Google logo on a wall
PHOTO: C.E. Kent

If you’re not familiar with the term "Insight Platforms-As-A-Service," fear not, it’s new.

Forrester, the term's creator, thinks Google is the lone leader in the space.

In its Forrester Wave: Insight Platforms-As-A-Service, Q3 2017 report (fee charged), analyst Brian Hopkins and colleagues Srividya Sridharan, Dave Bartoletti and Sara Sjoblom, defined it as "an integrated set of data management, analytics, and insight application development and management components, offered as a platform the enterprise does not own or control."

Forrester included eight vendors in its initial Wave: 1010data, Amazon Web Services (AWS), Databricks, GoodData, Google, IBM, Microsoft and Qubole.

Debating About Who Makes the Cut

Insight Platform-As-A-Service (PaaS) provides data scientists, data analysts, business analysts and others with everything they need to glean insights from data except the data and questions themselves.

When Hopkins, a vice president and principal analyst at Forrester, created this Wave, he wondered aloud who should be included. Google, Amazon, Microsoft, IBM and Oracle (which eventually did not make the cut) were no-brainers, he decided.

But what about Adobe Marketing Cloud, business services providers like FICO, cloud native BI vendors like GoodData and BIME, analytics vendors like TIBCO and Tableau or big data platform vendors like Pivotal and MapR?

Defining the Core Nature

To qualify as an Insight PaaS provider, vendors had to offer:

  • A multi-tenant PaaS.
  • Multiple kinds of analytics services, more than two types of data management services and insight execution services that help deliver analytic insights to other applications and processes. 
  • Significant integrations between their data management and analytics services.
  • Several enterprise customers that have built custom data analytics solutions on top of the vendor’s Insight PaaS offering that are in production.

Forrester judged its eight vendors based on its "current offering" (data management, analytics, insight applications and platform) and its "strategy" (product vision, execution roadmap, performance and planned enhancement.)

Google Stands Alone as a Leader

Google was the only vendor that earned a coveted Leader title. 

The analysts noted its "comprehensive set of analytics and machine learning tools, designed to fit the needs of every business." These include Google's App Engine and BigQuery PaaS components, its cloud-based Hadoop and Spark services and its “insight execution” features such as machine learning automation.

Google's Insight PaaS perceived flaws aren’t surprising: Forrester analysts reported its offering is too complex, and called its billing process “arcane.” 

However, it's worth noting here that Google has not had a serious enterprise business for very long, and as it seriously pursues customers, it will probably come to better understand what companies need.

IBM, GoodData, Databricks Strong Performers

There are plenty of big surprises here. First and foremost, perhaps, that much-acclaimed IBM Watson did not qualify as a Leader. To be fair, a Strong Performer grade indicates that a vendor has great capabilities to offer, but many expect more of Watson, given the hype. 

Forrester Wave authors reiterated some common customer complaints such as Watson Analytics still requiring manual data loading and that its Spark-based machine learning was not yet generally available.

Databricks’ selection as a Strong Performer is interesting for an altogether different reason: it's a startup among giants in this space. The analysts were impressed by its “polished user interface” and they called it “ideal for teams of citizen data scientists.” 

Its shortcomings should not be all that surprising to anyone who watches the space. Forrester said that business analysts might find coding in notebooks “challenging” and that its approach to real time was “limited to Spark.” (Databricks was founded to bring Apache Spark to enterprises.)

Mastering a New and Complex Field

GoodData was recommended for consideration by Forrester to companies that want to implement “lightweight PaaS” solutions for “business analytics systems of insight in the cloud.” 

Its limitations, according to the analysts, are that its analytics aren't real-time (there is at least a 15-minute lag) and that it lacks robust data science tools, among other concerns.