Recently, I was on a designer website trying to buy a leather belt as a gift, but all I managed to get was a repeat error code. I flagged the issue via the online chat system, but the service agent told me there were no issues reported, so I switched to an alternative site.
This incident, which undoubtedly happens on a regular basis, will be unknown to the CEO and the CMO, but it is essentially a leaky pipe for the business. Plus, I’m now harboring a tiny bit of resentment toward that brand, unbeknownst to them, too.
Consumer digital demands are growing, this we know. According to a survey we conducted this past summer (via Econsultancy), 73% of consumers will jump ship to a competitor if a website is slow. Products, apps and services have to be rigorously tested if they are to delight consumers. But as you can tell from my scenario and the survey results, it’s as much about testing the intangible concept of ensuring we are providing an amazing customer experience as it is testing to see if the app actually works.
What Does This Mean For Quality Assurance (QA)?
The QA department has long been responsible for checking the latest version of a digital product (e.g., a website or app) works before going live.
However, with the changes in digital demands, we need to turn QA from a compliance function to a profit center. Testing should no longer be about just ensuring a digital product works; rather, it should be about testing that it delivers business outcomes along with an amazing customer experience. Business outcomes depend on the business in question. They could be basket conversions and revenue for e-commerce businesses, lives saved in health care processes, or online insurance quotes converted into policies for insurance. Digital products now run the enterprise, so quality is actually about testing the business itself.
The Role Of The Business In Testing
Enter the role of the business tester, who will be focused on ensuring that software is fulfilling its purpose rather than whether it complies with a specification. The business tester is there to ask the critical questions: Are we sure these new developments are not going to fail, causing a disaster to the business? Are we sure these new services or apps are going to function, perform and handle excessive loads? Are we tracking our products against revenue and business performance?
So, Who Owns QA And How Must It Evolve?
In many cases, the QA team still lives deep in the technology organization. Testing often happens in two main ways. The first is manual testing, in which humans manually test websites and apps. This incorporates intelligence, but it can't keep up with the pace of modern "continuous delivery" in which releases have moved, as an order of magnitude, from once every six months to once every six hours. The second way, via automated regression tests, features the same test cases that are run on every new release. But these tests are not smart and don't recognize what might have changed in a product or new ways in which users are engaging with the product in production.
The Difference That Could Make Or Break Your Business
QA needs to be continuous, not a one-off thing. QA owners also need to learn from both pre-production and post-production data. It needs to be intelligent and show what’s changed so that testing learning and reflections can be made. Finally, it needs to be democratized and visible to the business to ensure it’s being measured against critical business criteria.
Here are some practical tips to get your QA process started:
• How can you make QA continuous? Continuous QA means that as soon as the software has changed, it can immediately be tested (i.e., the software is being tested continuously as it evolves). Tests need to be autogenerated -- and a promising technique here involves using machine learning -- so that the tests become "self-healing" as the system changes.
• How can QA learn from pre-production? It's important to learn the riskiest parts of your software in terms of quality. Mobile app screens with more than six inputs fields are five times more likely to have defects in them than screens with four input fields. Risks can also come from software submitted after 8 p.m., software components with multiple developers making changes or software components being updated by a new developer. Most development teams have all this information today, but they don’t analyze it in any structured way.
• How can QA learn from post-production? From post-production, you can learn how customers really use your software (i.e., which features do they use, which paths do they follow, etc.), and this helps optimize your testing. But you can even go a step further and prioritize the paths that lead to a successful result for the software (e.g., a customer buying something from a retail app, a flight landing on time for an air traffic control system, a doctor completing a consultation report within 20 mins of the appointment for electronic medical records). But the transformational part is that you can learn in post-production how technical behaviors affect your business outcomes. This means you can test pre-production for the technical behaviors and predict how these changes will impact customers -- and that’s really testing the business.
Conclusion
While life-threatening health care apps really underline why QA should sit at the heart of an organization, for businesses to thrive, uncovering and fixing the things that are slowing down a website and having a faster website can have a huge impact. Better conversion rates and lower bounce rates can deliver revenue, keeping you in the race against your competition.
