I recently joined Couchbase’s product team to lead Couchbase Analytics. This Analytics Service enables our customers to measure near-real time business operations, derive insights from data, and drive agile decisions to expand business growth.
Within my first few weeks at Couchbase, I had the opportunity to meet with a number of customers to learn about the “why” and the “how” behind their usage of Couchbase Analytics. This exercise helped my team think backwards from the customer point of view and assess which features customers care about most. More importantly, we learned how we could further take action on their feedback. In this blog, I’ll discuss:
- How Couchbase Analytics works
- Key business pain points we address
- Top customer use cases
Couchbase Server is an open source, distributed, NoSQL document-oriented database system designed to provide low-latency data management for large-scale interactive web, mobile, and IoT applications.
The Couchbase Analytics Service helps users to analyze their application data. Users do this by creating shadow copies (or datasets) of the data they would like to further analyze. When shadow datasets are created they are connected to the Data Service and any changes in the operational data are reflected in the Analytics Service in near-real time. Business users can then query the Analytics data using N1QL for Analytics without slowing down the operational data or query services. See Figure 1 below.
The Analytics Service is especially valuable when you don’t know every aspect of the query in advance – for example, if the data access patterns change frequently, or if you want to avoid creating an index for each data access pattern, or if you want to run ad hoc queries for data exploration or visualization. You can learn more about when to use the Query Service versus the Analytics Service here.
Addressing pain points and driving insights
To better understand our customers’ product usage journeys, we wanted to learn more about the business pain points that led them to use Couchbase Analytics. These are the top three pain points:
- Operational workloads being heavily impacted by analytical queries on operational data nodes. (Examples include placing orders, device and user profile management when processing analytical tasks to run operational reports, and dashboards like Customer 360.)
- A need to perform data science experiments in near-real time and derive faster consumer insights without having to first extract, transform, and load (ETL) data into an enterprise data warehouse where schema is less flexible.
- Inability to efficiently run complex analytical ad hoc queries using the Query Service, and as a result query performance is less desirable. (Examples of such queries include revenue impact for health care companies due to COVID-19.)
Customer use cases
We asked our customers how they use Couchbase Analytics to impact their business and what data analysis questions they ask to enable agile decisions. Here are some specific use cases that are representative of our customers as a whole:
Customer loyalty and promotion offers
This customer is one of the largest retailers in Europe with over 1,000 grocery stores. Their operational data streams from cash registers, web applications, scanners, and more than 1,500 point-of-sale systems with over 2 million cash receipts per year. Couchbase Analytics enabled them to isolate analytical workloads from operational workloads, provide fast response times, and streamline analyses of receipts, products, prices, and other data for over 18 million customers. Key questions answered: Based on consumer buying patterns, which customers and which products should I target in order to provide optimized promotional offers that will increase customer loyalty?
Customer segmentation and agile data science experiments
Another customer is a world leader in meal delivery with 16 thousand stores that delivers 3 million food items across 85 countries every day. They use Couchbase Analytics to target consumer offers in near-real time every few hours instead of taking weeks or months like they used to. They removed the need to ETL to the data warehouse and reduced time to perform data science experiments, thus enabling agile data mining models like propensity scoring based on order behaviors. Key questions answered: What customer promotions to offer based on the average transaction size, annual purchase frequency, and customer lifetime value driven by customer segmentation?
Healthcare revenue management
This last customer is an industry leader in revenue integration solutions for healthcare organizations such as hospitals and clinics. When they re-architected their entire data platform, they used Couchbase to meet increased data growth and improve scalability. They also shifted to Couchbase Analytics, allowing them to use the powerful N1QL for Analytics (SQL++) query language, partition data by time aggregations, and optimize system load to deliver predictable performance. Key questions answered: What are the effects of COVID-19 on business? What trends in visits, discharges, cancellations and procedures can help minimize the impact of missed revenue?
These past couple of months have been extremely motivating and exciting for me as I’ve heard from many customers about their winning moments of truth. Customer feedback is the fuel that improves our products, validates our product investments, and shapes our product roadmaps to support our customers’ strategies and successes. More importantly, customer voices help us cultivate a stronger relationship and a deeper understanding of their needs. Customer satisfaction should be a primary objective of any business, and I look forward to working with our customers to help them analyze data and achieve their goals using Couchbase Analytics.
To wrap up, below are the key benefits and differentiators of the Analytics Service:
Thanks to our product, engineering and marketing management teams for their valuable contributions to this post.