Human conversation can get very complicated and so can the questions you (or your accounting or bookkeeping clients) have about business and financial data.
Though not always perfect, asking questions and receiving answers in real-life conversation tends to be an intuitive and collaborative process. But when it comes to human-to-human interactions, we may take for granted the nuances and norms that enable such productive conversations. When we’re talking to a human, we can stop and clarify as many times as necessary to feel confident that we are understanding intent and effectively communicating what we need. Because we practice this all the time, we do all of this very naturally, perhaps even subconsciously.
When performing in-depth data exploration processes using Chata, asking complex queries can lead to powerful outcomes. But doing so can require a bit of critical thinking. Once you’ve got the basics covered, you’ll be querying naturally in no time at all.
Structure Your Queries Intentionally
At the beginning of your data exploration journey, you can think of structuring your queries in the context of a tree. Begin with the roots and branch out from there. The roots of your tree are the domain entity (or theme) that your query falls under (for example, “expenses” or “sales”). From there, you can branch off by applying different dimensions to your queries (like filtering or grouping data) in order to reach different outcomes.
First, you need to determine what type of answer you are looking to obtain. Specifically, are you looking for a calculation or are you looking for a list?
If you’re looking for a calculation, determine whether you are looking for a total, an average, a count, or a change. Common query words that can be used to return a calculation include:
- Change in
- Percentage of
This is a jumping-off point for almost any question you’ll want answered in your data exploration journey with Chata, but here are a handful of example queries you can start with:
“Total cost of managing business this month”
“Average days to pay invoices over $5000”
“Sum of all outstanding AR”
“Change in current ratio by month last year”
“Compare profit per year”
“Percentage of sales by class”
If you’re looking for a list, determine what you are interested in (e.g. are you looking at sales? revenue? expenses? etc.) and specify your query accordingly. Common query words that can be used to return a list include:
- List all
- Show me all
Some sample queries that will yield a list include:
“All products with inventory on hand”
“List all AR last month”
“Show me all expenses this month”
Narrow Down Your Question
Next, ask yourself if there are specific segments that you are interested in filtering or sorting, so you can get to the precise details you’re looking for.
In Data Chat, even if you’ve asked a high-level question, it’s easy to filter your results or ask a more specific follow-up query. Asking follow-up queries is part of diving deeper in your exploratory analysis process. You can also ask for a more detailed, filtered response within your initial query right off the bat. For example, you can ask for your results to be displayed “High to low” or “Low to high” depending on the information you’re after.
To return a filtered result, try including one of these terms in your query:
- Greater than
- Less than
- Over [x amount]
- Under [x amount]
You can also filter the data by time period by adding specific time conditions like:
- In the past year
- Over last 6 months
- Last month
- This month
Some queries you can try include:
“List all sales high to low”
“All sales greater than $5000”
“Sales less than $1000 “
“All sales last year”
Multiple filters can be applied within a single query:
“All sales greater than (filter #1) $5000 last year (filter #2)”
“Sales less than (filter #1) $1000 last month (filter #2)”
Next, consider whether or not to apply a groupable to your query. In this case, the clue is in the name: groupables are things you can group data by! Groupables might include things like customer names, classes, service names, product names, and timeframes.
To apply a groupable to your data, it’s helpful to use the word “by” in your query. Similar to filters, multiple groupables can be applied to a single query, for example:
“All sales by sales rep (groupable #1) by month (groupable #2)”
A single query can also include both groupables and filters. See example below:
“All sales by sales rep (groupable) by month (groupable) last year (filter)”
Through these examples, you can see that there are infinite ways to slice and dice your data with Chata. There’s also a lot of flexibility in the way you can frame or ask your questions. If you want “Total sales by month”, you can ask your query in exactly that way. Chata will also return the same information if you ask “Monthly total sales” or “Sales by month.”
If you can think of the type of answer you’re looking for and include relevant filters and groupables to ensure the data that gets returned meets your specific needs, you’ll be able to find answers faster and uncover insights more easily than ever before.