September 23, 2006

Scathing review of Oracle’s pre-Siebel BI products

Stephen Few offers a blistering review of Oracle Discoverer, its portal integration, and its UI in general. This fits well with what I said last November:

Obviously, Oracle has the potential to be a titan in analytics. But it doesn’t have its act at all together yet.

And so I agree with a couple of comments on Stephen’s post, to the effect of “Well, gee, no wonder that Siebel’s BI tools look like they’ll be the surviving technology.”

EDIT: Mark Rittman offers a lot of screenshots of Oracle’s Siebel BI Suite. If you look at other posts on his blog, you’ll see Discoverer as well.

September 11, 2006

My actual column on data mining

In a couple of recent posts about data mining, I referenced a Computerworld column due to run September 11. Wonder of wonders, they got it posted on the very first day. Here’s a link.

September 8, 2006

Where does data mining succeed, and why?

As previously noted, I have a Computerworld column coming out next week on data mining. The heart of the column is an enumeration of markets where data mining applications were having genuine success. Before I sat down to actually write the column, my list went something like this:

Read more

September 2, 2006

Further information on data mining

My September Computerworld column (I’ll post a link, no sooner than September 11) is about data mining. As promised in that column, here are some links and guides to further work on the subject.

September 2, 2006

KDD 2006 conference on data mining and knowledge discovery

I went to the KDD 2006 (Knowledge Discovery in Databases) conference in Philadelphia last week. It was an interesting, if weird experience. The conference had been billed to me as the place where all the world’s great data mining/KDD experts gather. This turns out to have been old news; the conference has apparently fallen off some the past 2-3 years. What are left are an academic conference and a small trade show that seem to be only loosely coupled. Here’s what I experienced at each.
Read more

July 21, 2006

Integrating BI with planning

One of my big themes these days is the integration of various kinds of analytics with each other, and with other kinds of IT. The following got a good reaction when I posted it in an SAP forum, in response to a question about integrating BI and planning. Read more

June 11, 2006

Second annual Text Analytics Summit

In Boston, June 22-23. Focused on text mining. VCs should consider going. My readers get a discount on their registration fees. More details here.

April 8, 2006

A blueprint for the analytical organization

I just ran across Andrew Stellman and Jennifer Greene’s great article What Corporate Projects Should Learn From Open Source. Basically, it’s a detailed argument for the virtues of the open source ethos — careful analysis, transparent analysis, open-minded analysis, honest analysis. Nothing real new there, but they did a particularly good job of spelling it out.

And then it struck me — their arguments don’t just apply to software development. It’s really just basic good management sense, very similar to what business schools and BI vendors have been trumpeting for years. Only in their case the motherhood-and-apple-pie rhetoric is bolstered by — well, by analysis.* After all, we can do fairly objective after-the-fact observations as to whether and to what degree any given development project has succeeded.

*At least of the handwaving sort — but does anybody really doubt that what surveys and anecdotal consensus seem to show about software development projects is essentially correct?

If I can, I’ll flesh out these ideas into one of my next columns.

April 8, 2006

UI musings

In the past, I wrote vigorously and often about UI. I knocked heads long ago about the superiority of GUIs to character-based interfaces, and even long before that about the advantages of OLTP (which we called “real-time” then) over batch processing. In the latter 1990s, I put a lot of time and effort into search, better alerts-management, and context-sensitivity in general. And recently I’ve focused a lot of my research on analytics, often with a theme of “Yeah, yeah, the server-side stuff is cool — but let’s talk about how people actually interact with this stuff.”

Still, I feel something has been lacking, probably because there just are so many different UI subjects to talk about. So here are some quick-hit thoughts on UIs. The first ones are from my Computerworld column running next Monday, which is called (with apologies to Sports Illustrated columnist Peter King), Six Things I Think I Think About UIs.

1. “A good GUI interface” is the most important feature a product can have. In many cases, the GUI is the feature set, whether we’re talking about operational apps, BI, or IT administration tools. For example, when I looked into the security market a few years ago, it turned out that Checkpoint’s rise to dominate the firewall market in the late 1990s came about because it had a good GUI rules-administration interface, while otherwise equal or superior competitors didn’t.

2. Web UIs are now, finally, much superior to the client/server systems they replaced. That wasn’t true until recently. But now they’ve leapfrogged client/server a little bit in pure GUI functionality. (I somehow like this article on the technology, even though I’m not sure what I learned from it.) And they’re always been way ahead in application navigability.

3. BI look-and-feel is on the upswing. Business Objects is a good example of this. They brought their thin client products up to client/server GUI standards. They fiddled around in usability labs with screen real estate and so on to polish the dashboard UIs further. And then they went out and bought what is now Crystal Excelsius.

4. Portal technology is headed for a boom. I have a whole whitepaper in the works on that one.

5. Natural-language interfaces are advancing too slowly. Unfortunately, big vendors remain clueless about language-based UIs. Enterprise search is a fiasco. Most single-site web search is even worse; in almost every case, it’s inferior to just googling on search string + site name. As for natural language/voice command/control and navigation – we’re nowhere, Inquira and Sybase AnswersAnywhere notwithstanding. (I bet you can’t name a single user of either product off the top of your head. To tell the truth, I can’t either, except that I’m pretty sure Inquira powers the websites of a couple big-name cellular providers.)

6. Microsoft Office is a huge question mark. Office is facing a huge, if slow-moving, threat from open source. And the product has basically been stagnant for years, in that few users have cared much about any of the newer features.

Microsoft’s stated and obviously sincere strategy is to make Office an important window in the world of database applications. The Proclarity acquisition this week is surely part of that. So are the moves to make XML important in live documents, which dovetail nicely with the XML file formats of Office 2007.

EDIT: See also: You can start to imagine a world of Office as a business application platform,” Witts said.

7. In particular, Excel is a huge question mark. On the one hand, the BI industry is doing ever more to make Excel into a viable BI client. On the other hand, they’re trying to replace Excel as the data storage engine of choice — and in some cases even as the client — for budgeting/planning/etc. It does seem to me as if server-based planning is sweeping the enterprise world. So where does that leave Excel? Will it ultimately be anything more than a glorified calculator?

8. Home UIs are challenging work ones. Back when I consulted a lot to AOL in the late 1990s, I (correctly, it turns out) warned them that their client’s lack of functionality in areas such as email and browsing would get them into big trouble, because users’ expectations were being set higher at work. Now the reverse is at times true. Home bandwidth has caught up with work bandwidth, and webmail is in some ways better than Outlook. Meanwhile, a few websites out there are actually pretty usable, annoying clutter notwithstanding — and most of them are focused on consumer shopping, e.g. Amazon, Land’s End, et al.

9. Usability labs are crucial. Back in the 1990s, usability labs were new. Microsoft and Lotus and Borland had good ones, and Oracle hired Dan Rosenberg away from Borland to set up theirs. Other than that, there mainly were third-party consulting firms, or very primitive inhouse operations.

Well, I’m still not convinced that very many inhouse usability labs accomplish much. But I do know that whether it’s inhouse or third-party, you must use a lab if you’re serious about offering a competitive product.

10. Rules-based interfaces are too primitive. This isn’t really an interface issue so much as a functionality one — but as noted above, the two are inseparable. True declarative rules interfaces, which function with the same flexibility as 1980s-era expert system shells, are way too rare. Executing a set of rules in a set linear order is not the same thing at all.

March 9, 2006

Issues in privacy

My March Computerworld column is an exhortation to IT workers to get involved in IT-related public policy issues. Probably the most complex and serious ones are in the area of privacy. Options I posed include:

• Do nothing.
• Maintain sharp limits on government acquisition and retention of information.
• Mandate that the government keep its information in separate silos.
• Create strong rules about how governments can use information once acquired.
• Hamstring corporate acquisition, retention, or use of information. (Much of the government’s potential data comes through private channels.)
• Various mixes and matches of the above.

My own view, which I plan to lay out and defend in a series of posts, is a complex one. Historically, protections such as the Fifth Amendment to the US Constitution have focused on limiting the government’s access to information. And this has been wise. As was shown for example by the misuse of FBI information under J. Edgar Hoover and the misuse of IRS information in the Nixon Administration (which Nixon claimed he didn’t innovate), once information gets into government hands its use is hard to control.

I see no good alternative to preserving these safeguards as long as possible. Maybe the Bush Adminstration should have gotten legislative permission for its data mining adventures and maybe the permission should have been denied, but the fact that they pursued them while circumventing the legal safeguards is utterly deplorable. That few people have suffered from the violationof these safeguards, or the worrisome provisions of the Patriot Act, merely proves that our multiple layers of safeguards are strong. It does not justify chipping away at them illegally, and I’m not real thrilled about legally undermining them either.

On the other hand — many of those safeguards are eroding fast. The amount of information the government can or potentially could obtain legally from credit card records, electronic auto toll tracking, etc. is staggering. National ID cards and passports are going electronic. The US Constitution apparently doesn’t prevent sensitive devices snooping into your house from the outside. Web surfing behavior is being submitted as criminal evidence. The government WILL have access to a very complete dossier on your activities, rather soon. Legislation to mandate this data be maintained in independent silos, while worthy, is just a stopgap.

And so in the US (and other developed countries, I would think), it is not just enough to fight government information acquisition. That’s a losing battle, especially since the most absolute safeguards can not be maintained in the face of the terrorist threat. Thus, I think we also need some reaffirmation — in principle, at least, if not in actual law — that “thou shalt not be hassled.”

I’m not yet sure what form(s) I think that needs to take.

← Previous PageNext Page →

Feed including blog about enterprise technology strategy and public policy Subscribe to the Monash Research feed via RSS or email:


Search our blogs and white papers

Monash Research blogs

User consulting

Building a short list? Refining your strategic plan? We can help.

Vendor advisory

We tell vendors what's happening -- and, more important, what they should do about it.

Monash Research highlights

Learn about white papers, webcasts, and blog highlights, by RSS or email.