“How do you stay up to date with technology?” asks how you pick new tools. More exactly, it asks which ones you think deserve your team’s time. Most engineers can recite a list of podcasts and newsletters. Nobody can check that list. So answer in four moves. Say what you follow and why it touches your work. Give one tool you adopted after a small trial, with its result. Give one tool you studied closely and turned down, with the reason. Then say what would make you look at it again. The second example does most of the work. Adopting new things shows curiosity. Turning one down for a stated reason shows you spend a team’s time with care. Below, an Android engineer builds this answer from two decisions. One is a screenshot testing trial that worked. The other is a Kotlin Multiplatform spike she argued against. The last section covers the case where you have adopted nothing lately.
Listing how you stay up to date with technology won’t separate you
Most engineers keep learning. The 2025 Stack Overflow Developer Survey asked about the past year. About 69% of respondents had learned new coding techniques or a new language. Technical documentation was the most common way to learn, at 67.8%. Blogs or podcasts came in at 34.8%. So “I read the docs and follow a few blogs” fits many of the people you compete with.
Some interviewers also score this question against a written standard. Amazon’s Leadership Principles include Learn and Be Curious. It says leaders are curious about new possibilities and “act to explore them.” The reading half of that is hard to prove in an interview. The acting half leaves evidence. A trial has a result. A decision has a reason and a person who agreed with it.
So spend one or two sentences on sources and the rest on decisions.
Sort what you follow by what it could change in your app
Take Marta, an invented Android engineer with five years of experience. She works on a fitness app with about 40 screens. She is interviewing for a senior role on the Android team of a travel booking company.
Her first draft listed six newsletters and two podcasts. Her second draft sorted them by what each could change. She reads the release notes for each library in her app’s build file. Those changes land in her code. She reads the behavior changes for each new Android version, because those break things she owns. She skims one weekly newsletter for everything else. She says that plainly.
That sorting is the first sign of judgment. It tells the interviewer she reads with a question in mind. That question is whether this changes anything for her team. The next two parts of her answer show what happened when the answer was yes.
The tool you adopted needs a trial and a result
“I started using X” is weak evidence. “I tried X on a small part of the codebase” is stronger, once you add what happened. Thoughtworks publishes a twice-yearly Technology Radar with four rings. Its FAQ says items in the Trial ring are ready for use. They are “not as completely proven” as those in Adopt. Thoughtworks is a consultancy that sells technology advice. The ring idea still gives you a useful order. Look first, then try small, then roll out.
Marta’s adopted tool came through her release-notes habit. Her team kept shipping small layout bugs on one screen, the workout summary. Each one reached users before anyone noticed. She read about a screenshot testing tool for Jetpack Compose. She proposed a trial on that one module for a quarter. The tests caught four layout regressions before release. Her lead then asked her to add the tests to two more modules.
Every part of that is checkable. There was a problem before the tool arrived. The trial had a fixed scope and a fixed length. Someone else made the call to expand it. If you only remember that you “started using” something, find those three facts before the interview.
A tool you turned down shows how you spend the team’s time
New tools carry costs that a demo doesn’t show. Dan McKinley made that case in “Choose Boring Technology,” published in March 2015. He wrote, “Let’s say every company gets about three innovation tokens.” Each new technology spends one. Older tools are cheaper because “their failure modes are well understood.” He wrote it about web companies in 2015. The same arithmetic applies when a mobile team adds a tool.
Marta’s turned-down tool is Kotlin Multiplatform. JetBrains makes Kotlin. Its product page says Multiplatform can share business logic between Android and iOS. Her app calculates workout calories on both platforms. The two versions had disagreed twice in a year. So she spent two weeks on a spike that moved the calorie code into a shared module.
The spike worked. She still recommended against it. Her company had three iOS engineers. None of them could review Kotlin. Every change to the shared code would have needed her approval. She wrote that up in one page. Her lead and the iOS lead agreed to wait.
Marta also named two conditions that would reopen the question. One was a third calorie mismatch in a year. The other was an iOS hire who reads Kotlin. The Thoughtworks FAQ calls its Assess ring things to look at closely, “but not necessarily trial yet.” Her conditions say when she would move it forward.
Marta’s full answer runs under a minute
Marta, her fitness app and every number in her story are fictional. Copy the four moves. Bring your own decisions to fill them.
I read the release notes for the libraries we ship and the behavior changes for each Android version. Those are what change my code. I skim one newsletter for the rest.
What I do with it matters more. Last year we kept shipping layout bugs on our workout summary screen. I’d read about screenshot testing for Compose, so I proposed a one-quarter trial on that module. It caught four regressions before release. My lead asked me to add it to two more modules.
I also turned one down. Our calorie math lives on Android and iOS. The two versions disagreed twice last year. I spent two weeks sharing it with Kotlin Multiplatform. It worked. But none of our three iOS engineers could review Kotlin, so every change would have waited on me. I put that in a one-page note. Both leads agreed to wait. We’d revisit after a third mismatch, or once iOS had someone who reads Kotlin.
Sources get three short sentences, because the question literally asks for them. The trial comes next, with a number. The refusal comes last, so the answer ends on her judgment. If they want more, the obvious follow-up is the one-page write-up. Marta can describe it from memory. A related question has the interviewer pick the technology for you. Our post on learning a new technology quickly covers it.
If you haven’t adopted anything lately, tell them what stopped you
Some engineers work where new tools barely get in. A bank may freeze its dependency list. A small team may have no slack for trials. Then the honest answer has no adopted tool. Don’t invent one, because the first follow-up will ask how the rollout went.
Give the evaluation you could run. Then name the gate it stopped at. Put Marta at a bank instead. Her version could go like this.
I wanted screenshot tests for our statement screen. New test dependencies need a security review that takes about a quarter. So I filed the request. Meanwhile I ran the tool on a personal project to check it first. The review is still open.
That answer still shows a problem, a test and a decision. The new team also learns that she will have trials ready the week their process allows one.