How much money do you waste fixing bugs?
The later a problem is identified in the development process, the more difficult and costly it is to fix. Not only can the bug create significant re-work but the farther down the path the production, generally the more complicated the fix becomes. According to the Center for Systems and Software Engineering, a fault corrected in production is 200 times as costly as a fault corrected in the requirements stage.1 Even a fault corrected earlier in the design phase is 4 times as costly to correct as a fault corrected in the requirements gathering phase.
Identified too late in the process bugs can cause projects to miss milestones and go significantly over budget. Worse still, once a bug is in production it can create a host of problems including down-time, inaccurate analytics, and poor performance.
Why do so many bugs get into Production?
Despite the known heavy costs of bugs, many get into testing and production. Each phase of development creates new ways of introducing these errors:
- Requirements gathering: Business requirements are difficult to organize and link to technical requirements and metadata which causes redundancy, incomplete data requirements, and inefficient architectures.
- Design: Communication errors between functional and technical teams create specifications that are incomplete and inaccurate. These faulty specifications become the foundation of all proceeding work.
- Development: Poor specifications lead to code that meets the wording but not the intent of the business users. Time constraints lead to poor or inefficient code.
- Testing: Testing efforts cover a small range of scenarios based on inaccurate specifications that do not test the actual code well. This stage provides false confidence in the quality of the code heading into production.
While there are many tools used to improve code quality, the primary source of serious defects is during requirements gathering and specification writing.
Agile Data Suite greatly reduces development costs by limiting the number of errors in the development process through accurate, complete specifications:
- Integrates technical and business requirements
- Prevents communication errors by creating an enterprise-standard data understanding
- Object-oriented business model and accurate GUI-based mapping reduce redundancy and re-work
- Designs and organizes requirements holistically, allowing strategic architecture design with tactical project build-out
- Centralized database creates a single version of the truth and eliminates spreadmarts
For more information on these and other Datric products and services, please contact us
1 Overview of Software Engineering Principles - 2001 USC
ADS overview brochure