Plausible Analytics

@plausible show me a 30-day breakdown of traffic sources for my-app.com, sorted by visitor count

Based on the last 30 days of data for my-app.com, here's your traffic source breakdown: Direct traffic accounts for 3,847 visitors (42%), followed by Google organic search with 2,156 visitors (24%), Twitter referrals at 1,423 visitors (16%), GitHub at 891 visitors (10%), and other sources totaling 683 visitors (8%). Your bounce rate from direct traffic is notably lower (31%) compared to social referrals (58%), suggesting direct visitors engage more deeply with your content.

This example demonstrates the Get Breakdown Stats tool, which segments your analytics by dimensions like source, page, country, or device. Requires read access to the Stats API. The breakdown respects your site's timezone setting and can filter by custom properties if you're on an Enterprise plan.

@plausible create a goal called 'signup_completed' for my-app.com and tell me how many times it fired in the last 7 days

I've created the 'signup_completed' goal for my-app.com. However, goal creation through the API requires the goal to be triggered by your site's tracking code before historical data becomes available. Since this is a new goal, there's no historical data for the past 7 days yet. Once your site starts sending 'signup_completed' events via the Plausible script, you'll be able to query conversion counts and analyze signup funnel performance.

This example shows a write operation followed by a read attempt, illustrating an important limitation: newly created goals won't have retroactive data. The MCP can create goals via the Sites API, but your frontend must instrument the tracking events. Goals are a paid feature, and you'll need both Sites API and Stats API access to create and query them.

@plausible compare this week's visitor trends to last week for my-app.com and explain any significant changes you notice

Pulling timeseries data for my-app.com: This week (Dec 9-15) averaged 487 daily visitors compared to 412 last week (Dec 2-8), a 18% increase. The spike occurred primarily on Tuesday and Wednesday, coinciding with your Product Hunt launch mentioned in referral data. Bounce rate improved from 52% to 47%, and average visit duration increased from 1m 24s to 1m 51s. However, Thursday-Friday showed a 23% drop-off, returning closer to your baseline. This suggests the launch drove quality traffic initially, but sustained engagement will require follow-up content or campaigns.

This example combines Get Timeseries Stats with the AI's analytical reasoning to surface insights across multiple metrics. The MCP provides raw timeseries data; the AI interprets patterns and correlates them with breakdown stats (like referral sources). Timeseries queries can span up to 12 months but become slower with finer granularity (hourly vs. daily intervals).

When Plausible wins for privacy-first campaign tracking

A 6-person SaaS marketing team runs weekly campaigns across Twitter, Product Hunt, and email. They need to see which sources drive signups without violating GDPR or slowing down their site with Google Analytics. Plausible's breakdown stats tool segments traffic by source and referrer in under 2 seconds, and the aggregate stats tool pulls conversion metrics for custom goals (like 'trial_started') without cookies. The API key auth means any team member can query data from Slack or a dashboard without OAuth friction. This setup works until you need session replay or user-level funnels—Plausible tracks events, not individual user journeys. If your campaigns generate under 10k monthly visits and you care about compliance, this MCP is the right call for daily standup reporting.

Use this MCP to watch Product Hunt spikes live

A 3-person product team launches on Product Hunt at 8am Pacific. They need to see visitor counts every 5 minutes to decide whether to boost the post or fix a landing page bug. The realtime visitors tool hits Plausible's API and returns current traffic in a single call—no dashboard tab-switching, no 30-second GA delays. The timeseries stats tool graphs the last 24 hours so they can compare this launch to last month's. This works because Plausible's API is fast and the team only needs top-level metrics during the launch window. If you need to drill into individual user sessions or A/B test variants in real time, this MCP won't help—it's aggregated data only. For launch-day traffic watching at small scale, Plausible's speed and simplicity beat heavier tools.

When Plausible handles weekly content retrospectives

A 5-person editorial team publishes 12 blog posts a month and reviews performance every Friday. They need to see which posts drove the most traffic, how long readers stayed, and which external sites linked in. The breakdown stats tool pulls top pages ranked by visitors, and the aggregate stats tool shows average visit duration and bounce rate for the past 7 days. The API returns clean JSON that feeds directly into a Notion database or Slack summary—no manual CSV exports. This setup breaks down if you need scroll depth, heatmaps, or per-paragraph engagement; Plausible tracks page-level metrics, not in-page behavior. For teams under 50k monthly pageviews who want fast, privacy-compliant content analytics without the Google Analytics learning curve, this MCP is the move.