Operational Business Intelligence

March 28, 2008 · Filed Under BI 

BI has traditionally been the software tool of choice for executives doing business planning: figuring out what to do by looking at historical data. Answering questions such as, what were the sales like in each territory this past year, how many widgets came out of each plant, etc. The traditional BI solution provided the important base for strategic decision making and for setting the performance targets to support the strategy. For business review sessions, the BI software tells the executive if the targets were hit or missed, by how much, and the data could be used to further analyze the reasons behind each success or failure.

The trend now is to put BI software to work in a tactical role. This is near real time BI: comparing and combining transactional data with information from the data warehouse to make faster fact based decisions. It is no longer necessary to rely on gut instincts or to try to extrapolate from experience. Take for example, the staffing a software support call centre. With the release of a new product, historical data will show an increase in call volume in the weeks following the release. However, for the release of a brand new product (as opposed to an upgrade version of an existing product) or the release of a product into a new, unknown market, accurate historical data does not exist. With near real time data, any assumptions made can be adjusted quickly enough to meet the business goals. If call volume is higher than expected, then line managers can decide to get more staff trained sooner to support the new product. The sooner this decision is made, than the better the odds are of meeting the strategic performance targets for serving customers. Better still, if the decision is based on real information, executive support for the manager’s decision will likely be stronger.

This tactical, operations orientated BI is supported by three other information technology trends:

1. The need for a quicker ROI on all IT projects.
2. The IT infrastructure is in place to support a company wide roll-out (for example, broadband connections to remote locations).
3. BI can be delivered via email, a web browser or on a mobile device, all of which are ubiquitous and familiar to non-technical staff.

From a IT architecture perspective, how to deliver operational BI must be carefully considered so as not to impact the transactional systems. The major vendors in the BI space all have products to support operational BI, for example, Cognos Now! or Business Objects XI. Using open source BI is an interesting alternative, both for reducing costs and for potentially better integration into an exisiting application framework.

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • StumbleUpon
  • Technorati
  • del.icio.us
  • Reddit
  • TwitThis
  • Live

Why bother with Business Intelligence?

March 20, 2008 · Filed Under BI 

The ROI for BI software can be calculated once you have nailed down the affected business processes and the specific tasks that business users will be doing. But before getting to that point, even before the project scoping has begun, the question needs to be asked: why bother? In a nutshell, here are my arguments:

  1. Leverage the investment in ERP or CRM software. Looking at transactional data is not usually helpful, but looking at it in the aggregrate is invaluable. Business analysts and executives can identify trends, see a snapshot of the business at a particular point in time, or isolate the data for a specific business line or geography. Enterprise software often ships with some canned reports. This is a good starting point, but the reports are usually too generic to be useful for very long. A BI solution can make it easy to build customized reports, or even let the power users create their own ad hoc queries.
  2. Increase the ROI of the disk storage solution. The cost of disk storage is decreasing, but the amount of data is increasing faster. BI gives another purpose to the data, thus increasing its value to the business.
  3. Report on information from disparate systems. Data resides in many places and in many different formats. BI software can combine the data, aggregate it, and allow the creation complex reports that otherwise would not be possible.
  4. Avoiding spreadmarts. Dumping everything into a spreadsheet is not maintainable in the long term. Valuable corporate data should be easy to access and to understand.
  5. Expose data to customers in a relevant and controlled fashion. With increased emphasis on transparency and self-service, exposing customer data to the customers themselves can drive customer satisfaction and become a marketing tool. Many customers would value knowing, in the aggregate, the cost, the time or the location of their transactions.
These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • StumbleUpon
  • Technorati
  • del.icio.us
  • Reddit
  • TwitThis
  • Live

What makes an open source project ready for business?

March 19, 2008 · Filed Under Open Source 

A cursory search on Sourceforge, Google Code or Freshmeat will turn up a number of interesting open source products in almost every imaginable category. But which ones are worth investigating for use inside the enterprise? Here are five keys to separating the wheat from the chaff:

  1. There have been recent additions to the code base and there are developers active on the project. The best evidence is a blog, wiki or a website devoted to the project. Look for recent and regular posts and links to download recent patches.
  2. The code is written in a mainstream language. Look for Java, C++, PHP, Perl, Ruby or anything else that has been widely adopted. Projects written in D, Gambas, or Erlang will have a harder time recruiting developers, and future support of the project could be problematic. While the fringe languages have many advantages in their respective niches, the more established languages have quality libraries and well documented APIs. This will not directly impact the business users, but it can help avoid a technology dead end.
  3. Transparent licensing. Open source FUD loves to focus on licensing issues. Be wary of obscure licenses and stick to projects that are using one of the more common ones such as GNU GPL, BSD or MPL.
  4. A commercially supported version is available. This is the most telling sign that an open source project is ready for business use. The vendor will provide testing, training, documentation and control over the code base. They will also be working hard to build a community of users, developers and partners who will support and further enhance the product. Check also for 3rd party add-ons, which is a sign that the product is maturing well. The most well known example of commercial open source is Red Hat Linux.
  5. Some marquee customers have signed on. Search around to see if large customers have made public their use of the open source project or the commercial counterpart. The large volume of users will have weeded out the most common bugs, and they might be sponsoring (or pressuring the vendor) for industry specific enhancements.
These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • StumbleUpon
  • Technorati
  • del.icio.us
  • Reddit
  • TwitThis
  • Live

More Funding for Open Source BI

March 18, 2008 · Filed Under BI, Open Source 

A post on the BI for business people blog comments on the recent round of funding for Pentaho, an U.S. based open source BI vendor. Pentaho offers a full featured business intelligence suite, and they have positioned themselves as a competitor against Cognos, Business Objects, Hyperion and Microsoft. However, in a head to head competition, I don’t think Pentaho is a winner. They can offer lower upfront licensing costs, but this is offset with the risk of an increased total cost of ownership.
I think where Pentaho, and other open source BI vendors, will gain market share is in the following situations:

  • In the SMB space, where cost is an issue and BI deployments are less complex.
  • As a development platform for managed or hosted service providers.
  • For ISVs looking to embed a full featured reporting or OLAP solution into their product.
  • Where heavy customization is required.
  • Where interoperability with other systems is complicated.
These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • StumbleUpon
  • Technorati
  • del.icio.us
  • Reddit
  • TwitThis
  • Live

Business Intelligence News Roundup

March 17, 2008 · Filed Under BI 

Business Intelligence News Roundup

These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • StumbleUpon
  • Technorati
  • del.icio.us
  • Reddit
  • TwitThis
  • Live

The Problem with Late Software

March 14, 2008 · Filed Under BI, Project Management 

With the delay of the highly anticipated database, Microsoft SQL Server 2008, the problem of late software releases is again receiving attention. An article over on TDWI highlights four problems a delayed software release can cause:

  1. Delayed implementation
  2. Delayed training
  3. Loss of project budget
  4. Loss of momentum

I would also add the following:

    • Rushed testing, resulting in potentially unforeseen problems during the go live.
    • The time and budget originally allocated for enhancements is reduced, affecting long term satisfaction with the solution.
    • The snowball effect: follow on projects are delayed.
    • Disappointment among business users. Their hard earned trust in the IT department is eroded.

      After reading through the article, I realized that these challenges are not specific to BI projects, or even projects based on off-the-shelf software. Any project with a wide audience of end users, executive level visibility and high expectations will be severely affected by pushing out a release date. This is a risk that must be considered and mitigated from the very beginning of any IT project.

      These icons link to social bookmarking sites where readers can share and discover new web pages.
      • Digg
      • StumbleUpon
      • Technorati
      • del.icio.us
      • Reddit
      • TwitThis
      • Live

      Copy-Paste in Google Apps with Firefox

      March 10, 2008 · Filed Under Google 

      By default, the Firefox browser limits the use of the clipboard. This is a real inconvenience when using Google Docs (imagine using Microsoft Word or Excel without being able to cut and paste with the context menu). Yes, the keyboard shortcuts work, but these are not always practical. Fortunately, there is a solution, which I found here. I was able to apply it on my Windows Vista laptop with an extra step to get around the security on the application directory.

      1. Type cmd in the Vista “Start Search” box.
      2. Press CTRL-SHIFT-Enter to open up a command prompt as Administrator. Vista UAC will prompt for permission to open up the command line in administrator mode.
      3. Navigate (cd) to C:\users\[username]\Application Data\Mozilla\Firefox\Profiles\[profile name] You might not recognize the profile name.
      4. Open the user.js file. It might not be there, because by default, the user.js file does not exist. Create it by typing this command: notepad user.js Answer yes to the prompt to create the file.
      5. Paste the following lines into the file:
      user_pref(”capability.policy.policynames”, “allowclipboard”);
      user_pref(”capability.policy.allowclipboard.sites”, “http://docs.google.com”);
      user_pref(”capability.policy.allowclipboard.Clipboard.cutcopy”, “allAccess”);
      user_pref(”capability.policy.allowclipboard.Clipboard.paste”, “allAccess”);
      6. Save the file and close notepad.
      7. Exit the command line prompt and restart Firefox.

      Now you will be able to cut, copy and paste from the right-click context menu in Google Docs. Unfortunately, this still does not work for image files.

      These icons link to social bookmarking sites where readers can share and discover new web pages.
      • Digg
      • StumbleUpon
      • Technorati
      • del.icio.us
      • Reddit
      • TwitThis
      • Live

      Twitter for business

      March 8, 2008 · Filed Under Web 2.0 

      I came across the following articles while looking for business uses of Twitter, on DoshDosh and on LinkedIn. Most of these ideas are creative ways to replace email or instant messaging. There is some overlap between Twittering and blogging, but I see blog articles as being more permanent, whereas Twitter is more for fleeting thoughts and tiny doses of information. One platform can nicely complement the other and drive reciprocal traffic. I found some good tools here for use with WordPress. I also put Twitter on my Facebook page, since I really had nothing else there. This is an easy way to re-purpose content.

      These icons link to social bookmarking sites where readers can share and discover new web pages.
      • Digg
      • StumbleUpon
      • Technorati
      • del.icio.us
      • Reddit
      • TwitThis
      • Live

      How useful is Twitter?

      March 5, 2008 · Filed Under Web 2.0 

      I have been experimenting with Twitter recently to see how it could be useful for both myself as a consultant and for my clients inside the corporate world. So far, I have found it helpful for the following situations:

      • when I have an idea for a blog post that I do not have time to explore at the moment
      • to make note of a link to something useful or interesting on the web
      • to put something on my Facebook page (via the Facebook Twitter app)

      Thanks to the Twitter API, there are some interesting mash-ups appearing for several niches. However, I think the most interesting thing about Twitter so far is the use of Twitter inside the firewall, to fill the communication gaps where email and instant messaging are breaking down.

      These icons link to social bookmarking sites where readers can share and discover new web pages.
      • Digg
      • StumbleUpon
      • Technorati
      • del.icio.us
      • Reddit
      • TwitThis
      • Live

      How not to automate

      March 3, 2008 · Filed Under CRM 

      I came across yet another rant against poor customer service. There is too often a direct link made between technology, specifically sales automation or CRM software, and poor customer service. I would take the blame up a level and look at the decisions that were made during the early stages of the CRM project. Was there an existing process in place that needed to be automated? Were decisions based on assumptions about what employees should be doing, as opposed to what actually happens? Were there any meaningful metrics in place to measure customer satisfaction and loyalty?

      People almost expect poor service, so when a company does provide a great customer experience, the brand is rewarded with a loyal following. If management is smart, this loyalty is turned into more revenue and profits. Certainly smart businesses are using technology to maintain a relationship with their customers, but the technology is secondary. It can only support what is already in place. Even the slickest, best implemented CRM or BI system will not help a company that is not committed to measuring and improving how the staff interacts with customers.

      Minutes later, I came across a similar story by marketing guru Seth Godin. There are countless posts like this all over the blogosphere. Is technology really to blame?

      I believe that technology can enable and enhance customer service, perhaps even bringing good service into the exceptional category by leveraging people’s time and making knowledge about your customers more accessible. But if customer service is broken at the top, then technology will be a waste of money at best, and at worst, it will accelerate the side effects of treating customers poorly.

      These icons link to social bookmarking sites where readers can share and discover new web pages.
      • Digg
      • StumbleUpon
      • Technorati
      • del.icio.us
      • Reddit
      • TwitThis
      • Live

      Next Page »