The Feature
Interview with Luke Kowalski from Oracle
August 14, 2006 on 8:43 am | by Marian Crkon | In Conversations | |
Print
|
Email
I had a privilege to talk with Luke Kowalski, an Oracle Corporate User Interface Architect. Mr. Kowalski and the User Experience teams are making sure that the applications are as functional and great value to their customers as they are usable, eye-pleasing and intuitive to their users. Read on for the excerpts from our online chat. You can join the conversation by providing your questions as comments to this post.
I appreciate you are doing this. Please tell us how you got started with Oracle.
I came from a couple of stints at startups. One was bought by Netscape, where I worked in server UI for a couple of years, and the other was almost bought by Microsoft. I have been at Oracle since late 1998. I worked on e-commerce applications, back when being on the team got you an invite to Larry’s house in San Francisco. I then managed the server UI group and subsequently transitioned to an architect position.
What are you responsible now?
I now bridge the User Experience groups at Oracle. I work on partnerships, policy, UI technology, usability, and a number of marketing efforts around User Experience. Most of my time is spent helping either the Apps Middleware team (Thomas Kurian), or the VP of Applications User Experience, Jeremy Ashley.
Will you be so kind and describe the design process? How do those different groups work together?
We have had human factors folks here for the last 14 years. It is an engineering driven bus, but the executives have made considerable investments in user experience. This tends to show in our applications, servers, and the tools UIs.
As with any user centered design process we start by talking to the end user. They help us gather requirements. We then validate those against existing or new marketing data. We transition to low fidelity prototypes and we show them to user groups and in the usability lab. The next step involves a technical validation. As in: “Can we build it?” We repeat, just like the shampoo bottle says: design, get feedback, iterate, test, etc.
The silver bullet in the design process is the idea that we get to embed design patterns (UI building blocks) into the developer tool (JDEV). Before…we had guidelines and a police force…it did not work.
In the When Design Is Not a Problem white paper you co-authored you talked about “overcoming barriers in technology, organizational structure, legal, marketing, documentation and quality assurance (QA), and development tools†to have a design impact. What barriers do you see ahead of you as you design Fusion Applications?
Very few. John Wookey bet a lot on user experience. It is one of the key differentiators and exit criteria for any team in the apps division. UX works very closely with Jesper’s strategy, Fusion Middleware, and other stakeholders to ensure that we deliver the right applications, of the right quality. Have you heard about our Design Partner Program? This is where we involve Peoplesoft, Oracle, Siebel and other customers early in the development process. They tell us what they need!
No, not really. Tell me how do you define a user? Is that a CIO or people who would use the software?
Decision makers are not always the same as the end users. They often have different needs. We cater to the end user (productivity), while keeping the business/financial prerogatives in mind.
You also mentioned the marketing barriers. I totally agree when you say that “business requirements can often be at odds with those articulated by the end users. CIOs making purchase decisions have very different needs from the employees actually using the softwareâ€.
Oracle has recently spent a lot of energy articulating its Fusion strategy to its customer base (CEOs and CIOs). What do you think is the best way to get end users passionate about the future releases?
Allowing users to participate in the design process has been motivating for us and the companies we work with. We usually get positive reviews when we invite folks over to the lab to see the new software, or we often go to them (contextual inquiry method).
I agree that nobody cares if a piece of software is usable, if it cannot be implemented or sold. But it also goes the other way: how do we avoid selling software that sells, looks great on datasheets but is unusable, or unfriendly to the users? How do you make sure bad design does not “fall through the cracks�
Every product goes through concept/prototype/final reviews. That is the process in the applications division. Design patterns will save some labor, of course. IOW, there are only 3-4 ways to do a Master/Detail, or a wizard flow, or whatever. Most important review is the conceptual one.
It is often too late to change anything when we see the finished product in the lab the only thing possible then is “lipstick on a pig”. Information architecture, feature set, and flows have to be right first. Layouts, widget choices, Look and Feel are almost secondary when it comes to usability, efficiency, or having the right tool for the task.
The devil is in details. When users do encounter bad, unintuitive design, what is their best chance to get it corected? And I do not mean software bugs that can be resolved through Metalink SRs, but mostly bad features that work “as designed”.
We mined Metalink for usability issues…not that successful. The best way is to sign up for the design partner program! We do work with account reps and I have dived in on a few usability fires one via Larry, for a Fortune 50 company. That was fun.
It seems like the design discussions happen at too high of a level. On the other hand, it takes very passionate and committed users to spend time documenting and requesting improvements.
But usability issues are bugs. We file them as such internally and get them resolved.
They are not bugs; they are “features”! Now we are getting somewhere…
User groups are an excellent venue. We work closely with OAUG.
The OAUG provides many good tools to users with the Enhancement Request System, Forums, etc. to communicate their needs to Oracle.
Out of the 4 main applications suites now owned by Oracle (Oracle EBS, Peoplesoft, JD Edwards and Siebel) which one you think is best designed? How are you choosing the best features? That must be a lot of fun!
We use cold and hard facts. Siebel and Peoplesoft have had a bit more of a design mindshare. Now we are creating the perfect blend.
And you do not have to wait till Fusion. We are making incremental improvements, adopting a new look and feel for R12, leveraging all the cool things that the application server will get us (declarative UIs, skinning, composite application, Workplace UI shell). A lot of things are possible with 10.1.3. Not that 11g will not kill the competition…
Seeing few R12 screens, they looked “Peoplesoftizedâ€. Where can people see some “sneak peeks”?
We will be showing lots of cool, new stuff at Open World. We are also showing some stuff to early adopters, companies in the design partner program, testing with target users, etc. We have done previews for user groups, at conferences, and just showed a whole bunch of stuff to analysts.
We have a couple of screenshots on the User Experience web site.
2 Comments »
RSS feed for comments on this post. TrackBack URI
Leave a comment
Powered by WordPress with Pool theme design by Borja Fernandez.
Entries and comments feeds.
Valid XHTML and CSS. ^Top^
[...] Lastly, I would also like to mention the blogs by Oracle Applications executives Jesper Andersen, Jim McGlothlin, Luke Kowalski [see an interview with Luke] and John Wookey. While one post a month (on the best of them) hardly qualifies them as blogs, they are more like personal press release pages, it is the intention and the effort on the executives’ part that counts. I like the idea of being able to provide comments to their statements. _uacct = “UA-526159-1″; urchinTracker(); [...]
Pingback by The Feature » Application Blogs I Read [and Why You Should Too] — December 26, 2006 #
[...] Interview with Luke Kowalski [...]
Pingback by The Feature » The Feature is One Year Old! — February 1, 2007 #