Keyword-Driven Testing: Content Reuse in App Development

If you've spent any time around software QA, you've probably heard of data-driven software test automation, but how about "keyword-driven testing"?

If you've spent any time around software QA, you've probably heard of data-driven software test automation, but how about "keyword-driven testing"? This newer, more modularized method is gaining popularity because it speeds time to market and lowers costs. Administered through an interface such as a content management system, keyword-driven testing also can allow business analysts (non-programmers) to assemble test scripts, pulling down the proverbial — often palpable — wall between IT and the business.


The Fortune 1000 see the high level of uncertainty they face about economic conditions, talent pool changes and other variables as a growth impediment. Many of these companies in the services, manufacturing and high-tech industries polled by Wirthlin Worldwide cited three main solutions they're pursuing: better analytic and BI capabilities, organizational cultures that adapt more easily to change and more robust IT infrastructures.
Software test scripts are conventionally composed ad hoc by a coder. Some software development tools help automate testing by recording tests that are run, allowing "playback" of the test routines. However, an entire test routine is rarely, if ever, applicable to more than one release of one application. Data-driven testing adds some modularity by keeping test input and output values separate from the test procedure, but the procedure itself is still in a monolithic script. Keyword-driven testing breaks the test procedure into logical components that can then be used repeatedly in the assembly of new test scripts.

The Allegis Group, a staffing and recruiting firm, continually needs to test a variety of applications. Because Allegis must ensure that its applications meet its performance thresholds and provide excellent user experiences, the company wanted to improve the scalability and flexibility of its testing processes.

Solutions the Fortune 1000 will use to fight uncertainty
91% BI and analytics
84% Nimble culture
74% IT infrastructure
"Both functional and load testing are critical during upgrades of our Web and client/server applications," says Eric Amo-Gottfried, the team lead for Allegis's WebQA group. Testing helps identify deficiencies early in the product life cycle to minimize costs and avoid giving users a bad experience. To speed the process, Allegis has adopted keyword-driven testing. The company uses modular test components, identified by keywords such as "login" or "verify" to automate low-level actions such as login and balance checks. Reuse means faster test script creation. Allegis has reduced its time to production by about half.

Keyword-driven testing is quickly becoming more prevalent, according to Forrester Research. "We are seeing more mainstream vendors producing these tools, which indicates a growing market" says Forrester analyst Carey Schwaber. The market's growth is due, in part, to the savings that companies can capture with this approach.

— Jerri L Ledford

Get Smart

A surprising 85% of any process is muda — Japanese for "waste" — according to the discipline of value stream mapping. Manufacturers are starting to change their businesses to get back as much of that 85% as possible. How? By using some of the knowledge captured in Inescapable Data: Harnessing the Power of Convergence by Chris Stakutis and John Webster (IBM Press, 2005). This easy-to-understand book inspires readers to take advantage of the staggering (sometimes anxiety-provoking) opportunities the increasing pervasiveness and interoperability of data offer. If you're a technology evangelist who isn't getting through to business decision makers, give them this book.