Skip to main content

Documentation Index

Fetch the complete documentation index at: https://fromhazel.ai/docs/llms.txt

Use this file to discover all available pages before exploring further.

​
How it works

Getting answers from your data takes three steps:
1

You ask a question

Type a question the way you’d ask a coworker. No formulas, no query language — just plain English.
2

Hazel queries your connected data

Hazel figures out which data sources to pull from — Shopify, Meta Ads, Klaviyo, and anything else you’ve connected — and runs the analysis behind the scenes.
3

You get an answer with charts, tables, and citations

Hazel responds with a clear answer, the right visualization, and inline citations so you can verify every number. Ask follow-ups to dig deeper.

​
Starting a conversation

Not sure what to ask? Here are some great places to start:
  • “What was revenue last week?”
  • “Show me top products by margin”
  • “Compare this month to last month across all channels”
  • “How many new customers did we acquire this quarter?”
  • “What’s our best-performing Meta campaign right now?”
Don’t worry about phrasing things perfectly. Hazel understands natural language, so ask the way you’d naturally say it.
You can start a new conversation anytime from the chat screen. Each conversation keeps its own context, so start a new thread when you’re switching topics.

​
Follow-up questions

Hazel keeps context within a conversation thread. Once you get an answer, you can keep going:
  • “Now break that down by channel”
  • “Show me just new customers”
  • “What does that look like as a percentage?”
  • “Compare that to the same period last year”
Each follow-up builds on what came before, so you can drill into your data naturally. When you’re ready to explore a different topic, just start a new thread.

​
File uploads

You can upload CSVs and images for Hazel to analyze alongside your connected data. This is great for things like:
  • A vendor report you received via email
  • A competitor’s pricing spreadsheet
  • A screenshot of a dashboard you want Hazel to interpret
Just drag and drop the file into the chat, or click the upload button. Hazel will read the file and use it in your conversation.
Currently only CSV and image files are supported. If an image upload errors, it’s likely too large — try a smaller file or crop to the relevant section.
Hazel can search the web to bring in context beyond your own data. This is a powerful feature for:
  • Competitor research — “What promotions is Brand X running right now?” “What’s Brand Y’s pricing strategy?”
  • Recent trends — “What are the trending products in our category?” “What are the latest DTC marketing trends?”
  • Industry benchmarks — “How does our email open rate compare to the industry average?” “What’s a good ROAS for our category?”
  • Company/address research — “What can you find out about this company?” “Look up this address for me”
Hazel combines web results with your data to give you richer, more complete analysis.
Next up: Learn the difference between Ask and Deep Thinking so you can pick the right mode for your question.