6 Ways to Speed Up App Testing

Editor's Notes

• #4: Leading brands trust Perfecto Mobile to help drive business success USAA: Leading Mobile First bank with a history of firsts – remote check capture (2009), voice assistant (2013) United: 3-star app. 15M downloads (73% iOS), 400K daily launches, 66% participate in loyalty program Dell: Top 10 Global e-commerce site ($4B) optimizing its mobile quality program, in anticipation of a complete traffic shift from web to mobile within 2yrs
• #6: Bar with 14% waterfall, agile Goals bar Yoram: 80% of market has agile processes 80% of these are new to agile The market has agile processes – but most are new to agile. Our survey audience is mostly agile and going completely to devops
• #10: Evidence is key in reproducing bugs. We’ve all been there, we can’t replicate the issue reported. Is something unique to their device impacting performance? Is my network more secure, and in turn not returning a bug? Am I following all of the same steps they took to find it? With a tool that records transactions and then lets you share the video recording with members of your team, you’re not going to suffer this challenge anymore. With an immediate way to locate bugs, there’s no more trial and error slowing down your velocity.
• #13: As mobile app dev teams adopt agile practices, this is a really common reason why things get slowed down. Say you’re on a monthly release cycle, you spend 3 weeks developing and reserve the last week or so for testing and fixing bugs. Improved visibility into the “state of quality” can speed up this process. For this we recommend embedding quality throughout the entire lifecycle, a concept we call Continuous Quality. QA feedback is always visible and developers know where their code stands at all times, since it’s constantly being tested. It's about having control over the quality of your mobile app throughout all stages of your application development lifecycle. Adopting this method will increase the speed of development and reduce the number and severity of defects reaching production – ultimately cultivating higher end-user satisfaction.
• #16: Automating test scripts, also a challenge that slows down release velocity, is an important factor for optimizing testing. Hand in hand with this is stabilizing your test environment. What happens when you run automated testing across an unstable environment? False negatives happen. This impacts all of the hard work put into automating your test scripts. How can you speed up test automation? Start with automating the provisioning of your test environment to ensure that it’s stable and will support your test automation correctly.
• #19: I just upgraded to the Galaxy S6. It came out a few weeks ago – did someone on your team pre-order it or wait at the store to purchase it? Better yet, did you get before it was released so that you could test your app against it before your end users had a chance to upgrade? Probably not – device readiness is a full time job in itself if maximizing test coverage is a high priority to your quality strategy. Here we say let someone else do it for you. Let your vendor acquire new devices, maybe even get them before they’re GA. Let them manage your test lab for you- they can easily dedicate experts to you. And make sure they offer an SLA. Many of our customers have a whole team or resources dedicated to this – it’s a costly time sucker. It costs about $6,500 per device per year to manage in house.
• #22: One of our telecoms customers suffered from this same issue – every time a major device was released to the market, they had staff who had to acquire the device and ship it to the first location where testing was being performed. This testing group had a limited amount of time to use the device for testing before it was to be shipped off to the next testing group. When devices got lost or timelines weren’t being met, the staff member had to physically travel with devices, from test group to test group, to ensure they made it into the right hands and on time. Travel costs with limited usage time made for an inefficient process. Once they chose to put devices in the cloud, the procurement, shipping and deadline challenges were completely eliminated and a scalable, on demand available-everywhere cloud emerged.
• #25: A large bank we work with suffered from this problem – obviously it’s a common problem. But they were able to solve their issue by implementing test automation. Their pre-automated test environment was heavily reliant on manual testing, and included duplicate labs and a 16-20 day regression cycle. After automating their testing scenarios, they saw a 90% time reduction per regression cycle. That’s bringing weeks-long regression cycles down to a matter of hours. Think of the increase to velocity! They were even able to quantify that they saved $40,000 per test cycle.