What are the advantages of using online business intelligence vs. traditional BI software?

If you are speaking of 3rd party BI web services compared to inhouse implementations i.e. Business Objects, Hyperion, MS SQL with Rerporting Services - I believe there are differences.

The greatest difference comes from the ability of inhouse support to analyze and construct a Knowledge Exploration/Decision Support facility (ETL, Store-Retrieve-Manipulate tools, User Workbench, Plug-n-Play KPIs,etc.) Even when this is supported by substantial platforms like BO or Cognos the faciliative expertise required from IT via Business Analysts tends to be a steep climb. Buying in expert consultants helps considerably but there is still an insidious aspect to the effort.

Most BI value comes from activities knowledge subdomains, either single departments or at specific interfaces. A clue here is that most BI projects lead by outside concerns are "incremental", focusing and building on one knowledge area first. This means that encompassing projects tend to take longer and portend more risk.

Given this, very high ROIs can be attained by allowing subdomain funtions to acquire outside web based services to create their own BI facility for analysis, decision making and reporting. These solutions tend have much shorter "Time to Benefit" cycles and can provide a solid basis for future inhouse "whole organization" efforts i.e. you get internal benchmarking and "best of (in)breed" in the process.

Part of the reason this works is that subdomain managers retain higher levels of ownership and responsibility - not to mention "credit" for the results, a motivating factor.

If the BI you have in mind includes considerable unstructured information objects then a web interface is the way to go. Ther main reason being the plethora of integratible tools accommodating a wide range of eMedia. This still begs important questions about the depth of understanding of the knowledge domain itself and the utility of such eAssests and associated retrival tools.

These robust eMedia strategies involve lots of effort in creating and maintaing taxonomis and indexing services. Even with "free" utilities costs of construction and maintenance should be included in the ROI before proceeding.

There are many advantages of SaaS or online BI over traditional BI, here are a few-

1) Small and medium-sized businesses often cannot justify the large CapEx of traditional BI, do not have the IT resources to run an on premise database and server system and are not sure that they have the breadth or depth of data to warrant such a system. Often their needs change and evolve far faster than software can be developed and installed.

2) Traditional BI demands technical expertise both to manage and use. Business users could not quickly perform analysis themselves; queries had to be made through the IT department and often involved long delays. As there is no software to install, there is no daily maintenance or routine tasks to be done, so nothing to require extra IT hands or expertise. This frees IT professionals to concentrate on strategic IT and growing the business.

3) With traditional BI having enough data to justify the BI spend is certainly a problem. The huge upfront costs and time investment often cannot be justified by the need for a couple of key decision makers to analyse a small number of records. SaaS BI gets rid of this problem as it provides fully functional BI suite available for just one person. SaaS applications can be rapidly scaled up or down as your business needs change.

4) Most SaaS products are lightweight and designed to complement the investment in data storage and manipulation that you already have. For example, Bime, a SaaS BI solution, plugs directly in to online and traditional sources to extract data, then allows creation of visualizations and dashboards with a few clicks.

5) Access and sharing were the original drivers of the SaaS cloud computing model so applications make it easy to get useful visualizations to the right people quickly by inviting people to your dashboard's URL or embedding it in a website or blog.
BI products have traditionally required huge upfront costs, had long lead times and
demanded technical expertise to use. SaaS allows powerful functionality without the need for any of these, making it a highly profitable investment for most organizations.

Hi Cody,

I'm a little late to the party, and you've already gotten some great answers from Kristy and Hrvoje about how SaaS is more affordable, less IT resource intensive, and easier for business users to use.

So I'll talk about a slightly different perspective, more along the lines of Hrvoje's comment about knowing what you need.

Here's something that traditional BI does well: if you have the time, money, and resources, it can create an ultra-customized, highly scalable solution that can answer an astonishing number of questions across multiple data sources. There's a reason that people pay millions and wait years for a high end Oracle OBIEE, SAP/BO, or Cognos implementation. When it works and you have the big budget and fully staffed IT team to run and manage it, it can be a beautiful thing.

Most of us don't have that kind of time, money, and resources to burn, however. That's where the bad stories about expensive, failed implementations, difficulty scaling/extending the solution, and complexity come from.

SaaS BI addresses the cost, usability, and complexity issues, but you have to know what you need, because not all SaaS vendors are the same. Some offer specialized applications, like sales analytics only, or can only go against a few tables or one data source. Others offer visualization and reporting, but don't have a data warehouse and ETL engine. Others offer a comprehensive solution, including ETL, data warehouse, analytics engine, and reporting.

If you'd like to know more about SaaS BI vs. traditional BI, check out this BI 101 article http://www.birst.com/bi101/index.shtml

And in the interest of full disclosure, I work for Birst, a SaaS BI company (www.birst.com).

- Barbara