Menu

Mattering Automations Plausible Analytics integration

Connect Plausible Analytics to Mattering Automations: the unified v2 stats query, realtime visitors, and site, goal, and shared link provisioning as workflow steps, authenticated with a Plausible API key.

The Mattering Automations Plausible Analytics integration adds Plausible's privacy-first, cookieless analytics to your automation canvas: query any historical stat through the unified v2 endpoint with metrics, date ranges, dimensions, and filters, read the realtime visitor count, and on Enterprise plans provision sites, goals, and shared dashboard links.

Setup takes a few minutes: create an API key in Plausible, attach it as a credential in Mattering, and test the connection. This guide walks through each step and ends with the full node reference.

The central fact of this integration is that Plausible issues API keys in two grades, chosen when the key is created. A Stats API key, a Business plan feature, serves Plausible: Query Stats and Plausible: Get Realtime Visitors. The six sites-surface nodes require a Sites API key, an Enterprise plan feature, and each says so in its description, so a 401 from them under a Stats key is expected behavior, not a broken setup.

Before you start

  • A Plausible account on plausible.io with at least one site set up. Every stats call names a site by its domain, so know the domain exactly as Plausible shows it.
  • An API key of the grade your workflows need: a Stats API key (a Business plan feature) for the stats nodes, or a Sites API key (an Enterprise plan feature) for the site, goal, and shared link nodes.
  • A Mattering workspace with access to the Automations app.

Create a Plausible API key

Mattering authenticates to Plausible with an API key, sent as a bearer credential on every request. Plausible offers two key grades and the grade is fixed at creation: Stats API keys read analytics, Sites API keys manage sites, goals, and shared links. Create the grade that matches the nodes you plan to use.

In your plausible.io account, open the account menu and go to Settings > API Keys.
Click New API Key and choose the key type: a Stats API key for reading analytics, or a Sites API key for provisioning.
Copy the key.
Plausible shows the key once, at creation. The grade is also fixed at creation: the two are distinct key types, not tiers of one key, so a workflow that needs the sites surface later means creating a separate Sites API key, not upgrading the existing Stats key.

Connect Plausible Analytics to Mattering

In Mattering Automations, open Connectors and click Add Connector.
Choose Plausible Analytics from the list.
Click Create new credential, paste your API key, and save.
The credential is stored encrypted in Google Secret Manager and never appears in workflow logs.
Click Test Connection and provide the domain of a site your key can access. Plausible has no authenticated read that takes zero parameters on a Stats API key, so the test reads that site's realtime visitor count, a free, read-only call.
Click Connect. All 8 Plausible Analytics nodes appear in your canvas palette.

Verify it works

After connecting, open any automation canvas and check the palette: a Plausible Analytics section should list the query and action nodes. Drop Plausible: Get Realtime Visitors onto the canvas, set site_id to your site's domain, and run a test step. The response body is a bare number, so the live count lands directly at data rather than under a wrapper key; a 200 with a number, even 0, means the key and the site reference are wired correctly. The call is free and read-only, which makes it the credential health check for this integration.

Then run Plausible: Query Stats with the visitors metric and a 7d date range to confirm the main surface: rows arrive at data.results, each with a metrics array and a dimensions array. On a Sites API key, Plausible: List Sites is the natural check instead, since it takes no parameters and lists the sites the key can reach. A 401 from the sites nodes under a Stats API key is the key-grade boundary described above, not a setup error.

Node reference

8 nodes, generated from the connector itself so this reference always matches what ships.

Read and query nodes 5

Plausible: Get Realtime Visitors

plausible_realtime_visitors

Get the number of visitors currently on your site (anyone who triggered a pageview in the last 5 minutes). Needs a Stats API key. The response body is a bare number, so the count lands directly at data (e.g. data equals 21), not under a wrapper key. This is the one legacy v1 stats endpoint still shipped: the v2 query API has no realtime range, so this remains the documented way to read the live count. Free to call and read-only, which makes it the connector's credential health check.

Plausible: Get Site

plausible_get_site

Fetch one site's details by domain: timezone, allowed custom properties, and the tracker script configuration (which optional tracking features are on). Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Plausible: List Goals

plausible_list_goals

List a site's configured goals (custom events and page visits) at data.goals, each with its id, goal_type, event_name or page_path, and display_name. Goal ids and names feed the event:goal dimension of the Query Stats node. Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Plausible: List Sites

plausible_list_sites

List the sites your API key can access, with each site's domain and timezone at data.sites. Cursor-paginated. Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Plausible: Query Stats

plausible_query_stats

Query any historical stat from your Plausible dashboard through the unified v2 query endpoint: pick metrics, a date range, and optional dimensions and filters, and get rows at data.results (one metrics array plus one dimensions array per row). This single node replaces the legacy v1 aggregate, timeseries, and breakdown endpoints: aggregate is a query with no dimensions, a timeseries is a time dimension, and a breakdown is any other dimension. Needs a Stats API key (Business plan feature; default rate limit 600 requests per hour).

Action nodes 3

Plausible: Create Goal

plausible_create_goal

Create a goal on a site, or find the existing one: the endpoint is an idempotent create-or-find PUT, so re-running it never duplicates a goal. Two kinds: goal_type event (matches a custom event by name) or goal_type page (matches visits to a page path, wildcards supported). Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Plausible: Create Shared Link

plausible_create_shared_link

Create a shared dashboard link for a site, or find the existing one by name (idempotent create-or-find PUT). Returns the secret URL at data.url, ready to send to a client or embed in a report - anyone with the link can view the dashboard without logging in. Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Plausible: Create Site

plausible_create_site

Add a new site to Plausible programmatically, e.g. when provisioning analytics for each new customer site. Returns the created site with its tracker script configuration. Reversible: the site can be deleted from the Plausible dashboard (this connector ships no delete node on purpose). Requires a SITES API key (an Enterprise plan feature on plausible.io, created as its own key type under Settings, API Keys); a Stats API key gets a 401 here.

Using the integration

One query node for the whole dashboard

Plausible: Query Stats replaces what used to be three separate endpoints. Pick metrics (visitors, visits, pageviews, bounce rate, visit duration, and more), a date range (relative keywords like 7d and month, or an explicit date pair), and optional dimensions and filters. No dimensions returns an aggregate, a time dimension returns a timeseries, and any other dimension returns a breakdown by page, source, country, or whatever you chose. Filters compose with and, or, and not, so a question like blog traffic from search, excluding one country, is a single call.

The realtime count and its bare number

Plausible: Get Realtime Visitors returns the number of visitors on the site right now, meaning anyone who triggered a pageview in the last five minutes. Plausible's response body is a bare JSON number, so the count sits directly at data with no wrapper key; map data itself into the next step. This is the one legacy v1 read the integration keeps, because the v2 query has no realtime range, and it doubles as the cheapest possible credential check.

The two key grades in practice

One credential slot, two possible grades. Most users store a Stats API key and use the two stats nodes; the six sites-surface nodes then return Plausible's 401, and each node's description names the key it needs, so the error explains itself. Enterprise users who want provisioning store a Sites API key instead. The Plausible docs present the grades as separate products, so the integration makes no claim that one grade serves the other surface.

On the automation canvas

Plausible nodes chain with every other node: a Monday schedule can run Plausible: Query Stats for the week's numbers and land the digest as a report task in Mattering Productivity, a spike check can compare today against the trailing numbers kept in a Mattering database and open an investigation task when traffic moves, and goal ids and names from Plausible: List Goals can feed the event:goal dimension for conversion reporting.

From AI assistants over MCP

Every node speaks MCP, the open protocol for connecting AI assistants to tools, so Claude, ChatGPT, and Matter can answer traffic questions through the same nodes your workflows use. Because one query covers aggregates, timeseries, and breakdowns, an assistant can answer how the blog did this month and where that traffic came from in a call or two. The stats surface is entirely read-only, which makes it a safe surface to hand an agent.

Provisioning on the sites surface

With an Enterprise Sites API key, onboarding a customer site becomes a workflow: Plausible: Create Site adds the site (removable from the Plausible dashboard; the integration ships no delete nodes), Plausible: Create Goal configures event or page goals, and Plausible: Create Shared Link returns the secret dashboard URL to send the client. The goal and shared link creates are idempotent create-or-find operations, so re-running the workflow never duplicates them.

Troubleshooting

Test Connection fails with a 401
The key is invalid, was revoked, or picked up stray whitespace when pasted. Plausible shows a key once, at creation, so a lost key means creating a new one under Settings > API Keys and updating the credential on the connector.
The stats nodes work but the sites nodes return 401s
That is the key-grade boundary, not a fault. The stored key is a Stats API key, and Plausible: List Sites, Plausible: Get Site, Plausible: List Goals, Plausible: Create Site, Plausible: Create Goal, and Plausible: Create Shared Link all require a Sites API key, an Enterprise plan feature created as its own key type under Settings > API Keys. Create a Sites API key and store it as the credential for workflows that provision.
Plausible: Get Realtime Visitors returns just a number
That is the endpoint working. Plausible's realtime response body is a bare JSON number, so the count lands directly at data with no wrapper key, and a response of 0 is a valid answer meaning nobody triggered a pageview in the last five minutes. Map data itself into the next step of the workflow.
Plausible: Query Stats returns no rows
Check three things in order: the site_id is the site's domain exactly as it appears in Plausible, the date range actually covers days with traffic, and the filters are not excluding everything. An empty data.results with a 200 is Plausible answering honestly that nothing matched, which is different from an error status, which points at the key or the site reference instead.
Requests start failing after many calls in a short window
Plausible API keys are rate limited to 600 requests per hour by default. Scheduled sweeps rarely approach that, but a backfill loop iterating over many date ranges or sites can. Space the runs out, fold several questions into fewer, richer Plausible: Query Stats calls (one query with a dimension replaces many single-purpose calls), and retry after the window resets.
There is no node to send events or pageviews
Deliberate. Plausible's event ingestion endpoint identifies visitors by client IP address and user agent, so events posted from an automation runner would all carry one server's IP and one user agent, and the site's traffic would be attributed to a single phantom visitor. Install Plausible's script on the site instead; this integration reads analytics and provisions sites, it does not write traffic.
Plausible: Create Goal or Plausible: Create Shared Link keeps returning the same record
That is idempotency, not a failure. Both are create-or-find operations: if a goal with the same definition, or a shared link with the same name, already exists, Plausible returns the existing one rather than creating a duplicate. That is what makes the nodes safe to re-run inside workflows; to get a genuinely new shared link, give it a different name.

Plausible Analytics is a trademark of its respective owner. Mattering uses platform names to identify the services these integrations connect to; no endorsement or affiliation is implied.

Start Free Explore our Products