Application quality assurance testing: what can SA learn?
There is a clear trend towards more agile development and DevOps activities, says Raffi Margaliot, senior VP and GM of Application Delivery Management Product Group at Micro Focus.
Software testing has taken on a new significance in companies, but also brings new challenges with it. This trend was explored in-depth in the article Quality at speed, where it highlighted the findings of the latest World Quality Report.
To summarise, software increasingly runs the world and more companies are using software as part of their core products and revenue streams. This is radically altering software quality testing, which is now being measured around customer satisfaction and alignment with business goals.
"The time to market for software has dropped radically," says Raffi Margaliot, Micro Focus's Senior Vice-President and General Manager of Application Delivery Management Product Group. He also co-authored the report.
"There is a clear trend towards more agile development and DevOps activities. The rapid adoption of IOT and the API economy is making the situation even more acute. It's leading a move away from test centres of excellence and other more traditional testing methodologies. Companies have to look at their QA processes and modernise them."
The new levers of QA
The WQR makes for interesting reading and highlights three key points. The first is the expanding role of artificial intelligence both in how it influences quality testing and how it can be used as a QA tool. The former refers to the challenge of testing AI software in a company, while the latter considers how companies can use AI to improve and speed up testing. It's perhaps the most obvious example of how radically software testing is changing to accommodate new business models, said Margaliot.
"AI is revolutionary both for business purposes and testing business services and products. But it is a very new field and while we see a lot of experimentation, the knowledge, expertise and maturity required to apply it to QA still fall short. Using AI for testing also creates roles that are very different from the typical tester profile."
But, AI is nonetheless very important to keep on the testing roadmap because it can play a role in driving QA's most potent new concept: automation. Automation of testing is drastically reducing costs, even when applied to software made in older waterfall environments. It is core to the idea of "quality at speed", yet isn't about speeding up testing as much as improving the coverage of testing. The WQR boldly refers to test automation as the single biggest enabler of maturity in QA and testing.
The third technology trend making waves in QA is blockchain. For the uninitiated, blockchain is a distributed ledger system that ensures accurate and nearly incorruptible record-keeping. Popularly used to support crypto-currency, it actually can replace most trust mechanisms. When speed and oversight are crucial, as it is in QA, blockchain can be very welcome. At least 66% of respondents to the WQR said they are using blockchain, though many of these instances may still be proofs of concepts or prototypes.
"Blockchain brings its own complexities to the table," Margaliot explains. "There are big concerns around security-related risks, data leakage, integration and a few other factors. But, as companies explore and mature around blockchain, those can be addressed. We should keep in mind that blockchain is suited for digitally mature organisations and many businesses aren't there yet."
What can SA do?
Institutions of all kinds, businesses, parastatals, academia, etc, need to evolve to make software part of their inner workings, revenue streams and operating models. This is a fact. Just consider the software inside cars: vehicle manufacturers don't call themselves software companies, but software creates their competitive differentiation. African companies are not immune, if anything, adapting to this trend faster will create opportunities to leapfrog global market incumbents.
But, there is a twist to the tale: so-called digital transformation is crucial to stay relevant in today's world. Yet its ingredients. cloud platforms, DevOps and such, have led to increased spend on infrastructure, tooling and re-organisation, leading to a spike in QA and testing budgets.
"Taking companies into the digital era is a good thing and brings a lot of benefits," says Margaliot. "But all change brings new challenges and companies must know what they are getting into. The knock-on effect of digital transformation on QA shows that this is not a revolution you just deliver in a box. It requires new thinking and new roles. For example, AI testing requires new types of validation and verification. You can't just fit the new on top of the old."
Automation of QA processes faces several barriers, including challenges with predictable and re-usable test data and test environments as well as shortages of skilled and experienced test automation resources. On the other hand, a lack of QA automation is creating testing bottlenecks due to the new speed of business mentioned earlier in this article.
What should companies do to move ahead while managing the new complexities of QA? The report recommends increasing the level of basic and smart test automation, but to do so in a smart, phased manner. Implement a non-siloed approach for test environment and data provisioning.
"Organisations need to build quality engineering skills beyond SDETs [Software Development Engineer in Test]," says Margaliot. "To optimise QA spends, organisations must first be able to track them. Enterprises must also start developing a testing approach for AI solutions now."
Local companies may believe they still have time, that these are trends impacting foreign markets. But that is a fallacy for several reasons, not least the improved performance and efficiencies that come with digital business models. Testing cannot lag behind, not if African business intends to remain competitive in the near future.