IT Briefcase Exclusive Interview: Web Analytics and Big Data Delivery with Pentaho’s Ian FyfeJuly 18, 2012 No Comments
Ian Fyfe is the VP of big data product marketing at Orlando and San Francisco based Pentaho
There is no question that one of the biggest challenges for Big Data today is properly filtering and putting to use all of the information that has been collected.
Ian Fyfe, VP of big data product marketing at Orlando and San Francisco based Pentaho, outlines below ways in which companies can not only make good use of the information they are collecting, but also capture the full benefit of the data.
Q: What, in your opinion, are the most important factors affecting Big Data Delivery?
A: Everything about big data analytics is challenging, including ingesting and manipulating data, through to visualization and analysis of the data to gain meaningful business insights. The new generation of big data platforms centered around Apache Hadoop and NoSQL databases such as MongoDB, Cassanda and Hbase achieve their massive scalability, high performance and ability to manage unstructured data by rejecting the established relational database model. But as a result, traditional SQL-centric data integration and business analytics simply don’t work, or at best are retro-fitted but without being able to keep pace with the volume, velocity and variety of truly “big” data stores.
To summarize, here are the four required attributes of a successful big data web analytics solution:
1. Removes technical barriers to using the chosen big data platform, whether that be Hadoop or a NoSQL database. This means visual development interfaces for creating and orchestrating data management and analytics tasks.
2. Makes it easy to connect your big data platform to the rest of the enterprise’s data stores. This means a wide variety of connectors for Hadoop, NoSQL databases, as well as traditional relational databases, data exchange formats, and enterprise applications.
3. Lifts the limitations of big data storage and data processing platforms. This means for Hadoop, eliminating the delays inherent in accessing data across large clusters of computers, and for NoSQL databases, this means eliminating the limitations around querying data such as the ability to perform joins, group, and sort data.
4. Provides complete business analytics capabilities – not just one piece of the solution. This means everything from reports, dashboards, interactive visualization and analysis, and predictive analytics.
Q: How do Big Data Web Analytics play a role in this?
A: Big data web analytics is a great example of a business problem that can only be efficiently solved for both cost and data volume reasons using big data platforms such as Hadoop and NoSQL databases. But understanding how visitors interact with web properties, and correlating that with other business activities such as marketing and sales campaigns is absolutely critical to the success of more and more companies. So finding an analytics solution that can work with the enormous volume of semi-structured click-stream data and finding the hidden nuggets of insight contained within is in turn critical.
Q: How will this new partnership between Pentaho, Semphonic, and Infobright help organizations find a way to meaningfully represent digital behaviour at the customer-level?
A: The challenge with organizations’ Web analytics efforts is finding a way to meaningfully represent digital behavior at the customer-level. Together the three companies offer a customizable framework for Web analysis and digital measurement. Semphonic offers state-of-the-art techniques for tracking customers across multiple identity keys to understand, classify and measure the success of individual touch points, and then stream the data using a consistent customer-level model. Pentaho’s business analytics platform provides interactive visualization, dashboards, and data exploration capabilities for the massive amounts of web data stored in Infobright’s high performance database. CMO’s, product managers, or marketing analysts can then rapidly answer questions regarding Website usage, customer behavior, marketing effectiveness and site performance. Combining the partners’ technology and expertise, companies can quickly uncover meaningful insights from big data Web analytics.
Ian Fyfe is the VP of big data product marketing at Orlando and San Francisco based Pentaho. He is a key member of the big data team at Pentaho. He brings over 20 years of experience in the business analytics software market with roles spanning consulting, product management and product marketing. Ian started his career by co-founding a business intelligence startup and has worked at Business Objects, Informix, Epiphany, Pivotal, PeopleSoft and Jaspersoft. His favorite hobbies are skiing, rock climbing, mountain climbing and kiteboarding.DATA and ANALYTICS , Featured Articles, Featured Blogs, Fresh Ink, Inside the Briefcase