Scrunch integrations: API, MCP, exports, BI and webhooks
A buyer-focused map of Scrunch APIs, MCP access, CSV and PDF exports, Data Studio, warehouse loading, GA4, Agent Traffic connections, webhooks, and plan gates.
Quick answer
Which Scrunch integrations are documented, and what can each one actually do?
Scrunch has a substantial integration surface: aggregate and raw-response APIs, a hosted OAuth MCP, CSV, Excel and PDF exports, a private Data Studio connector, GA4 referrals, Agent Traffic ingestion, and documented warehouse-loading patterns. The Data API is listed on Enterprise, while MCP is documented for any Scrunch plan. Public webhook documentation is narrower, covering pipeline-stage callbacks rather than a full event catalog. Buyers should map direction, data depth, history, and entitlement before implementation.
Key facts and evidence
- Data API gate
- Enterprise on the current pricing pageEvidence: Scrunch pricing and plan comparison
- MCP entitlement
- Documented for any Scrunch plan with account permissions appliedEvidence: Scrunch hosted MCP overview
- API history
- Latest 90 days for aggregate Query and raw Responses APIsEvidence: Does Scrunch show change over time, Scrunch API products and operation
- Data Studio
- Private connector for Agency and Enterprise organization adminsEvidence: Using Scrunch Data Studio templates
- Webhooks
- Pipeline-stage completion callbacks publicly documentedEvidence: Scrunch developer documentation indexVerification: partially verified. The public index confirms a narrow callback use case. Buyers needing arbitrary alerts should request the supported event catalog.
Which Scrunch integration path fits the job?
| Integration path | Documented use | Access or limit | Best buyer fit |
|---|---|---|---|
| Query and Responses APIs | Aggregate metrics and raw answer-level records for analysis and pipelines | Data API is listed on Enterprise; public API history is 90 days | Teams building a governed warehouse or custom BI modelEvidence: Scrunch developer documentation index, Scrunch API products and operation, Scrunch pricing and plan comparison, Does Scrunch show change over time |
| Hosted MCP | Live analysis, configuration operations, and Agent Traffic in compatible AI clients | Any Scrunch plan; 1,000 items per request and incomplete page-audit coverage | Analysts and agents that need conversational read and controlled write accessEvidence: Scrunch hosted MCP overview |
| Exports | CSV or Excel tables, PDF reports, and scheduled API pulls | Static files lose live permissions and API scheduling still needs entitlement | Ad hoc sharing, audit snapshots, and simple offline analysisEvidence: Scrunch data export options, Scrunch pricing and plan comparison |
| Data Studio | Private multi-brand templates for visibility and Agent Traffic reporting | Agency or Enterprise admins; connector is not in the public gallery | Agencies and enterprise teams standardizing client dashboardsEvidence: Using Scrunch Data Studio templates |
| Traffic and referral inputs | GA4 referrals plus CDN, host, WordPress, and custom Agent Traffic ingestion | Each brand connection and site environment must be configured | Teams joining visibility with visits, sessions, conversions, and AI bot activityEvidence: Connect AI Referrals to Google Analytics, Supported CDN integrations for Agent Traffic |
| Webhooks | Callbacks when documented processing pipeline stages complete | No public general-purpose event catalog was found | Engineering workflows tied to Scrunch pipeline completionEvidence: Scrunch developer documentation indexVerification: partially verified. The public index confirms a narrow callback use case. Buyers needing arbitrary alerts should request the supported event catalog. |
Checked 25 August 2026. Direction matters: some connections bring analytics or logs into Scrunch, while APIs, exports, Data Studio, MCP, and webhooks move or expose Scrunch data elsewhere.
What can the Scrunch API return, and how much history is available?
Scrunch documents aggregate Query data and raw Responses, plus APIs for Agent Traffic, AI Referrals, audits, signals, sitemaps, AXP rendering, deployment, and configuration. The raw path includes response-level material suitable for a data mart.
The public history window for Query and Responses is 90 days. Scrunch publishes an incremental loading pattern for BigQuery, Snowflake, Redshift, and Databricks, which matters if the team needs a longer governed record.
When should a buyer use Scrunch MCP instead of the API?
MCP is the easier path for an analyst or AI client that needs live, permission-aware questions and configuration operations. It uses OAuth, inherits account access, and is documented for any Scrunch plan.
The API is the better fit for repeatable engineering pipelines and warehouse loads. Scrunch lists the Data API on Enterprise. MCP also has a 1,000-item request maximum and does not fully expose URL details, sitemap, or page-audit operations.
Can Scrunch feed BI tools without a custom engineering project?
Scrunch documents table exports to CSV and Excel, PDF reports, and a private Data Studio connector. The current Data Studio help page scopes organization-admin access to Agency and Enterprise and says the connector is not listed in the public gallery.
For a governed warehouse, the published pattern uses the Responses API and upserts stable records into BigQuery, Snowflake, Redshift, or Databricks. That path still requires API entitlement and an owned ingestion job.
How does Scrunch connect AI referrals and bot traffic to outcomes?
A per-brand GA4 connection uses read-only OAuth to show AI referrals, sessions, and conversions. Agent Traffic is different: it ingests server or edge evidence through documented Cloudflare, WordPress, Vercel, Akamai, and custom API paths, with more CDN and hosting guides available.
Buyers should scope both connections. GA4 describes human referral behavior after a click; Agent Traffic describes automated AI systems reaching the site.
Does Scrunch have webhooks for alerts and workflow automation?
Scrunch's developer index documents webhook callbacks for pipeline-stage completion. That supports engineering flows waiting for Scrunch processing to finish.
The current public index does not present a broad event catalog for visibility drops, citation changes, or user actions. A buyer needing general outbound alerts should request the exact event list, authentication method, retry behavior, and delivery history.
Evidence and method
Scrunch separates aggregate and raw data APIs
The documented Query and Responses split gives BI buyers a clear choice between ready-to-use metrics and answer-level records for custom analysis.
Evidence: Scrunch developer documentation index, Scrunch API products and operation, Does Scrunch show change over timeMCP includes permission-aware operations
Scrunch's hosted OAuth MCP is more than a static knowledge connector, because it exposes live data and configuration tools while preserving account permissions.
Evidence: Scrunch hosted MCP overviewWarehouse loading has a published pattern
The data-mart guide names common warehouses and an incremental upsert method, reducing ambiguity for teams planning durable response-level history.
Evidence: Load Scrunch data into a data martTraffic connections cover both referrals and bots
GA4 provides human referral outcomes while server and edge integrations provide AI bot evidence, allowing buyers to keep two different traffic questions separate.
Evidence: Connect AI Referrals to Google Analytics, Supported CDN integrations for Agent TrafficWebhook scope is documented narrowly
Scrunch confirms pipeline-completion callbacks, but the lack of a public general event catalog means broader workflow claims should be tested during procurement.
Evidence: Scrunch developer documentation indexVerification: partially verified. The public index confirms a narrow callback use case. Buyers needing arbitrary alerts should request the supported event catalog.How we checked this page
We reviewed Scrunch's current developer index, API, MCP, export, BI, warehouse, GA4, and Agent Traffic documentation, then classified each connection by direction, data depth, limit, and plan gate.
- 1. Separated data extraction, conversational access, reporting connectors, referral analytics, bot-log ingestion, and callback webhooks into distinct integration jobs.
- 2. Mapped every entitlement statement to current pricing or help documentation and did not infer access from the existence of an endpoint.
- 3. Compared Trakkr using current official integrations, pricing, and MCP documentation checked on the same date.
- Limitation: We did not create API keys, authorize a live connector, inspect private rate limits, or test webhook delivery and retry behavior.
- Limitation: Private enterprise connectors and negotiated access may exist beyond the current public documentation.
When is Scrunch or Trakkr the stronger integration fit?
Choose Scrunch when raw-response extraction, a documented multi-warehouse loading pattern, Data Studio for many brands, or deep site bot ingestion is central. Its public developer surface is broad and its MCP is documented for any plan.
Choose Trakkr when the team needs a wider documented set of workflow destinations for actions, alerts, task tools, collaboration, CMS publishing, and webhooks. Trakkr offers MCP on every paid plan, while REST API and Looker Studio sit on Scale.
Yes. Scrunch documents aggregate Query and raw Responses APIs plus several traffic, audit, signal, sitemap, AXP, deployment, and configuration endpoints. Current pricing lists the Data API on Enterprise.
Yes. Scrunch documents a hosted OAuth MCP for any Scrunch plan, with live analysis, configuration, and Agent Traffic operations subject to the user's existing permissions and published limits.
Scrunch documents a private Data Studio connector with multi-brand templates for Agency and Enterprise organization administrators. It is not listed in the public connector gallery.
Scrunch publishes a Responses API loading pattern for BigQuery, Snowflake, Redshift, and Databricks. The buyer owns the incremental extraction and upsert pipeline, and needs API entitlement.
The public developer index confirms callbacks for pipeline-stage completion, not a complete general event-notification catalog. Confirm required event types and delivery behavior before relying on webhooks for operations.
Sources and related reading
See how AI talks about your brand
Enter your domain to get a free AI visibility report in under 60 seconds.