplexity, in whatever your team adopts next.
What it can do
Every tool an analyst reaches for
More than 40 tools — reads across posts, accounts, audiences, dashboards, and targets, plus light writes like tagging and slide building. Nothing destructive is ever exposed.
Search every post
Query posts across every connected account — by content, format, platform, tag, or time window.
search_postsget_post_details
Pull the numbers
Account, group, and platform stats — reach, engagement, growth — over any window you name.
get_account_statsget_platform_stats
Know the audience
Demographics and geography for accounts and individual posts, straight from the source.
get_account_demographicsget_post_audience
Keep posts organized
Apply and remove tags in bulk, using the same taxonomy your team already runs in Measure.
add_tags_to_postslist_user_tags
Build the deck
Create reports, add slides, and update them from a prompt — no deck tool required.
create_reportadd_slide_to_report
Compare and track
Side-by-side account and post comparisons, plus the targets you're pacing against.
visual_compare_accountsget_target_progress
Why it's different
Where the Measure MCP goes further
New AI summaries & content detections
It reads the content, not just the numbers
Your AI tools can now pull Measure's AI read of the content itself: a written summary aggregated across every frame of multi-media posts — carousels included — alongside transcripts and computer-vision data. More than 20 detections cover subject matter, content quality, brand partnerships, ad disclosure methods, and more — down to a post's aspect ratio.
get_ai_summary_values get_post_content
Detections include
Slides, not just answers
17 visual tools build real report slides — side-by-side comparisons, metric charts, demographics, post grids. “Make me a deck” is a tool call, not a copy-paste job.
visual_compare_postsvisual_search_posts
More than 40 tools deep
Reads across posts, accounts, groups, dashboards, and targets. Light writes like tags and post groups. The same coverage your dashboards have — through one connection.
get_dashboard_widget_datacreate_post_group
Ships its own manual
The server publishes its own reference docs as MCP resources — so clients form valid calls on the first try instead of guessing at your data model.
data modelquery referencevisual tools guide
Ask it like you'd ask an analyst
- Recap last week's launch — which formats carried reach?
- Build a deck comparing our top 10 Reels this quarter.
- Which spring carousels disclosed the brand partnership — and did disclosure cost us reach?
- Tag everything from the spring campaign, then tell me how it performed.
- Does 9:16 beat 1:1 for us on Instagram this quarter?
- How are we pacing against this month's engagement target?
Setup
Connected in minutes
-
Turn it on
MCP access is off by default. A workspace admin flips it on once in settings — that's the whole rollout.
-
Add Measure to your client
Point Claude, ChatGPT, or Perplexity at your workspace's MCP server. The server URL lives in your Measure settings.
-
Sign in
A standard OAuth flow scopes the connection to your user and your workspace — the same permissions you have in the app. No API keys to paste.
-
Ask your first question
Start with last week's recap. Every answer is pulled live from your workspace the moment you ask.
Included on every plan — and on the 14-day trial.
Client-by-client walkthroughs, prompts, and FAQs: Measure in Claude · Measure in ChatGPT
Locked down by default
Measure Studio is the data source, not the model. Nothing leaves our infrastructure unless you connect a client and it makes a request — and you choose the client and model tier that handle what comes back.
Read the full security details- Opt-in per company. Disabled by default — an admin enables it before anyone can connect.
- OAuth 2.1 with PKCE. Every token is bound server-side to a single user and a single workspace.
- Scoped to you. Clients see only what you'd see logged in — same roles, same workspace isolation.
- Reads and light writes only. Tagging and slide building, yes. Destructive account-level actions, never.
- Stateless by design. No conversation transcripts stored — only standard access-log metadata.
- Revoke anytime. Disconnect a client to immediately invalidate every token it was issued.
Questions, answered
The short version of how it works. For the deep dive, see the security and data pages, or book a demo.
What is MCP?
MCP — the Model Context Protocol — is an open standard that lets AI tools securely call live data sources. Measure Studio ships an MCP server, so any compliant client can query your social data with your permission.
Which AI tools work with it?
Claude, ChatGPT, and Perplexity are the ones teams reach for most today, but the server speaks standard MCP — any compliant client can connect, including ones that haven't shipped yet.
Is it included in my plan?
Yes. The MCP server is included on every plan — Starter, Pro, and Enterprise — and it works during the 14-day trial, so you can query your data from Claude on day one.
What can a connected client actually do?
More than 40 tools' worth: every read your dashboards can do — posts, stats, audiences, dashboards, targets — plus light writes like tagging posts, creating post groups, and building report slides. Nothing destructive is exposed.
Can it see inside the content itself?
Yes. Measure's AI layer is part of the surface: written summaries aggregated across every frame of multi-media posts like carousels, transcripts and computer-vision data, and more than 20 detections covering subject matter, content quality, brand partnerships, ad disclosure methods, and more — down to a post's aspect ratio.
Is my data used to train models?
Not by us — Measure Studio is the data source, not the model. You pick the client and the model tier that handle each answer, so the connection can meet your own data-residency and training requirements. We store no conversation transcripts.
How is this different from Archer?
Archer is the AI analyst inside Measure Studio — it reads, asks, and recaps in-app. The MCP server brings the same live data into the tools you already work in. Same numbers, different desks — and both are included.
Take your numbers with you.
14-day trial · No credit card · Connect Claude in minutes.