We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.
Start a Discussion
BI
user-pic

What is the key to improving BI's ease-of-use?

Vote 0 Votes
In asking about the key developments for BI in 2011, one thing that kept coming up was the need to improve BI's ease-of-use.  So what is the key to making this happen?

7 Replies

| Add a Reply
  • The key is self service, so tools where you don't need the expert/power user, but a tool with all the data at you fingertips. With search/query options to find the data you are looking for and the ability for you to make yourself your own reports and dashboards.

  • The key feature is the tool transparency. The software should get out of the way and let the user work with data. It should also blur the boundary between data visualization and data manipulation letting the users making simple simulations, classifications etc.

    How this can be achieved, is yet to be defined; if I had a proven, working, recipe, I were rich.

    Happy New Year to everyone!

  • Ease of use, in BI or elsewhere, is a worn-out idea that shouldn't still be an issue in 2011. After all, what "use" is supposed to be "easy," and what other variables does that overlook?

    Starting with APL in the late seventies until today, I find the things that made analytical software difficult to learn and to apply have not changed much. The presumption was that pretty GUI interfaces would free people from command line, code-driven work, but what was often overlooked was that the underlying process was still the same, and GUI's often made it more difficult. All of those tables, drop-downs, lists, indented hierarchical categories, little boxes at the bottom of the screen that said "More" or "Advanced," that no one ever understood; the need to click, click, click in a certain order to get the result you wanted. A mess.

    Unfortunately, we see this same awful situation on the business/consumer web, too, though the best examples of developing applications that really suit people's needs are there too.

    But it's almost 2011. We have on the order of 7 or 8 magnitudes more computing power at our disposal (I could be wrong, it could be more), wouldn't it make sense for the computer to do a little more work? Wouldn't it make sense for the data to be a little smarter and know how to organize itself in a given domain?

    Data exists in an organization as an independent entity that has to be "mastered" and "QA'd" and integrated before it can be put to useful purpose. This is still the same old industrial assemblyline mentality laid down by the fathers of our business computers in their short-sleeved shirts and pocket-protectors.

    I have spoken, directly, to perhaps thousands of business users of BI, and this is what I've learned. Forget ease of use, give people tools that are relevant to the work they do, which you can't do unless you go out and find what they really do. And make it instantly understandable and incorporate the ability of the machine to learn and even anticipate what people need. Don't just inform them, be a part of the process, actively and passively.

    -Neil Raden

    • Meh - That all sounds fantastic, but BI always suffers from "Garbage In, Garbage Out" syndrome.

      Sure I can help you set up the perfect system that works perfect... as long as there is perfect data.

      Data input has to come from somewhere. For many businesses, and especially SMBs, there is no one-touch automated solution - data gets put in by: secretaries, interns, distracted managment, bored ceos.

      If there isn't an easy interface that makes sense and helps them input that data, it ain't gonna happen.

      Excel trumps a lot of new solutions because it is familiar not necessarily easy or better.

      Things will change tho - the new generation workforce are going to expect facebook like functionality, extensive use of suggested input (AJAX!) etc.

      A common fatal flaw to SO MANY TERRIBLE BI solutions is the belief that the end user knows what to do. Learn from great UIs (OS X, etc.) and learn that it's not about capturing everything, it's about capturing what is needed.

  • Celebrating the year end and a new year, let me post a radical idea.
    BI tools to drop any user interfaces (UI), and have UI to 100% depend upon MS Excel and/or OpenOffice Calc.

    I believe that the notion, "the final destination of any information is Excel" isn't a joke. It's simply a truth.

    As long as I have experienced, I always had to use Excel to add a final touch.

    I hope you disagree!! ;-)

    I wish you all, a happy and peaceful new year!!

  • Adrian,

    I should have elaborated the concept of smarter data. When the meaning of data and its relationships are carried around with it, a lot of GIGO will disappear. The representational frameworks we use for data today, such as data warehouses and MDM provide only a rigid, purpose built result in a doomed effort to provide that "single version of the truth," if they work at all. In the full-bore MDM projects I've examined, I can only say they're like a heart transplant - you feel great if you survive the procedure (and luckily for heart transplant patients the survival rate is much better than it is for MDM).

    A semantic approach, which has been proven effective in many areas beyond enterprise software, such as B2B/B2C web apps as well as defense and intelligence mega-data apps is, in my opinion, the only real solution. Semtech has some problems to work out, such as the cumbersome nature of RDF stores and poor query response from SPARQL, but try to remember what a mess SQL was for analytical reporting 15 years ago (and still is in many cases).

    As for setting up the perfect system, no thanks, that wasn't my implication, but if you make the data smarter and the manipulation and presentation systems more adaptive, we'll forget about ease of use and the distinction between enterprise, SMB and individuals be irrelevant.

    I'm not waiting under water for any of this. These innovations are coming from smaller companies and may never see the light of day.

    -NR

  • Taking an instance of a day in the life of a business decision maker, the following are the typical activities in his/her realm of work.

    Step 1 is to know what happened - Dashboards, Scorecards, Advanced Visualization, Embedded BI, Self-service BI etc. serves this need.

    Step 2 is to understand how it happened - Decision maker wants to check & review the facts with the team to ascertain what led to the observed situation. Collaborative BI techniques can help in this direction.

    Step 3 is to evaluate what could happen - What-if scenarios, Simulations, Predictive Analytics, Forecasting etc. helps the business user to understand the future state if certain decisions are implemented.

    Step 4 is to make happen - Realm of Actionable BI where the user can initiate certain actions from within the BI environment. These actions, will typically involve interaction with IT platforms like BPM, SOA etc., and will result in changes to business processes.

    The BI techniques underlying the 4 steps have to improve in its own way but the key message am driving at is that - BI has to blend-in seamlessly with the day-to-day activities of business users (knowledge workers) in an organization, so as to be used more easily & effectively.

Add a Reply

Recently Commented On

Monthly Archives

Blogs

ADVERTISEMENT