Agents
What it is:
Automated, repeatable workflows that pull data, analyze it, and generate insights or reports on a recurring basis.
Best for:
Recurring weekly or daily reports
Budget recommendations
Inventory forecasting
Creative analysis
Cross-platform data synthesis
Example Use Cases:
“Send a weekly Meta performance report every Monday at 9 AM.”
“Notify us if any ad set has NC ROAS < 1.5.”
“Forecast which SKUs are at risk of stockout and generate this report on a weekly basis.”
Key Features/Differentiators:
Can pull from multiple sources (Meta, Shopify, Google, Klaviyo, etc.)
Logic + Analysis + Reporting + Output steps
Supports scheduling
Deterministic and consistent meaning it will always give you the same output.
Outputs to Slack, Email, Google Sheets, Feed, Dashboards
When to use over the others:
When the insight needs to run on a schedule
When analysis includes logic, thresholds, or multi-step comparison
When you need a repeatable business process
Agents have access to all the tools
Build Mode (inside Moby)
What it is:
Conversational way to build an Agent by describing what you want in plain English. Moby will scaffold the Agent for you.
Best for:
Quickly creating complex Agents without touching the step-by-step builder
Users uncomfortable with SQL or structured query tools
Iterating on Agent requirements in real time
Example Use Cases:
It’s the place to ask Moby questions that are related in an exploratory manner, then turn the outputs into an agent
“Create an Agent that identifies our top SKUs and which channels drive them.”
Key Features/Differentiators:
Natural language interface
Clarifies goals and requirements as you build
Creates full Agent flows: data > analysis > output > schedule
Can review/edit the output in the visual Agent Builder
When to use over the others:
When you know what outcome you want and KNOW the steps to get there
When the user knows what they want but doesn’t want to build step-by-step
When speed and ease matter
When you want to guide a customer toward recurring insight automation
Deep Dive
What it is:
Moby’s long-form research tool for answering complex, multi-layered business questions using structured AI reasoning.
Best for:
One-off, exploratory questions that span multiple datasets
Identifying underlying causes or correlations
Discovering complex patterns within your business
Example Use Cases:
“Why is our Meta ROAS down this month but TikTok is up?”
“What factors influence high LTV across our customer base?”
“Which channels are most effective for acquiring high-AOV customers?”
Key Features/Differentiators:
Deep Dive creates a research plan, pulls data from different sources, and synthesizes results
Visual, step-by-step output
Can convert findings into a recurring Agent with one click
Not deterministic and not consistent. Outputs may vary even with the same prompt.
When to use over the others:
When you know what outcome you want but DON’T know the steps to get there.
When the business question is “why” or “how” not just “what”
When you’re exploring new ideas or diagnosing problems
When a client wants deeper, strategic insight (not just metrics)
Deep Dive does not have access to all tools that agents do
Custom Dashboards
What it is:
A user-defined dashboard view that displays real-time metrics, charts, and visualizations relevant to the business.
Best for:
Ongoing monitoring of key metrics
Executive-level reporting
Embedding Agent outputs for visibility
Creating shared views for different departments (Ops, Media, Finance)
Example Use Cases:
“Show me ROAS by platform, NC ROAS, and AOV in one view.”
“Create a Product Performance tab for my eCom manager.”
“Add a top-line P&L view with daily trends.”
Key Features/Differentiators:
Deterministic. It’s the same data every single day.
The data is in real-time unlike an agent or deep dive that needs to be run.
Drag-and-drop UI to customize layout
Embed Agent outputs directly
Assign user-level permissions
Supports visual charts, KPI blocks, tables
When to use over the others:
When someone needs a centralized home for metrics
When stakeholders want real-time data at a glance
When embedding Agent outputs for cross-functional reporting
Moby
What it is:
A conversational interface for querying your data using natural language — designed for ad hoc questions, on-the-fly analysis, and quick insights across your connected platforms.
Best for:
Quick answers to unstructured questions
Exploring performance anomalies
Strategic or creative brainstorming with data
Ad hoc comparison and benchmarking
Live support during customer calls or team syncs
Example Use Cases:
“How did our Klaviyo flows perform last week?”
“Which campaigns had the highest ROAS in May?”
“Why did revenue dip last Thursday?”
“What was the open rate trend for our welcome email over the last month?”
Key Features/Differentiators:
Freeform natural language input — no prompt engineering required
Fast, one-off access to data without building workflows
Supports follow-up Q&A in conversational thread
Pulls from multiple data sources (Meta, Shopify, Klaviyo, Google)
Ideal for exploring before formalizing in Build Mode or an Agent
When to use over the others:
When you need quick insights or exploratory analysis
When you’re unsure what you’re looking for yet
When follow-up questions or deeper digging is expected
When you’re on a live call or need flexible conversation-style support