braze · MCP Server

Braze + Improvado MCP — Campaign Intelligence, No Dashboard Required

Improvado gives your AI agent direct access to Braze data through an MCP server. Query send volumes, open rates, conversion funnels, and segment sizes in natural language. Works with Claude, ChatGPT, Cursor, and any MCP-compatible tool.

46K+ metrics ·Read & Write access ·500+ platforms ·<60s setup

Read

Read: Pull Any Braze Metric Instantly

Stop exporting CSVs and building manual Braze reports. Ask your AI agent for campaign performance, churn cohorts, message fatigue signals, or A/B test results — across any channel, segment, or date range. The MCP server handles Braze API calls.

Example prompts

"Which email campaigns had open rates below 15% last month? Break down by segment and send time."

40 min → 30 sec

"Show me push notification CTR by platform for the last 30 days. Compare iOS vs Android."

20 min → 15 sec

"Which active canvases have had a drop in conversion rate over 20% in the last two weeks?"

1 hr → 1 min

Works with Claude ChatGPT Cursor +5

Write

Write: Update Campaigns Without Leaving the Chat

Your AI agent reads Braze data and acts on it. Pause underperforming campaigns, update segment filters, and trigger A/B test variants — without navigating the Braze dashboard.

Example prompts

"Pause all in-app message campaigns with click-through rates below 2% that have been running over 14 days."

30 min → 45 sec

"Update the re-engagement canvas entry criteria to exclude users active in the last 7 days."

20 min → 30 sec

"Clone the Spring Sale email campaign, update the send date to next Friday, and increase the control group to 20%."

1.5 hrs → 10 min

Every action logged · Fully reversible · SOC 2 certified

Monitor

Monitor: Catch Engagement Drops Before They Compound

Set up watches on the metrics that matter. Your AI agent monitors Braze campaigns continuously and flags anomalies — message fatigue, deliverability issues, or canvas drop-offs — before they erode engagement.

Example prompts

"Alert me if unsubscribe rate for any email campaign exceeds 0.5% within 24 hours of a send."

Manual → auto

"Every Monday at 8am: send a digest of weekly send volume, average open rate, and top 3 converting canvases."

2 hrs → auto

"Flag any canvas step with a conversion rate drop of more than 15% vs the prior 7-day average."

Manual → auto

Alerts sent to Slack, email, or your AI agent

Full cycle

The Closed Loop: Read → Decide → Write → Monitor

Your AI agent doesn't just surface data — it acts. Adjust pricing, update product descriptions, manage inventory, apply discounts — all through natural language. The MCP server translates intent into API operations.

Every phase runs through the same MCP connection. One protocol, all platforms, full governance. No switching between tools.

Ideate

Launch

Measure

Analyze

Report

Iterate

One conversation. All six phases. Every platform.

The daily grind

Common problems. Direct answers.

Challenge 1

Canvas Performance Is Invisible Until It's Too Late

The problem

Multi-step canvases accumulate drop-offs across 10+ steps. Identifying which step is leaking users requires pulling step-level data, calculating drop-off rates, and comparing against historical benchmarks — all manually. By the time someone notices, weeks of send volume are wasted.

How MCP solves it

Improvado extracts full canvas step-level data from Braze and makes it queryable via AI. Ask for drop-off rates at every step in one question. The MCP server surfaces the leak immediately, with context on when the drop-off started.

Try asking

Show step-by-step conversion for the onboarding canvas. Where's the biggest drop-off?

Answer in seconds

All data sources, one query

Challenge 2

Segment Overlap Causes Message Fatigue

The problem

Users enrolled in multiple canvases receive overlapping messages from different teams. There's no consolidated view of how many messages a user segment receives per week across all campaigns. Fatigue builds silently until unsubscribe rates spike.

How MCP solves it

Improvado normalizes send-level data across all Braze campaigns and canvases into a unified model. Query total message frequency by segment, identify overlap, and spot fatigue signals before unsubscribes climb.

Try asking

How many messages per week are users in the high-value segment receiving across all active canvases?

Full detail preserved

No data loss on export

Challenge 3

Cross-Channel Attribution Is Guesswork

The problem

Email, push, in-app, and SMS each live in separate Braze reporting views. Comparing which channel drives the most conversions for a single cohort requires exporting four separate reports and manually joining them. Attribution decisions are made on incomplete data.

How MCP solves it

Improvado's data model unifies Braze channel data into one schema. The MCP server lets your AI compare email vs push vs SMS conversion rates for the same cohort in a single query, with no manual exporting.

Try asking

For the re-engagement cohort, compare email, push, and SMS conversion rates over the last 30 days.

Unified data model

Compare anything side by side

👥 Teams

One Framework. Five Roles. Zero Setup.

Same MCP connection, different workflows for every team member. Each role asks in natural language — the MCP server handles the complexity (rate limits, auth, schema normalization, governance) behind the scenes.

Agency CEO

Portfolio health. Client risk. Revenue signals.

Media Strategist

70% strategy, not 70% ops. Auto campaign QA.

Marketing Analyst

Zero wrangling. Cross-platform. AI narratives.

Account Manager

QBR decks auto-generated. Call prep in 30s.

Creative Director

Performance-to-brief. Predict winners before spend.

FAQ

Common questions

What Braze data can I access through the MCP server?

Campaign and canvas performance metrics (send volume, opens, clicks, conversions, revenue), segment membership and sizes, A/B test results, message fatigue indicators, subscription states, and channel-level breakdown across email, push, in-app, and SMS.

Does this work with Braze canvases or just campaigns?

Both. You can query individual campaigns and multi-step canvases, including step-level conversion data. Canvas step analysis — which is notoriously hard to get in bulk — is one of the most common use cases.

Which AI tools work with this Braze MCP server?

Any tool supporting the Model Context Protocol. That includes Claude Desktop, ChatGPT, Cursor, Windsurf, Gemini, and custom applications using MCP HTTP transport. Claude is the most commonly used client due to native MCP support — all through Improvado's hosted MCP server.

Can the AI write back to Braze or just read data?

Both. Read operations cover all analytics and reporting data. Write operations include pausing campaigns, updating canvas settings, adjusting segment filters, and triggering test sends. Permissions are scoped to what your Braze API key allows.

Is my Braze data secure through the MCP server?

Yes. Improvado stores all API credentials in an encrypted vault under SOC 2 Type II controls. Your AI agent sends queries through Improvado's secure proxy — raw API keys are never exposed to the AI model. Prompt injection protection is built into the server.

How long does setup take?

If you're already an Improvado user with Braze connected, your data is ready. Open the AI Agent at app.improvado.io/agent and start querying. For Claude Desktop or Cursor, add one line to your MCP config — under 60 seconds.

Stop Reporting. Start Executing.

Connect your data to an AI agent in under 60 seconds. The closed loop starts with one conversation.

SOC 2 Type II GDPR 500+ Platforms