When Complex Business Questions Lead To Simpler Queries

What questions to ask and how to easily find the answer


As a data analyst, you’re often caught in a bind. You want to ask the right business questions that will lead to meaningful insights and growth. But it seems as though the more insightful and powerful your questions are, the more complex they (and your SQL) become. To save yourself from creating endless code lines, you’ll need to use an agile query language that gives you the ability to analyze user behavior.

So what are the most important questions you need to be able to answer? How do you create the right query that asks that business question?

In this post, we’ll show you the types of queries you’ll need to ask based on your most critical business questions, as well as how to get the answers to your queries by breaking them down into small steps.

What Series of Events Lead You to the Desired Action?

It’s rare that we wake up one day having magically achieved our goals. Generally, you need to chart out a path that breaks your goals into smaller steps. That’s why in your online business it’s helpful to be able to analyze which series of events led to a specific action.

Maybe your new business objective is to encourage more free players to become paid players. That means you have to analyze your data to discover the common path users take towards conversion. Such path analysis could reveal, for example, that most players collect the free bonus as the second action as well as show that even though they pressed 'Buy' to purchase more coins, they did not complete the purchase.

But what if you want to only analyze users who pressed 'Buy', or only to look at users who completed a certain series of actions within the same day? Or within the same session? That’s when your query might get too complicated.

Once you’ve adopted an agile query language that is built for such path analysis, you’ll receive deep insights quickly, with a simple query of less than 10 lines of code!

Which Online Campaigns Generate the Highest ROI?

On the other hand, let’s say you’d like to optimize your marketing spend on online campaigns, and focus on the most effective campaigns.

To understand campaigns' ROI, you need to analyze the performance of the different online campaigns, Google Adwords campaigns, Facebook campaigns, or email campaigns to understand which campaign or which channel was most effective in bringing in new installs or any other conversions. This can be measured by analyzing users’ LTV and cross-referencing it with the customer acquisition cost (CAC). Once you have a winner you can focus on that type of campaigns or channel and lose the less effective.

This ability to wholly query and analyze raw data from different sources is one of the greatest advantages with advanced digital intelligence solutions.

Are My Users Staying or Going?

Retention. That’s what it seems to all come down to, doesn’t it? The more the users engage with your site or app, the more likely they are to convert. But you’ll need to constantly keep your eye on any drops in the retention rate. Cohort analysis is commonly used to look at retention and the opposite, Churn. Cohort analysis shows clearly the segments of users that stick to your product.

Or maybe you just want to analyze backwards, from the conversion event. Reverse cohort analysis is used to analyze backwards from the point of a certain event to gain a better understanding of what actions preceded and why they behaved the way they did.

For example, we can analyze which users purchased an item after watching the product video. From here we can deduce what types of product videos are more successful at conversion.

Cohort analysis can get quite cumbersome when using traditional querying. When using a query language that is built to analyze user segments over time, can facilitate this needful analysis.

Dream Big by Taking Small Steps to Achieve Your Goal

The more ambitious your goals are, the more likely you are to get stuck. That’s why it’s so important to break down your goals into smaller, achievable steps. By adopting an agile query language, built for time-series analysis, you’ll be able to simplify analysis of raw data and have greater understanding of how your users behave.


Guy Greenberg is the Co-Founder & President at CoolaData, a leading behavioral analytics platform. Before founding Cooladata, he was the co-founder and CEO of Gilon Business Insight, which was acquired by Ness Technologies in 2010. At Ness Technologies, Guy served as Senior Vice President for Global BI and Big Data, where he worked with some of the largest corporations in the world. With over 20 years of experience in big data and startups, he is an active angel investor and adviser of several Big Data startups.

Vision small

Read next:

Big Data Forecasting In Pharma