ThoughtSpot Spotter MCP

@ThoughtSpot what were our top 5 revenue-generating products last quarter, broken down by region?

Based on your ThoughtSpot data model, the top 5 products last quarter were: Product A ($2.3M, led by EMEA), Product B ($1.8M, strongest in APAC), Product C ($1.6M, North America-heavy), Product D ($1.4M, evenly distributed), and Product E ($1.1M, Latin America focus). This query ran against the 'Sales Performance' worksheet and returned results in 1.2 seconds. Note that figures reflect closed deals only; pipeline data requires a separate query.

This example shows natural-language querying against your existing ThoughtSpot data models. The MCP translates the prompt into ThoughtSpot's search syntax, executes it, and returns formatted results. You'll need OAuth access to the specific worksheets or liveboards the query references—scope your connection accordingly.

@ThoughtSpot create a liveboard pinning our monthly active users trend, churn rate by cohort, and NPS score over the last 12 months

Created liveboard 'Product Health Dashboard' with three visualizations: (1) MAU line chart showing 18% growth from Jan to Dec, (2) cohort churn heatmap highlighting elevated churn in Q2 2024 signups, and (3) NPS gauge averaging 42 across the period with a dip in August. Liveboard is now accessible at [example-url]. You can share it with your team or schedule email snapshots from the ThoughtSpot UI.

This demonstrates programmatic liveboard creation—useful for templating dashboards across teams or automating reporting workflows. The MCP constructs the liveboard via ThoughtSpot's API but doesn't control sharing permissions; you'll configure those in ThoughtSpot directly. Creation may take 10-15 seconds for complex visualizations.

@ThoughtSpot analyze our customer support ticket volume by category and suggest which areas need staffing adjustments based on resolution time trends

Ticket volume analysis shows 'Billing Issues' spiked 34% month-over-month (now 1,200 tickets/month) while average resolution time climbed from 2.1 to 3.8 days. 'Technical Troubleshooting' remains flat at 800 tickets/month but resolution time improved to 1.5 days. Recommendation: consider reallocating 1-2 support engineers from Technical to Billing, or investigate root causes in the billing workflow—three specific ticket threads show recurring confusion about invoice line items.

This pairs ThoughtSpot's data retrieval with the AI's analytical reasoning. The MCP fetches raw metrics; the AI interprets trends and proposes actions. Accuracy depends on your data model's completeness—if resolution time isn't tracked consistently, recommendations will be speculative. Always validate suggestions against operational context.

When OAuth2 analytics access beats manual dashboard exports

An 8-person SaaS team runs weekly revenue reviews where the VP of Sales, two account execs, and the founder compare pipeline metrics across regions. Without this MCP, someone exports CSV files from ThoughtSpot on Monday morning, pastes them into a Google Doc, and the team argues about whether the numbers are current. With OAuth2 authentication, the MCP pulls live ThoughtSpot data into a shared Switchy workspace where the team annotates trends and flags outliers in real time during the call. The trade-off: if your ThoughtSpot instance has row-level security or complex entitlements, OAuth2 scope might not surface the right subset for every team member. This works when your team shares a single analytics view and needs to stop the export-paste-stale-data loop.

When live BI queries replace static support runbooks

A 12-person customer success team triages accounts flagged as at-risk each morning. The current process: a CSM logs into ThoughtSpot, runs three saved searches (usage drop, support ticket spike, payment issues), screenshots the results, and drops them in Slack. The team then debates next steps in a thread that's hard to search later. This MCP lets the team query ThoughtSpot directly from a Switchy workspace, tag accounts that need outreach, and build a living triage log that references the exact query parameters used each day. The boundary: if your success team doesn't already trust ThoughtSpot as the source of truth, adding an MCP won't fix upstream data quality problems. This is for teams who live in their BI tool and want to stop context-switching between Slack, screenshots, and the actual platform.

When feature adoption metrics need to live next to Jira tickets

A 6-person product team meets biweekly to rank feature requests in Jira. The PM manually pulls ThoughtSpot dashboards showing which features get the most usage, pastes screenshots into a Notion doc, and the team votes on what to build next. Two weeks later, no one remembers which usage numbers justified the decision. With this MCP, the team queries ThoughtSpot adoption metrics directly in the Switchy workspace where they're already debating priorities, links each metric to the corresponding Jira ticket, and creates a decision trail that survives the meeting. The catch: if your product team prioritizes qualitative feedback over quantitative usage, this MCP won't change your process—it just speeds up the data-lookup step. Buy this if your roadmap lives or dies by BI metrics and you're tired of stale screenshots.