Connect – a Simba Technologies Blog

Strata September Playlists

 So I had a few moments yesterday to pull together playlists for the recent Strata event. For whatever reason, Oreilly has not posted much material on the actual Strata Conference (Sept 22 & 23). But the JumpStart (Sept 19) and Summit (20 & 21) are nearly completely online. These are sessions I attended that I felt are eye-opening, new ideas or just well-articulated points of view. Let me know what you think.

Strata JumpStart playlist

Strata Summit playlist

Strata Conference playlist

 

Strata 2011

Some of you may know I've been working on a few ideas in the last while. Of late, I've been talking to some of you and attending conferences/webinars. I'm in New York for O'Reilly's Strata this week. Apart from a couple of misdirections, the day has turned out rather well. Day 1 was the so-called Jumpstart day before the Summit and Conference. By that, the organizer meant for day 1 to provide a preamble to the topic at hand: Big Data. I had rather low expectations for day 1 mainly because I couldn't see how much any one speaker could fit into 35 minutes. After all, we managed to fit 14 sessions, two coffee breaks and one lunch into one nine hour day.

But, like a good hors d'oeurve, the day has whetted my appetize for what's to come in the remainder of the week. The wide ranging topics spanned people (e.g., recruiting data scientist and Google's Project Oxygen), methodology (Simon Wardley's "disturbance in the flow", Jennifer Zeszut's agile marketing), process (Michael Hugos on Starbucks' supply chain) and policy (Michael Nelson on organizational transparency). What stood out was that nearly all speakers were presenting their hypothesis with sound data and/or analysis. (This really shouldn't surprise since the conference is about using and working with data.)

It should be a good week.

 

Excel BI Survey 2011

How Do Excel and Business Intelligence Integrate in Your Organization?

Simba invites you to participate in a short industry survey on the use of Microsoft Excel in Business Intelligence (BI).  Microsoft Excel is the most-widely deployed and used tool in the BI arena.  This survey is intended to provide deeper insight and understanding for which Excel is used in combination with other applications and as a stand-alone technology for delivering actionable information.  Many companies have told us that their user communities have embraced Excel as the “tool of choice.” We would like to learn how pervasive Excel use truly is within Business Intelligence environments.

The survey will take less than 10 minutes to complete.

Results will be shared with participants shortly after the survey closes on May 18th, 2011.  Results will also be used as the foundation of a future report and white paper.  No individual data will be revealed.

Take the Survey Here 

Thank you for your participation!

An Excel PivotTable is not like a regular Excel Spreadsheet Table

Contrary to folk wisdom, if it walks like a duck and quacks like a duck, it’s not always a duck!

Most people use Simba’s MDX Provider for Oracle OLAP to create an Excel PivotTable to summarize and analyze data in an Oracle OLAP Database cube.  However, once the data is in the PivotTable, many users would like to add additional columns or calculations to their analysis.  In a normal spreadsheet, that’s easy, right?  Insert a column, type in the formula you wish to use (e.g. ratio of actual vs. target), and Excel will propagate that formula to every row in your column.

Even though a  PivotTable looks like a regular table in an Excel workbook, it isn’t one.  In reality, it is a single object that is displayed in a worksheet in a familiar format.  As you drill down or expand rows or columns, your PivotTable will grow and shrink as necessary, and any data you have in cells next to your PivotTable on the right or below it will be overwritten as it expands (though at least Excel warns you that you are about to overwrite some cells).   In fact, we recommend that if you must show additional information on your worksheet that is not part of the PivotTable, that you put it above or to the left of your PivotTable so it remains persistent. 

Because your PivotTable is one single object, you can’t simply split it down the middle to insert a new row or column with your own formula.   But, there is good news – there are ways to add a new row or column to your PivotTable with your own formulae.  With PivotTables, it’s all about your data source. 

• In the rare case that the source data for your PivotTable is in one of your other worksheets in your Excel workbook, you can’t insert anything you want, but you can insert a new calculated field.  Just click anywhere in your PivotTable, and select from the ribbon Home>Insert>Insert Calculated Field.
   
• In the more likely case that your data is a connection to an external data source, you need to go to your data source to add the calculation.  In the case of Oracle OLAP, you would use Oracle’s Analytic Workspace Manager (AWM) to add another calculated measure or calculated member to your cube.  After you have “maintained” your cube, the new member or measure shows up in the PivotTable field list, and you can add that as a row or column in your PivotTable.

However, if the above doesn’t apply to you, e.g. you want to add something that is not a calculated field, or you do not have access to the Oracle AWM or the cube in order to create a new calculated member or calculated measure, you can still have your way.   Here’s how:

1. Click anywhere in the PivotTable
2. On the ribbon, select PivotTable Tools > Options > OLAP Tools > Convert To Formulas

Your PivotTable has been turned into a regular table, with each cell containing its own formula to extract data from the database.  Your data will continue to be connected to your data source, and you can insert a row or column wherever you like in your table, with any value or calculation you need.   What you have lost, however, is the ability to change the structure of the PivotTable by adding or removing hierarchies, and you can no longer drill down or up through the various levels – your PivotTable has become static in structure but still refreshable from the OLAP data source.   Perhaps that’s exactly the trade-off you are looking for.

And that is how you get all your ducks in a row (or column) in a PivotTable!

Happy Holidays!

Timeline on Oracle's acquisition of Sun

I was reading an interesting little timeline on eWeek about how Oracle ended up acquiring Sun Microsystems.  You can read it here http://www.eweek.com/c/a/IT-Infrastructure/Anatomy-of-a-MultibillionDollar-IT-Acquisition-894207/?kc=EWKNLDAT05142009STR1.  Interesting timeline and also interesting how IBM and HP were also both involved.  In my experience in business, many deals don't happen because in uncertain times, people are afraid of making decisions for fear of being wrong.  What is interesting in the above timeline is that both IBM and HP could have acquired Sun, but ultimately, it was Oracle that made the bold decision to do so.

This reminds me of when my partners and I decided to acquire Simba in 2000.  Initially, I was against the deal.  One of my partners, Dale Reimer, was a strong supporter of doing the deal.  My hesitation was ultimately a good thing because it allowed Dale to negotiate a lower price.  However, it was Dale's boldness that got us to pull the trigger and acquire Simba.  Of course, the next challenge was taking what we had bought and making it even more successful.  I have found that when there are good deals to be had, human nature seems to always make us suspect as to "is this deal too good to be true?"  That is the point where the smart people are able to determine what to do.  Oracle seems to be able to pull the trigger at such times (so far so good) and my old partner Dale was also able to pull the trigger.  I need to make sure I learn from both Oracle and Dale Reimer. 

SAP's BI Chief Juggles Independence, Integration

InfoWorld has published a very interesting interview with John Schwarz, the former CEO of Business Objects who now heads the SAP Business Objects division. Definitely worth a read http://www.infoworld.com/article/08/04/02/SAPs-BI-chief-juggles-independence-integration_1.html. I have always been a fan of the data warehouse appliance and John speaks a bit about SAP’s BIA “We are also borrowing technology [from SAP] called Business Intelligence Accelerator, or BI Accelerator, an in-memory data management tool that was built by SAP primarily to improve the performance of access to massive data structures inside BW. We have attached to BI Accelerator our query and our search capability, we will probably also attach our OLAP client, so we'll be able to do queries or search or slicing-and-dicing of the in-memory data cube with lightning speed. Ultimately, I'd like to use the technology outside of the SAP BW context as well, so we'll have the ability to do in-memory analytics everywhere, so that's a very exciting development on the BI side.”

The future of SAP’s BIA is definitely something to watch. From past experience, the independent Business Objects was a big fan of the data warehouse appliance. At last year’s Sapphire, Business Objects was demonstrating a Netezza data warehouse appliance in their booth. At last fall’s Business Objects Conference in Orlando, Netezza had a big presence and you even had the Chief Transformation Officer from Business Objects speaking at a session that featured Netezza. Business Objects is definitely keen on the data warehouse appliance space. SAP has been strongly promoting their BIA – I have been to a number of sessions in the past at Sapphire and at the SAP BI and Portals conference where lots of good things have been said about BIA.

From John Schwarz’s comments, he is suggesting he would “like to use the technology outside of the SAP BW context as well”. This to me is a very interesting comment. In the SAP context, the BIA is an add-on to the SAP BW product, which can significantly boost performance. The question is, what will BIA look like outside of the SAP BW context? SAP is a huge and powerful organization, but one thing they currently lack is a strong database product. Their main competitors – Oracle, IBM and Microsoft – have major database products and HP is investing heavily in NeoView. Will BIA outside of the SAP context be a new and powerful database? SAP currently generates a lot of revenue for companies like Oracle whose database is used in a lot of SAP deployments. It makes sense that this is something that SAP wants to change.

John Schwarz also talks about integration into the SAP stack and also having independent BI applications. This is also another area to watch.

The OLAP Report

I had an opportunity to dive into the latest OLAP Report this weekend. In the latest publication, Nigel Pendse talks a lot about the consolidation in the BI space (http://www.olapreport.com/consolidations.htm) and he has an interesting chart that shows all the acquisitions in the space. His use of colours on this chart is especially interesting because you see how much consolidation there has been in recent years, as opposed to just acquisitions. In my blog on January 15th, I wrote about the consolidations in the BI space and Nigel’s chart visually highlights just how much consolidation did happen.

If Larry Ellison was running Microsoft, would he buy Yahoo?

That is a very interesting question that I never thought to ask myself. However, when you consider this question, think of the success that Larry Ellison and Oracle have had in acquiring 39 companies since 2005. With Oracle’s stock up so much in that period of time, it seems that investors in the stock market think that Larry and Oracle are doing something right. So, what would Larry buy if he was running Microsoft?

Here is an article by Randall Stross in the New York Times that is suggests what Larry might do... http://www.nytimes.com/2008/02/24/business/24digi.html. Definitely worth a read, whether or not you think Microsoft can learn from Larry.