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

How Do You Prevent BI From Becoming Too Much BI?

Vote 0 Votes
From this Nari Kannan blog, When is BI Too Much BI?, where he states, "Too many fractional BI efforts stand the risk of moving in the opposite direction too much. Sometime too much BI may be harmful also!" What's the best way of preventing this from happening?

9 Replies

| Add a Reply
  • I don't know, but it's an excellent question. In fact, it's a more important question than the usual ones, about technology.

  • This is a great question. And, rather than supply a specific answer (indulge me here), I'll try to point out a clarifying observation.

    We are currently in the 3rd wave or generation of business intelligence and determining the right amount of BI (in any given organization) will likely come in the 4th wave or generation.

    First generation BI was about analyzing data in any way possible, simply because it was so sorely needed and the tools were raw and special-purpose (remember the '80s?). Second generation BI marked the great expansion of BI and Data Warehousing, with emphasis on centralization and control (basically the '90s through early this past decade). As the 2nd generation aged, the tools and projects became more complex and costly to implement, which is a major factor causing the third generation of BI.

    For at least the last 5 years, we've seen a new and disruptive expansion of BI. This expansion is occurring within a wide variety of applications (embedded BI), directly at the end-user level (going around the IT departments) and in smaller organizations than ever (because of the now-distributed expertise from the first two generations). This 3rd generation, necessarily, is about more BI and in a much less costly and simpler form. And, so Nari's question is both expected and timely.

    Will the fourth generation of BI create tools, skills, projects, and diplomacy that yield just "the right amount of BI"? I don't know. But, I'm looking forward to a front row seat that helps to define.

    Brian Gentile
    Chief Executive Officer
    Jaspersoft

  • It’s all about perception.
    “BI? in the eyes of a Business User is nothing more than getting the right information when they need it the most whether in terms of trends, patterns or alerts – they don’t care how the information is delivered to them apart from the delivery channel.
    On the other hand a technologist could dabble with a host of available BI platforms to deliver something very simple and in the process confuse the Business users with a plethora of fancy User Interfaces.
    The key is whether you are solving a problem or are you adding to it - The discretion is truly yours!

  • BI becomes too much BI when the data and the tools exceed what’s necessary for a user to make a good business decision. In other words, effective BI should involve delivering the right data in the tool that is commensurate with the end users ability to use it. For example, getting a simple list of overdue accounts in a spreadsheet format is all that is needed for a person in accounts payable to create a top ten overdue accounts list – there is no sensible reason to force this person to use a much more complicated and sophisticated tool. Conversely, a systems analyst would need (and be able to use!) BI tools that are infinitely more powerful and feature rich as well as access multiple silos of data.

    John Kitchen
    CMO & SVP
    Datawatch

  • The only way to have too much business intelligence would be if there was a way to have too much of any kind of intelligence. Maybe the problem is too much of too fast business? Is that possible?

  • user-pic

    I believe that BI itself doesn't become too much, it's rather people side issues. As one of BI users, what I have observed are that as long as people use BI as a tool and use their insights and really think about their businesses, BI never becomes too much. But once people become BI dependent and use less insights and think less, BI becomes seemingly too much.

    Kengaku, Satoru
    COO
    S&S Global Services

  • BI is like any other function within an enterprise. Once you have multiple instances you run into the danger of producing unwanted costs and/or redundant and inaccurate information. Looking at this from an enterprise architecture standpoint, I think the problem is both functional (business architecture) and data-related (information architecture). Each domain needs to have redundancy driven out, especially data.

    Coalescing data into one source of truth is a critical tactic in driving out waste and increasing the value of the information in an enterprise. Sure, its easier to draw this pattern on a white board than implement. Creating common formats and disambiguation rules to collect the data from disparate sources is a challenge. But organizations can't afford not to. I like to think that data is not just an asset but a competitive weapon when used correctly.

    Mr. Kannan gets it right. This type of action takes some corporate, um, fortitude to pull it off:

    So what may be the answer? Somebody should be responsible for distilling a single version of the truth - pick any one version and consistently follow it for action. As long as it is the same version it will effectively show what is important - trends!

  • Interesting question. Also raises the question of why nobody (atleast to my knowledge) talks of too much ERP or CRM or SCM. I guess the non-deterministic nature of the BI world (aided by both inductive & deductive reasoning) has a lot to do with the notion of 'too much BI'.

    Too much BI can be looked at in multiple perspectives:

    1) Departments within an organization running their own BI. The solution to this problem as Eric puts it beautifully in his answer above, is to remove redundancy at both the business and information architecture level.

    2) Too much analysis (slicing & dicing of data) without commensurate improvement in the quality of business decisioning. In other words, the marginal utility of analysis falls after a certain point and good BI users should know when to stop 'boiling the (analytical) ocean'.

    3) BI lacks the process rigor, of say, a Six Sigma methodology. The result is such that all good things associated with a well defined process methodology (reuse, knowledge management, measurements, documentation, etc.) are typically not enforced in a BI program, giving the perception of too much BI.

    A well defined process architecture in conjunction with a well thought-out business & information architecture will go a long way in alleviating the problem of too much BI.

    • Hello Karthikeyan

      This might a bit beyond this specific topic, but I recently have had a similar discussion about "too much ERP or CRM or SCM" or enterprise system with a couple of my friends.
      There are cases, where:
      ex. a) Recently, a wholesaler's sales assistants don't memorize anything about products or sales transactions. Whenever one of my friends calls sales assistants, they ask about a product ID, and/or a transaction ID, and take time waiting for screens pop up and reading those. Until a few years ago, those folks had memorized important portions of their businesses. At least, as a customer, my friend could feel people were more responsible for their businesses.
      ex. b) In a warehouse, even though piles of un-shipped products get bigger and bigger, warehouse people had just continued receiving products, since an ERP continued printing out expected product receipts. Those folks didn't warn until their warehouse ran out of storage spaces. But the the time warning was raised was too late, the company had to write of huge amount of dead stocks. The cause was duplicate purchase and production orders got issued multiple times, because of wrong setups in MRP and SCM.

      I cannot tell where a red line is drawn. But when people become more dependent upon ERP, CRM, SCM, or whatever, people start think less about their businesses. Meaning those systems become relatively too much, I suppose.

      Can better BI prevent such problems? I think Yes and No. Yes for short term with new signals and criteria implemented to a management BI screen or whatever. But in a long run, this passage ends up with too many signals and criteria. I think we need to continue educating people that people are responsible for businesses, not systems to prevent too much xx.

      Kengaku, Satoru
      S&S Global Services

Add a Reply

Recently Commented On

Monthly Archives

Blogs

ADVERTISEMENT