Clickstream Analytics

Currently, there are several tools like Google Analytics to analyze data. While it is easy to track page visitors, unique visits, devices and browser using Google Analytics data, it gets difficult if you have to mash it up with other web log data sources for analysis.

It also gets challenging for analysts to get quick insights mashing up structured with unstructured data when using visualization tools such as Power BI, Tableau, QlkView, Cognos and Excel.

The reason for deep clickstream analysis is to effectively determine all aspects of customer behavior across multitude of dimensions and intelligently drive the demand and supply cycles. Having a customer, demographics, category, product, time, region, device and experience would give immense value to the decision making process of any online business.

Clickstream analysis is being implemented by our customers to do:

  • A/B testing to optimize visitor, campaigns, out-bound email, etc.
  • Monitor in-page and intra-page analytics to track usability issues
  • Determine consumer segments and their click paths within the site
  • Abandon last-click marketing attribution models to measure spend.
  • Analyze click data with other dimensional analysis on  brands, demographics, geo locations, pricing, etc
  • Fidelity analysis of and to increase consumer stickiness on site.
clickstream-solution

Clickstream Services

LogicMatter provides framework and devOps team to optimize and support engineering and business teams to help in the following:

  • Cross-Domain tracking optimization
  • A/B testing planning and execution
  • In-page analytics strategy and analysis
  • Tag and Category optimization and  issues
  • Dashboards and KPI definition
  • Marketing campaign and ROI measurements
  • Digital analytics offshore devOps staff augmentation
  • Collate, filter, cleanse, convert, prepare and organize data for analysis
  • Support gathering of data feeds in various protocols (REST, JSON, WebServices, SMTP, FTP, HTTP), from many data sources
  • Data processing & pipeline automation
  • Manage, Design & Implement operational data store (ODS) and analytical data store (ADS)
  • Support Dashboard and Report authoring in  Tableau, Power BI, Excel, Cognos, QlkView and D3  visualization platforms
clickstream-ods

Hadoop Solution

One type of distributed solution we support uses Hadoop components to assemble a data framework for processing incoming web logs / hits / clicks / video usage data in a continuous streaming manner. The architecture style adopted is that of a large data pipeline, assembled using several pre-configured components from the Hadoop open source.

  • The orchestration of automation needs to gather, prepare & organize information for analysis.
  • Several parts of the Apache Hadoop are configured to seamlessly work together to perform the activity of data processing.
  • The ODS uses HBASE & HIVE and ADS uses HIVE to store data. The figure below shows the high level system architecture.
  • The system provides horizontal scale — which means adding additional nodes as data volume or velocity increases without the need for re-engineering the solution.
clickstream-hadoop

Types of Analytics

Which categories are interesting to consumers?

Drill down into products within categories?

Drill down the various click paths of visitors?

Where are the consumers coming from?

What are the number of visitors and their interest level?

How many visitors are IE vs Chrome users?

How many are mobile vs tablet visitors?