> ## Documentation Index
> Fetch the complete documentation index at: https://docs.marcenta.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agentic Analytics: Explore Marketing Data with AI Assistance

> Use Agentic Analytics to explore your marketing data with structured or natural-language queries, then analyze results with AI Insights.

Agentic Analytics (Queries) is an interactive data explorer that lets you ask questions about your marketing data without writing code. You can compose a query by selecting a source system, metrics, dimensions, and filters, and Marcenta returns the results as a table and chart. You can also type a question in plain language and the AI will interpret it, run the query, and explain what it found - all in the same interface.

<script type="application/ld+json">
  {`{
    "@context": "https://schema.org",
    "@type": "FAQPage",
    "mainEntity": [
      {
        "@type": "Question",
        "name": "Do I need SQL to use Agentic Analytics?",
        "acceptedAnswer": {
          "@type": "Answer",
          "text": "No. You can run structured queries through the UI or ask questions in plain language."
        }
      },
      {
        "@type": "Question",
        "name": "Can I save and reuse my analysis?",
        "acceptedAnswer": {
          "@type": "Answer",
          "text": "Yes. Query conversations are saved in the sidebar so you can reopen, refine, and rerun prior analyses."
        }
      },
      {
        "@type": "Question",
        "name": "What should I do after finding an issue in Queries?",
        "acceptedAnswer": {
          "@type": "Answer",
          "text": "Move to Incidents Triage for prioritization and use Dispatch for daily operational context."
        }
      }
    ]
    }`}
</script>

## Navigate to Agentic Analytics (Queries)

Select **Queries** in the left sidebar to open Agentic Analytics. The interface is split into a **conversation area** on the right and a **conversations sidebar** on the left, where previous query sessions are saved for you to revisit.

## Run a structured query

<Steps>
  <Step title="Open the query composer">
    Select the input area at the bottom of the Queries page to open the composer.
  </Step>

  <Step title="Select a source system">
    Choose the data source you want to query. Available sources are:

    * **Google Ads**
    * **Google Analytics 4**
    * **Google Search Console**
    * **LinkedIn Ads**
    * **HubSpot**
  </Step>

  <Step title="Choose metrics and dimensions">
    Select one or more metrics and the dimensions you want to group by. The available options update based on your selected source and the data Marcenta has ingested.
  </Step>

  <Step title="Set the date range">
    Pick a start and end date for the query. You can also add filters to narrow results to a specific campaign, country, device, or any other available dimension.
  </Step>

  <Step title="Run the query">
    Submit the query. Results appear as a data table and, where applicable, as a chart.
  </Step>
</Steps>

## Available metrics

The metrics available depend on the source system you select. Common metrics across sources include:

<Tabs>
  <Tab title="Performance">
    * `impressions` — total ad or search impressions
    * `clicks` — total clicks
    * `ctr_pct` — click-through rate as a percentage
    * `conversions` — total conversion events
    * `spend` — total ad spend
    * `cpc` — cost per click
  </Tab>

  <Tab title="Web analytics">
    * `sessions` — website sessions (GA4)
    * `users` — unique users (GA4)
    * `bounce_rate_pct` — session bounce rate (GA4)
    * `avg_position` — average search result position (Google Search Console)
  </Tab>

  <Tab title="Pipeline (HubSpot)">
    * `deal_value_created` — value of deals created
    * `deals_created` — number of deals created
    * `deals_closed_won` — number of closed-won deals
    * `contacts_created` — number of new contacts
  </Tab>
</Tabs>

## Frequently asked questions

<AccordionGroup>
  <Accordion title="Do I need SQL to use Agentic Analytics?">
    No. You can run structured queries through the UI or ask questions in plain language.
  </Accordion>

  <Accordion title="Can I save and reuse my analysis?">
    Yes. Query conversations are saved in the sidebar so you can reopen, refine, and rerun prior analyses.
  </Accordion>

  <Accordion title="What should I do after finding an issue in Queries?">
    Move to [Incidents Triage](/features/flags) for prioritization and use [Dispatch](/features/dispatch) for daily operational context.
  </Accordion>
</AccordionGroup>

## Related links

* [Dispatch](/features/dispatch)
* [Incidents triage](/features/flags)

## Page trust signals

* Last reviewed: May 20, 2026
* Maintainer: Marcenta Team

## Available dimensions

Dimensions let you group and slice your data. Common dimensions include:

| Dimension       | Description                                              |
| --------------- | -------------------------------------------------------- |
| `date`          | Day-level time series                                    |
| `campaign`      | Campaign name                                            |
| `channel_group` | High-level channel grouping (e.g., Paid Search, Organic) |
| `source`        | Traffic source                                           |
| `medium`        | Traffic medium                                           |
| `ad_group`      | Ad group name (Google Ads, LinkedIn Ads)                 |
| `keyword`       | Keyword term (Google Ads, Google Search Console)         |
| `country`       | Country of the user or target                            |
| `region`        | Sub-national region                                      |
| `city`          | City                                                     |
| `landing_page`  | Landing page URL                                         |
| `device`        | Device category (desktop, mobile, tablet)                |

Additional dimensions available for LinkedIn Ads include `job_function`, `job_title`, `seniority`, `company_size`, and `industry`.

## Add filters

Filters let you narrow a query before it runs. Each filter consists of a dimension field, an operator (equals, not equals, contains, greater than, etc.), and a value. You can add multiple filters and combine them with **AND** or **OR** logic.

<Tip>
  Filters run at query time, so they reduce the dataset before Marcenta aggregates results. This makes filtered queries faster and more focused than filtering a large result set afterward.
</Tip>

## Use natural language

Instead of composing a query manually, you can type a question directly into the input area. For example:

* "Show me clicks and impressions by campaign for the last 30 days from Google Ads"
* "What were my top 5 landing pages by sessions last month?"
* "Compare conversion rates by device for LinkedIn Ads this quarter"

Marcenta interprets your question, constructs the query, and returns the results with a text explanation of what it found.

## AI Insights

After any query runs, you can select **AI Insights** to get a written analysis of the results. The AI looks at patterns in the data—trends, outliers, top performers, underperformers—and summarizes them in plain language. You can ask follow-up questions in the same conversation to dig deeper.

<Note>
  AI Insights works on the result data that Marcenta already retrieved. It does not re-run the query; it analyzes the rows already returned. For very large result sets, Marcenta analyzes up to 500 rows.
</Note>

## Save and revisit queries

Every query session is saved automatically as a conversation in the left sidebar. Select any past conversation to reload the context and continue where you left off, or start a new session from the top of the sidebar.

## Export query results

To use query results outside of Marcenta, copy the data from the results table or use the export option where available.