The retail industry has seen some major changes in the last few years. The pandemic has not just forced every retailer to think about their online presence, but also to take a step further and enable the systems to be measurable & help in business decisions. With the rapid changes in the ecosystem, business stakeholders were ready to invest more in eCommerce solutions that could enable the retailer to be omnipresent. But the big question was how to measure the success of these investments?
The default KPIs included tracking the user behavior, measuring usability, and fixing the gaps to increase conversion rate. So analytics and customer data became the guiding points.
Google provides seamless integration of tools such as Google Analytics (GA), BigQuery (BQ), and Looker that help in ingesting, assimilating, and presenting the necessary data in the required format to make data-driven decisions.
The data captured by analytics tools like GA always need a reporting & visualization tool to represent the data in a meaningful way. There are many options and the choice depends on the business requirements. A simple requirement to understand the user behavior on the website can be addressed by out-of-the-box dashboards, compared to a need to track & identify the impact of a custom dimension in an interactive dashboard.
The GA, BQ, and Looker documentation would provide complete detail on data views, dashboards, and explorations. We will not get into those details here, but we can review “What to expect” from each platform. It is important to know the limitations of each tool so that it can be used for the actual purpose that it is designed for.
Google Analytics is a simple to use analytics tool that can capture all the user events on a website and provide out-of-the-box dashboards/explorations to visualize the captured data. This is a comprehensive tool to understand the clickstream data behavior. It includes reports for :
- Real-time data
- User/Audience data
- Acquisition data
- Engagement data
- Monetization/Conversion data
- Retention data
It also supports advanced eCommerce tracking that provides insights into users’ shopping behavior.
Retail commerce websites might interact with many external tools designed for specific needs like social media marketing, email marketing, CRM, etc. With just the web analytics data we would be limiting the information required to make data-driven decisions. This would need a system to ingest data from multiple sources and provide a consolidated view.
For advanced analytics, the data from GA can be exported to BQ and enhanced by merging it with data from other sources like social & email marketing, offline/pos data, other marketing vendor data, etc. This data provides a consolidated view of customer data that helps in analyzing customer behavior.
Another major advantage of having the analytics data, is the out-of-the-box support for machine learning models using BigQuery ML (BQML). This provides easy access to write ML models using SQL.
The enriched data on BQ can be visualized using Looker. It provides a customer-centric view with out-of-the-box analytic blocks that implement best practices for data analysis. The reports/dashboards can be easily downloaded and/or directly shared with anyone without access to the system.
Data-Driven Business Cases
Although brick-and-mortar has carried the majority of traffic for a very long time, eCommerce has seen significant growth in the last decade or so. The last two years of the pandemic have forced offline store traffic to go online. This change in behavior along with the general trend for retail to be omnichannel has created new expectations. It has forced retailers to think of it as channel-less, decoupling the dependency on a specific channel, and responding in a comprehensive way using data.
Data analytics plays a crucial role in analyzing user behavior on any given platform. So the emphasis on analytics increased and it transformed to be the core of all data-driven decisions taken by business stakeholders. All these changes in a very short period demand a few important updates to the retail operation. It brought focus on the below-listed business cases that have a dependency on collecting the right data points to provide a feedback loop and make knowledgeable decisions based on data:
- Features like Buy Online Pick up In-Store (BOPIS), Store locator, Store Inventory became core/necessary to operate a retail business
- Importance of Omnichannel support as customers jumping between different touch points like web, mobile & tablet became more common
- Need for a consolidated customer view using Customer data platform utilizing data warehouse on the cloud for easy migration & scaling to satisfy the growing demand
- Advanced analytics dashboards that can capture comprehensive user behavior
- Applications built on AI/ML models that can help in predicting user behaviors like Recommendations, Personalization, Demand forecasting, etc with more accuracy
Analytics provides the required fundamental unit to enable all enterprise applications to capture the required data.
In this series of blogs, we will dive deeper into one of the above-listed business cases to demonstrate how we can use the Google cloud services effectively.
Current state of maturity
Google’s ecosystem has realized the necessity of data-driven decisions and simplified the tool integrations so that it takes out the complex overhead of data transfer between systems. In its current state, linking GA data to BQ or other sources like Google Ads is very simple and only a few clicks away.
Let’s review the maturity of the following tools in the Google ecosystem:
The 4th iteration of Google analytics has a major change in the fundamental architecture. It has considered the user behavior changes that have brought in major emphasis on omnichannel and made a single view for tracking users on Web and Mobile Apps.
It has also simplified the type of data collected and made everything an “event”. It also supports complex ML-based attribution models, data drill down using explorations, simplified data import process, and quick product linking to name a few.
As a fully managed cloud data warehouse service, BigQuery goes through frequent updates to keep up with the ever-changing demand. It provides built-in support for Machine Learning using ‘BQML’ that has proven to be a gamechanger for building ML models using simple SQL statements.
The recent updates on the row-level and column-level security features show how quickly Google can adapt and make the necessary data security changes. Another major update is ‘BigQuery Omni’ that uses ‘Anthos’ to support multi-cloud analytics by accessing data on AWS and Azure cloud.
BQ also provides a BI engine SQL interface to interact with Looker, Tableau, PowerBI, etc. This makes it easy to visualize the data on BQ using any popular or custom BI tools.
Looker is a BI platform that can sit on top of data sources like Oracle, MySQL, or BigQuery to provide the necessary drill-down view and visualization. It has a built-in integration required to communicate with the BQ engine. This tight integration provides a quick linking option between Looker and BQ. It also enables Google cloud’s security tools like Cloud DLP (Data Loss Protection), supports integration with BQML, and customize/build data models using ‘Looker blocks’ accessing data from BQ.
Data-Driven Decisions stay at the core of the last mile for retail commerce operations. Multiple surveys reveal that organizations have been very poor in utilizing the data to derive proper actions to achieve business goals. With the maturity in data analytics, ML, and visualization, deriving the data in the required format for business decisions has been easier and more accessible. So it has caught the attention of major retailers and is becoming one of the main areas of focus.
This is the first in the series of the three blogs where we discussed the need for data-driven decisions for retailers, the business cases scenarios, and the current state of maturity of the tools that can be used to derive the data.
In the next section, we will dive deeper into a business case and see how we can use Google’s advanced analytics ecosystem to solve the business problem.