commentBlock = $("#comment_post_32592").find('.comment-blocks'); commentBlock.find('.loading').hide(); commentBlock.find('.see-all-comments').hide(); commentBlock.html("
<\/a>
\"it_user195735<\/a>
it_user195735<\/a>System Analyst , Business Intelligence at a tech services company<\/span><\/div>
<\/span>Consultant<\/span><\/div>
<\/i><\/div>
<\/i>Report as inappropriate<\/a><\/div><\/div><\/div>

I have been working on IBM Cognos for over 8 years from the earlier versions to the most latest.This product has evolved a lot over time and has always been in pace with the business needs and data reporting requirements which have changed tremendously over the years.IBM Cognos is far more mature product than most of the competitors , but keep in minds its expensive.The product is now moving more towards self serve analytics and predictive analytics.\n<\/p><\/div>

<\/i>Like<\/span>(0<\/span>)<\/a><\/i>Reply<\/span><\/a><\/div>
<\/div><\/div>
<\/a>
\"it_user164874<\/a>
it_user164874<\/a>Practice Leader<\/span><\/div>
<\/span>Vendor<\/span><\/div>
<\/i><\/div>
<\/i>Report as inappropriate<\/a><\/div><\/div><\/div>

I have to dispute some of the comments above. I believe the above author\'s comments are outdated.<\/p>\n\n

Firstly, IBM Cognos has a great ROLAP offering in the form of Dynamic Cubes. <\/p>\n\n

Cognos Series 7 was a set of products most of which almost all were built in house. Cognos Query was acquired. PowerPlay, Impromptu, Upfront, Architect, Transformer, Visualizer were all built in house. It is fair comment that there were differences as PowerPlay exclusively used muti-dimensional sources, Impromptu exclusively accessed relational sources, and did an OK, though not outstanding job, on star/snowflake schemas.<\/p>\n\n

Cognos 8/10 was built entirely in house, and all components share the same query engine. It has none of the legacy of Series 7.\n
Cognos TM1 was acquired from Applix back in 2007, <\/p>\n\n

As I posted earlier, I am an ex-Cognos product manager.<\/p><\/div>

<\/i>Like<\/span>(0<\/span>)<\/a><\/i>Reply<\/span><\/a><\/div>
<\/div><\/div>
<\/a>
\"it_user175269<\/a>
it_user175269<\/a>Owner at a tech services company<\/span><\/div>
<\/span>Consultant<\/span><\/div>
<\/i><\/div>
<\/i>Report as inappropriate<\/a><\/div><\/div><\/div>

The shortcomings you describe likely are the results of 1) COGNOS\' MOLAP architecture and 2) no single metadata. <\/p>\n\n

MOLAP is fine for fixed or slowing changing reporting, where what you are looking at is well known and you have plenty of time to respond to new needs. It is less than ideal for analysis because by its very nature, the analyst doesn\'t necessarily know everything they are going to question, thus the need for analysis. To get over this hurdle, you should look at a tool with a ROLAP architecture. <\/p>\n\n

COGNOS has been built over the years through acquisition and \"integration\" of disparate products and has always had a problem with trying to integrate the metadata from each. This lack of complete integration not only contributes to errors in the data across functions, but also contributes to stability problems.<\/p>\n\n

An additional problem, from an analyst\'s point of view, is that as your MOLAP-based system becomes larger it becomes ever more difficult to manage the number of cubes and decide which cubes may be needed for analysis. And then, if you don\'t have a/any cube(s) to answer the analyst\'s questions, a new cube needs to be built, which negates the speed advantage of using a cube. This is much less of a problem with a ROLAP system because the developer is not pre-calculating answers, but is waiting until run-time. While this can cause a major performance hit, ways have been developed to overcome that.<\/p>\n\n

If you have not followed the BI industry closely but have only followed developments with COGNOS, get hold of a recent copy of Gartner\'s Magic Quadrant for Business Intelligence and Analytics Platforms, a Business Application Research Center (BARC) BI Survey, or a BI Scorecard. \n
You can usually get these as marketing literature from the various BI players.\n
Disclaimer: I\'m a MicroStrategy-certified BI consultant.<\/p><\/div>

<\/i>Like<\/span>(0<\/span>)<\/a><\/i>Reply<\/span><\/a><\/div>
<\/div><\/div>
<\/a>
\"it_user164874<\/a>
it_user164874<\/a>Practice Leader<\/span><\/div>
<\/span>Vendor<\/span><\/div>
<\/i><\/div>
<\/i>Report as inappropriate<\/a><\/div><\/div><\/div>

In my experience performance is a function of design. Well designed solutions perform well, poorly designed solutions don\'t.<\/p>\n\n

For example a poorly designed data structure mapped to a DMR model in Framework Manager may not perform well. A well designed PowerCube (molap) can perform very well, especially if you understand what hierarchies will be challenging and address these through design. Well designed Dynamic Cubes can handle large data volumes and perform very well.<\/p>\n\n

If you want maximum performance for your data sources, with scalability, then Dynamic Cubes are the way to go. Even better if you put them on top of a BLU database, and better still if you use the columnar database options.<\/p>\n\n

I am sorry to hear you can\'t find a skilled IBM Cognos partner in your part of the world to help you. There are many good ones, and some not so good. This is true of any vendor\'s partners.<\/p>\n\n

Good luck with your investigation of Tableau. Just remember you are comparing apples and oranges since they don\'t do the same thing. IBM Cognos has many capabilities and is pretty good at them all. Tableau is great at one thing only.<\/p>\n\n

The closest comparison would be authoring dashboards in Cognos Workspace Advanced with RAVE visualizations, and delivering them using Workspace, all on the web, vs doing this client server with Tableau. <\/p>\n\n

(I am an ex-Cognos product manager, and currently work with an IBM Business Partner - about 17 years of Cognos BI experience)<\/p><\/div>