New in Amperity: More Ways To Put Customer Data to Work
Four new capabilities are live in Amperity, and together they close the gap between resolved customer data and the actions teams take on it.
Key Takeaways
The Amperity MCP server is generally available. Connect Copilot Studio, Claude, ChatGPT, or any MCP client to your tenant with your existing login, and it stays scoped to that user's permissions.
Recommended Actions (public preview, retail) reads your unified data and hands back a ranked list of plays, each with a target segment and expected impact, ready to send into a journey.
The Event Propensity model scores who's likely to take a revenue-generating action in the next 30 days and builds the audience for you.
Paid Media Measurement & Optimization pulls Meta and Google performance in next to your customer data, so AmpAI can answer spend questions in plain language.
Amperity MCP server is generally available
Your teams can now connect any MCP-compatible client, including Copilot Studio, Claude, and ChatGPT, straight to your Amperity tenant and work in natural language. Amperity MCP server, built on the Model Context Protocol (MCP), hosts this connection as a single service, so there's nothing to install locally. Connecting a client points it at one endpoint, mcp.amperity.com, with a setup guide for each supported client.
Once connected, a user can inspect data, run Identity Resolution, build queries and segments, create activations, and run campaigns and journeys, all by asking. More than 200 tools are available to connected clients by default. Every request authenticates with the user's existing Amperity credentials and is scoped to what that user is already allowed to see and do, so governance holds no matter which AI tool sits on top.
This puts governed customer intelligence inside the tools teams work in every day, without copying data into another system to get there.
Recommended Actions, now in public preview for retail
Recommended Actions turns your unified customer data into a prioritized list of marketing plays instead of a dashboard you have to interpret. Each recommendation names the opportunity it addresses, the customer segment it targets, and the expected business impact, along with the reasoning behind that estimate.
Recommended Actions is powered by AmpAI. It groups customers into value tiers, drafts the analysis, and explains why a given play is worth running. When you decide to act, you send the target segment straight into a journey or a campaign, so a recommendation becomes activation in a few clicks rather than a handoff between teams.
Recommended Actions is in public preview for retail customers. Contact your account team for more details.
The Event Propensity model
The Event Propensity model predicts how likely a customer is to perform a revenue-generating event in the next 30 days, then groups customers by recommended audience size and by ranking so you can build the audience that fits the campaign.
Point it at the events you care about: loyalty program signups, credit card signups, a repeat booking for a hotel stay or a flight. It joins the Predictive Models family alongside predicted CLV, product affinity, and churn propensity, and its output drops into the segment editor for immediate use.
Paid Media Measurement & Optimization
Paid Media Measurement & Optimization brings your Meta and Google performance data into Amperity, where it sits in the Customer 360 database next to your customer data. A measurement courier pulls the data in on a daily schedule and backfills roughly the prior year on the first run.
Because that spend data lands alongside customer context, AmpAI can analyze it in natural language, without SQL. Ask AmpAI to rank campaigns by return on ad spend, compare Facebook and Instagram over the last month, or flag campaigns with a declining click-through rate, and the answer draws on the familiar paid media metrics you'd expect: ROAS, CPA, CTR, CPM, and CPC.
Built to run on your lakehouse
These capabilities run on a foundation that increasingly lives where your data already does. Amperity Bridge and Bring Your Own Storage keep customer data in your lakehouse, and Bring Your Own Compute (BYOC), in preview for Databricks, runs customer intelligence workloads on your own Databricks compute, so the data never leaves your environment.
Get started
These capabilities are available to Amperity customers. Talk to your Amperity account team for more information, or contact sales to see what your customer data can do.