Complex Requests
Teela isn't just for simple lookups. It can handle sophisticated, multi-step analysis that would normally require a data analyst, a series of queries, and a lot of spreadsheet work. This page shows you how to make the most of Teela's advanced capabilities.
What Counts as a Complex Request?
A complex request is any question that requires more than a single, straightforward query to answer well. This includes things like:
- Multi-step analysis: "What's our customer lifetime value, and how has it changed over the past year?"
- Comparisons: "How does this quarter compare to the same quarter last year, by region?"
- Trend analysis: "Show me the monthly trend of new signups over the past two years"
- Root cause analysis: "Why did churn spike in March?"
- Combined metrics: "What's the relationship between support ticket volume and customer retention?"
How Teela Handles Complex Questions
When you ask a complex question, Teela may approach it differently than a simple one. Instead of running a single query, it might:
- Break the question into smaller parts and answer each one
- Run multiple queries to gather the data it needs
- Combine results across different tables and time periods
- Put together a narrative that ties everything together
You don't need to tell Teela to do any of this. It recognizes when a question needs a more thorough approach and adapts automatically.
Investigations: Deep-Dive Analysis
One of Teela's most powerful features is investigations. When you ask an analytical "why" question, Teela goes beyond a simple answer and conducts a multi-query analysis.
For example, if you ask "Why did revenue drop in Q3?", Teela might:
- Look at overall revenue trends to confirm the drop
- Break it down by region to see if the drop was concentrated somewhere
- Check customer churn during that period
- Examine changes in average order value
- Look at product mix changes
Then it brings everything together into a clear narrative: "Revenue dropped 12% in Q3, primarily driven by a 23% decline in the West region. This coincided with the loss of three enterprise customers and a seasonal dip in new orders."
Investigations give you the kind of analysis that normally takes hours of work, delivered in moments.
Building Up to Complexity Through Conversation
One of the best strategies for complex analysis is to build up gradually through a series of follow-up questions rather than trying to pack everything into one massive question.
Example: Step-by-Step Analysis
Instead of asking everything at once, try this approach:
- Start broad: "What was our revenue last quarter?"
- Add a dimension: "Break that down by region"
- Compare: "How does that compare to the previous quarter?"
- Identify anomalies: "Which region had the biggest change?"
- Dig deeper: "Why did the West region drop so much?"
Each step builds on the last, and Teela remembers the full context of your conversation. This approach is often more effective than a single complex question because:
- You can verify each step as you go
- You can change direction if something catches your eye
- Teela has clear context at each point in the conversation
Comparing Time Periods
Teela is great at time-based comparisons. Here are some effective ways to ask:
- Quarter over quarter: "Compare Q1 to Q2 revenue by region"
- Year over year: "How do this year's sales compare to last year, month by month?"
- Before and after: "Compare the 30 days before and after our pricing change on March 1st"
- Seasonal patterns: "Show me monthly sales for the past three years and whether there are seasonal trends"
When comparing time periods, be as clear as possible about which periods you want to compare and what metric you're interested in.
Trend Analysis
Trend questions help you see patterns over time. Some good ways to phrase them:
- "Show me monthly revenue trends for the past year"
- "What's the weekly trend of new customer signups?"
- "How has average order value changed over the past 6 months?"
- "Is our support ticket volume going up or down?"
Teela will often present trends visually as charts, making it easy to spot patterns at a glance.
When Teela Takes a Multi-Angle Approach
For particularly complex questions, Teela may let you know that it's approaching the question from multiple angles. This is its way of being thorough. It might say something like "Let me look at this from a few different perspectives" and then run several analyses to give you a comprehensive answer.
That's actually a good thing because it means Teela is working hard to give you the most complete and accurate answer possible.
Exporting Complex Results
When Teela completes a complex analysis or investigation, you can export the results in two formats:
Excel Workbook
The Excel export for complex analyses includes multiple tabs:
- Summary tab with the main findings and narrative
- Supporting data tabs for each query Teela ran during the analysis
This format is ideal when you want to do further analysis, create your own charts, or share the raw data with colleagues.
PDF Report
The PDF export gives you a polished, presentation-ready document with:
- A clear narrative of the findings
- Key data points highlighted
- A structure that's easy to share with stakeholders
This is perfect for dropping into an email, sharing in a meeting, or attaching to a project update.
Tips for Complex Requests
-
Build up through conversation. Start simple and add complexity through follow-ups. This is usually more effective than one giant question.
-
Be specific about what you want to compare. "Compare revenue" is okay, but "Compare monthly revenue by region, Q1 vs Q2" gives Teela much more to work with.
-
Use investigations for "why" questions. When you want root cause analysis, frame your question with "why" and let Teela do the multi-query heavy lifting.
-
Take advantage of exports. Complex analyses are often worth saving. Use the Excel or PDF export to keep a record or share your findings.
-
Don't worry about being too ambitious. Teela will do its best with whatever you throw at it. If a question is too broad, Teela may ask a clarification or suggest a way to narrow it down.
Keep Learning
- Getting More From Your Questions: Explore follow-ups, drill-downs, data reasoning, and other ways to dig deeper
- Writing Good Queries: Foundational tips for phrasing clear, effective questions