Skip to main content

MASTER AI

AI Chat (Master AI)

Written by Mariano San Juan

Master Metrics - AI Chat

With Master Metrics' AI Chat, you can set up each of your agency's clients and chat directly with their data: ask how an account is performing, request optimization recommendations, and generate visual reports, all in the same place where you already have your data sources connected.

How does it work?

  1. Create a client (for example, "Store X") and add whatever context you consider relevant: industry, goals, business specifics, margins, etc. The more context you give it, the better the AI's responses and recommendations will be.

  2. Connect its data sources: Meta, Google, Tienda Nube, and the rest of the integrations available in Master Metrics.



  3. Start chatting: ask how the investment is performing, request specific metrics, comparisons, reports by ad, by ad set, by country, by audience, by day or by week, among many other variables.




    As the conversation progresses, the AI generates downloadable reports that you can adjust with your feedback (adding tables, changing the level of detail, combining sources, etc.) to share with your team or with the client.

    You'll also have access to the complete conversation history per client, so you can go back at any time to see what was discussed and what was recommended over time.

    Benefits of using it

    • Recommendations based on each client's real data: the AI automatically identifies, for example, which ad should be paused or changed and why, without you having to cross-reference the data manually.

    • Visual reports editable through feedback: you request adjustments in natural language ("add a table with the ads and their images", "show it to me by week") and the report updates instantly, ready to download and share.

    • Combined metrics across platforms: you can request blended metrics, such as ROAS considering Tienda Nube sales together with Meta and Google spend, something that would normally require building a separate spreadsheet.

    • Time savings in decision-making: instead of reviewing reports and building analyses from scratch, you ask the AI directly and move faster toward the next step.

    • Organized history per client: every back-and-forth with the AI is saved and organized by account, so you can pick it back up whenever you need to.

    Differences vs. using it from Claude (or another general AI)

AI Chat in Master Metrics

Claude (or another general AI)

Cost

Included in your Master Metrics plan, no separate subscription required

Requires a Claude subscription plus payment for an MCP to connect to the data source.

Knowledge of your business

Comes with its own benchmarks to compare your accounts' performance

Has no industry references; only works with what you share with it

History per client

Automatically saves all conversations organized by client

Does not natively keep an organized history per client

Context of each account

Already knows which account, platforms, and data belong to each client

You need to clarify which client and account you mean in every conversation

Channel

You can also chat via WhatsApp

Only used from its own platform or app

Dashboards and alerts

You can create dashboards and alerts directly from the conversation

Does not generate dashboards or alerts within the platform

Connection to platforms

Data sources (Meta, Google, Tienda Nube, etc.) are already integrated and maintained by Master Metrics

You have to connect and maintain each platform's APIs yourself

Focus

Lets you focus on your business and on decisions, not on technical infrastructure

Requires time and technical knowledge to keep integrations running

In short, Master Metrics' AI Chat gives you everything a conversational AI offers, but already integrated with your data, with context for each client, with history, and with the ability to generate dashboards and alerts, without the burden of maintaining technical connections on your own.

Share it with clients via WhatsApp so they can chat with their own data:


Did this answer your question?