Tools for the Business Analyst? Better Late than Never

Current Articles | RSS Feed RSS Feed

Tools for the Business Analyst? Better Late than Never

Submit to Digg digg it |  Add to delicious  delicious |  Submit to StumbleUpon StumbleUpon | Submit to Reddit reddit 
SinkingIn the 90's the Standish Group with their "CHAOS Report" quantified the alarming rate of project failure in the software development industry and the associated costs businesses incurred as a result. It also identified poor requirements as a primary contributing factor. Over the last twenty plus years, the CHAOS Report has consistently told us that the project failure rate has not dramatically improved and poor requirements continue to play a leading role.
RM or Requirements Management was a term coined to cover the entire problem set at the time and a push was underway to convince development organizations to stop relying on office automation tools (Word, Excel, PowerPoint, Email) to document and communicate their project's requirements throughout the SDLC. Solutions such as DOORs, RequisitePro, and CaliberRM were introduced to the market as RM repositories. Textual requirements were typed directly into these tools instead of a Word document and they could be used to track each requirement's priority, status, due date, risk factors, etc. Edit histories were automatically recorded and trace relationships could be established and maintained between requirements for more meaningful analysis of both requirements coverage and the impact of proposed change. The focus moved from the document to the individual requirements themselves.
However, this early RM approach assumed the correct requirements were entered into the repository to begin with and that was a false assumption. As the old adage goes "garbage in, garbage out". It doesn't matter how well you version, baseline, trace requirements to each other. If the starting requirements are incorrect, incomplete or missing altogether then the team will surely encounter some unpleasant and very costly surprises further down the line.
RD or Requirements Definition was born of the acknowledgement that requirements had to be discovered and defined properly before they could be managed forward through the SDLC. As you might expect, solutions appeared in the marketplace to help facilitate the accurate definition of requirements: DefineIT, SteelTrace, and Profesy (later renamed "Requirements Center").
Additionally, there was focus placed on the visualization, prototyping, simulation of requirements to enable true validation that requirements are correct, complete, and aesthetic concerns, important to the business users, were not left to developers to figure out. Again, solutions like Apptero, Simunication, and others appeared on the scene.
Today the industry knows this collective space as RDM: Requirements Definition & Management and it includes visualization/simulation. However, the desired goal is not a collection of standalone products each addressing their own specialty but rather, a single integrated solution that addresses all of the challenges analysts face day to day, effectively becoming their workbench.
Developers have had their IDEs and integrated source code repositories for decades. Likewise, QA has had their test management and execution platforms for years. Finally, emphasis has been placed on the intrinsic importance of requirements and getting them right and the BA community will get the attention they deserve in the way of holistic solutions catering to their needs as a defined profession. It seems we're the last in line but better late than never.

 


Comments

An excellent overview of history behind RDM. Without a doubt "single integrated solutions" within the BA domain is an admirable goal; I would only offer that even then structure must remain 'open' enough so that integration can flow down the lifecycle. So that the RDM suite can work with change mgmt can work with development IDE can work with QA tools etc.  
 
In other words, its all about the lifecycle, not just the sub-elements of it (like RDM for instance).
Posted @ Tuesday, December 15, 2009 11:54 AM by Mark
I agree with Mark. First, that it's a good overview. Second that the optimal structure would be "open" rather than "closed". (Do we understand what this means? I think so.)  
Coding and Testing are necessarily closed environments: they don't want non-experts mucking around or providing uninformed opinion. In business analysis, we do to a certain extent. We always trying to capture the business context, the real business problem, and we get insights from unexpected sources. One of typical errors in settling prematurely on an inadequate definition of the requirements. So keeping the process open to many eyes is important.  
 
Wiki software addresses some of these challenges. I expect to see it more widely used in BA work, though I have yet to convince a client to implement it.
Posted @ Tuesday, December 15, 2009 4:00 PM by Christopher Burd
Eek, sorry for all the typos! Any chance of implementing a preview function?
Posted @ Tuesday, December 15, 2009 4:01 PM by Christopher Burd
Yes, I completely agree. RDM as an integrated solution for the Business Analyst domain cannot standalone as an island in the greater SDLC sea. Exposure, review, questions, comments, etc. are our friends and all combine to help us narrow in on a requirement set that is truly accurate, complete, unambiguous, etc. True validation is our goal and once we achieve it, we need to communicate it effectively to the rest of the team and no better way to do that than directly through our teammates own workbenches via integrations.
Posted @ Tuesday, December 15, 2009 4:21 PM by Chip Carey
Now if we can just get the decision makers of organizations to realize this and have them look into these types of solutions, the world will be a better place.
Posted @ Wednesday, December 16, 2009 5:25 PM by KR
I've noticed more demand for collaborative tools over the last year, and this will certainly impact business analysts, probably for the better. Of course, there's the risk that clients and stakeholders will become so good at collaboration that they won't need so many business analysts to aggregate the requirements. Disintermediation, I think it's called. I don't seriously think this will happen. 
 
In the mean time, BAs are in an ideal position to lead the implementation of these collaborative systems. It may be a useful niche over the next few years. 
 
BTW, I am uncertain, and tend to be skeptical, about specialized BA tools playing a big role. I think we will be using customized versions of standard tools - especially SharePoint, and collaborative extensions to open-source products like Joomla and Drupal. 
 
Somewhat related: I blogged a talk by Philippe Taillefer on how collaborative software may affect management styles. You can can read it here:  
 
http://infodesignnotes.wordpress.com/2009/11/19/phillippe-taillefer-on-collective-intelligence-tools/
Posted @ Thursday, December 17, 2009 10:59 AM by Christopher Burd
Post Comment
Name
 *
Email
 *
Website (optional)
Comment
 *

Allowed tags: <a> link, <b> bold, <i> italics

2009 Blueprint Software Systems Inc. All rights reserved
Sitemap | Privacy Policy | RSS Feed | Subscribe to receive Blueprint News
News Rss Feed
Follow me on Twitter