Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

BLOG: Mobile BI and Cloud BI evaluation criteria

Boris Evelson | March 12, 2013
Get specific with your suppliers or you will face problems


Cloud BI

  • Additional licenses. In addition to the subscription fees, does the cloud BI platform require any additional licenses/subscription fees for apps to run and/or to connect to data sources?
  • Architecture/mode. Is the BI cloud platform hosted, via a managed service provider, or multitenant SaaS? If SaaS, is it multitenant or virtual encapsulation?
  • Automodeling. Does the product autosuggest a best data model (star schema) based on input data types and content?
  • Cloud type. Public, private, or hybrid? If public, can clients connect via VPN? Is the hosting platform SAS 70 Type II certified?
  • Control over upgrades/maintenance. What kind of control and flexibility does the vendor allow you to have over upgrades and maintenance? This is usually different for hosted versus multitenant vs virtually encapsulated platforms.
  • Data model customisation. What is the level of data model customisation extent and flexibility?
  • Data persistence. Does the product store/persist data in an analytical cloud based data store or simply pass data through from data sources to BI applications?
  • DW. Is data warehouse functionality included in the cloud BI offering?
  • Elasticity. What is the level of computing resources, storage, and pricing elasticity? Can resources be added or reduced (including changing payment plans) instantaneously, monthly, quarterly, yearly? Who provides the elasticity capabilities, the BI vendor or the cloud infrastructure provider?
  • ETL. What is the sophistication of the cloud BI product data ingestion / ETL process?
  • Free, starter solution. Does the vendor offer a free starter try-before-you-buy option? If yes, what are the limitations?
  • Functionality. Does the product support pixel-perfect complex report writing, ad hoc querying, OLAP or OLAP-like analysis, data visualisations including dashboards, advanced analytics (predictive modeling, data mining, statistical analysis)?
  • Native multitenant cloud SaaS architecture. Was the product natively architected as multitenant cloud SaaS? Or was it converted from an on-premises single-tenant BI platform? If the latter, can the product be deployed in a multitenant environment with zero modifications?
  • Offline. What is the product's disconnected/offline capability? Will it automatically cache most recent reports?
  • On-premises. Does the product also offer an on-premises version of your software? If yes, is it exactly the same or a different product? Does the product offer a clear migration path from cloud to on-premises?
  • PaaS. Does the product offer vertical- or domain-specific solutions only, or does it offer a full cloud BI SaaS development environment, AKA platform-as-a-service (PaaS)?
  • Public cloud platforms. Does the product support popular public cloud platforms, such as Amazon, Google, Salesforce, and Windows Azure?
  • Self-service. What percentage of the product functionality can be instantaneously procured, implemented, deployed and used by end users with no IT or vendor involvement?

 

Previous Page  1  2 

Sign up for CIO Asia eNewsletters.