Adam Gartenberg's Blog

Business Analytics and Optimization, IBM and Social Marketing

Breaking: new version of IBM Mashup Center and new IBM Cognos 8 Mashup Service

I'm in the "nachos and news" program being put on for those of us participating in the inaugural IOD blogging program, and Anant Jhingran, CTO of Information Management, just shared exclusive news on the new version of IBM Mashup Center and unveiled the new IBM Cognos 8 Mashup Service.  The press release won't cross the wire until tomorrow morning (I'll link back to it when it does) but you can find the details below and on a few of the other blogs, like Todd "Turbo" Watson, Vincent McBurney, and ReadWriteWeb (Check back in for additional links).

I've written about mashups before. For those of you that might not have prior exposure to them, Anant described them as enabling the last mile of getting data to users by enabling end user to mix and match information themselves.  

There are three key pieces to what's new with v2.0 of IBM Mashup Center:
1.  Mashups only as good as data that have access to.  Tremendously expanding data sources can natively connect to.  
2.  Only as good as interaction paradigm - expanding interaction and ability to create widgets.  
3.  Customers are saying this is great, but I own data, and will I let some random LOB guy come in and access my data?  Am I really going to open it up?  So adding controls.

The new IBM Cognos 8 Mashup Services exposes Cognos content so that it can be surfaced in mashups (IBM's or others) or other places/processes (while still ensuring that the appropriate controls on access, masking data, blocking IPs, etc. are in place.)

The result is to greatly extend the reach of Cognos.  One of the key themes of the conference is on the importance of deliver information and insight to the point of impact, and combining Cognos reports with mashups allows for the information to be presented in the format most relevant to end user who needs to access that data.

Anant stressed that this was not just vision - mashups are actually being deployed by many of the clients.  (I'll have details on one such client below.)

Jennifer Hanniman gave a demo of how use BI content in mashup to address business situation. For example, a sales team is being asked to increase their sales goals for the third quarter by 10%.  They can look and see that sales were growing in Q1 and that there's a better picture in Q2.  But if they're going to sign up for that increase, they'll want to look at more than just past trends.  They need to understand where opportunities will come from and who will meet it.  They also need to drill into the specific sales reps in Seattle branch, for example, and can look at opportunities from CRM system, but all the necessary information is not in one place.  They also need to see what conditions are like - market data - for the accounts whose sales they might want to accelerate (e.g., if they just announced a bad quarter, the chances of getting a quick sale might be a lot lower).

By creating a mashup, they can bring in a Cognos report on BI, showing plan vs. actual sales data, broken down sales rep.  Clicking on a sales rep name shows the plan target and how they are doing against that target.  It can also show the top accounts for the selected rep, and bring up Google search information for each of those companies.  The result is that they have the information they need to see whether it makes sense to plan on closing that account earlier in the year.

There are many companies using IBM Mashups today, and the press release highlights the stories of Kent County, UK, Wells Fargo, and AMEC Paragon.  I'll write more about those tomorrow; we were joined at the session by Anton Langeler from O'Neill, the groundbreaking maker of wetsuits and much more, and he told a very passionate and compelling story about the value mashups are bringing to the company.  

The company that literally started out of a surf shack has now grown to 225 product groups, with many distributors and 60,000 retailers,  It's important that O'Neill have a view into the complete supply chain, not only to get data out of supply chain but also to share data back out to distributors and retailers.

They started with Cognos, and built data warehouse, and reporting above it.  They are now using mashups to frame how they collaborate with retailers.  They built a web interface - a widget - in which they can start a return process with a licensee.  The result is they can read data on what was sold by the distributor, using data across the complete supply chain, in addition to using it locally within the company.

The advantage of doing this as a mashup instead of traditional programming is that they need 25 diff. places from which to get the data, and it's hard to get data from all those places, bring together, add some inforamtion.  Now, with XML, can also update all the systems behind it.  In the past, could build application, but were ending up with hundreds of different local applications.  Instead, now have just one source where make changes, and distribute them out through the mashups..  

Anton explained how strong the widgets are.  For example, the stores have a lot of sales data, and they wanted to be able to combine it with weather data, as weather can have a big impact on understanding their sales.  Now, read weather data, combined with retail sales, and bring everything together on the surface.  Traditionally, would have had to bring all the data into the warehouse to do the reporting, blowing up the amount of data in the warehouse.  Now they can can just combine it and materialize it on the fly as needed (with the added benefit of allowing them to be much more spontaneous.)

As another example, Anton cited many processes that involved downloading data from the warehouse, transferring it to Excel, making manual annotations on printouts, and passing the paper around.  It used to take weeks to identify returns, for example, a process that's now done online and pretty much in real time.  

One more benefit O'Neill has found is the savings of over 2 million Euro in modifications that otherwise would have been needed in their underlying data sources.  Because reports can be assembled via mashups, there is a lot less customization needed to the underlying data sources, making upgrades "10 times easier," and also providing a lot more knowledge about the environment to the BI team.

The current version of the IBM Mashup Center is available for demos in the Lotus Greenhouse.