« An Airline Story | Main | Knowledge-driven decision support systems and Decision Services »

More on requirements gathering

Another great post on requirements from Scott over on his blog - Types of requirements gathering. Worth reading as good background for anyone gathering requirements and rules (which are not the same, remember). A couple of rules-centric comments first:

  • Rules derived from regulations and policies are probably going to fall in the "known" category
  • Always add the original policy manuals and regulations to your document search - there's often a ton of rules there
  • When using existing systems it is really important not to blindly extract rules as you will end up with very technical ones, not business rules
  • You might find some rules or rule types using the "unknown" requirements techniques but you will likely do better if you focus on types of rules and let the development and evolution of rules happen later (by empowering business users to make their own rule changes) - this will better support change-time and business agility

As for analytics:

  • Replacing existing analytics is the only way they are likely to show up as "known" requirements
  • Using analytics to add value in a new way is almost certain to be an unknown requirement - people will need to be poked to come up with new analytics. "What if you could predict this, would it help?"
  • You can try watching experts and see what reports they use. Asking them why they use that report, what they get out of it, can be a good way to find new analytics too

In general you need also to keep asking "what decisions are you making" or "what decisions is the system making" to find the decision services you will need.

Technorati Tags: , , , , , ,

First time on the EDM blog?
Subscribe to the EDM blog feed or check out some other recent posts:


TrackBack URL for this entry:

Listed below are links to weblogs that reference More on requirements gathering:


Scott Sehlhorst

Thanks James! I seem to be saying that a lot.

The analytics portion would seem to definitely fall in the "unknown unknowns" bucket - it is (or at least should be) pathfinding. With a focus on the goals of the business, you can ask the question "What would you need to know to predict goal X, or track progress on goal Y". I think that that would be more effective than "If you could do X, would it tell you anything?"

It's also harder to do, at least for former technologists like me, who are genetically wired to look for applications of cool ideas.

The comments to this entry are closed.

Search Site



  • enter your email

Upcoming Events

  • FICO Tools & Analytics User Forum 2012
    BERLIN: September 11-12, 2012 LONDON: September 18-19, 2012 Gain new insights for improving business performance through advanced analytics and decision management tools.