Digging deep on AI with Niklas Lindstrom

Niklas Lindstrom, our CTO and Co-Founder who was recently named one of the top CTO’s in the analytics startup space, appeared recently on the Hashtag Trending Weekend Edition podcast to give a deep dive into AI and unitQ.

Interviewed by Jim Love, publisher of IT World Canada, Nik discusses how unitQ makes sense of data — via AI — that unitQ ingests to provide our customers a real-time understanding of the Voice of the Customer. This parsed data enables organizations to build exceptional customer experiences based on that user feedback.

Human in the loop

Nik explains the “human in the loop” concept, where unitQ uses “data labelers” and machine learning to train the AI to accurately recognize what people are saying about organizations on Twitter, Reddit, Discord, app store reviews, chatbots, support tickets, and you name it.

“So we need to have very detailed labels on it. And it’s a challenge for, I think, all AI companies out there that they have to tackle, that they have to treat the data as the most valuable asset they have,” Nik says.

The collective mind

Nik says building an AI is a symbiosis between humans and machines.

“So an AI can never become better than me if I would be the only data labeler. But if we work as a team here, Jim, we can train an AI to become better than one of us, because we have a collective.”

What’s more, data always needs constant training and auditing because it can become biased, Nik says.

“You need to build and you need to train the AI with data that you trust, something that you 100% rely on. But you also need to keep auditing AI. So even after it’s trained, you need to make sure that it’s doing what you expect it to do, so it doesn’t get any assumption to say something wrong,” Nik adds.

Listen to the podcast here.

To learn how unitQ and its AI-driven platform can help your organization understand the Voice of the Customer to help create exceptional user experiences, schedule a demo today!

David Kravets is the Senior Content Marketing Manager for unitQ.

Best Practices

Helpshift advice: How to collect mobile app feedback

App feedback is essential in all app building because it is important to create an app your users will love and prevent churn. Companies such as unitQ are leading the field in analyzing app user feedback by making it easier to search, monitor and analyze feedback with its AI platform.

Why do you need feedback? Why is feedback critical?

In-app user experience will make or break your retention funnel. A well-executed feedback loop and player support can guarantee app longevity and drive financial success. Feedback will help your developers build better games and QA can more quickly replicate bugs, while support staff better addresses user frustration.

Feedback, like any other measure used to determine the success of your app, allows you to test the performance of your app. How your app performs correlates directly with your customers’ satisfaction and LTV (lifetime value). To make sure your app is satisfactory to your customer, gather qualitative data and insights directly from your users.

Continue reading this article on the HelpShift blog where it originally appeared.

Best Practices

Best practices for organizations to discover bugs when Apple releases iOS 16

Organizations of all sizes are likely to see a spike in product quality issues in their applications if the debut of iOS 16 in the coming days is anything like what happened when Apple released iOS 15 a year ago. In every industry unitQ examined — from business and productivity, to dating and social networking — complaints from mobile app users that the updates broke core features increased dramatically, according to a unitQ analysis of App Store reviews on more than 4,400 iOS apps.

Missing notifications in dating apps spiked 114% from the month before to the month after iOS 15 was released on September 20, 2021. For the same time period, upload issues in the photo-video industry jumped 132%; the mobile gaming sector saw a 300% increase in user complaints that their hero “died for no reason;” meanwhile, concerns from users of finance apps that their “data disappeared” jumped 150%. The list goes on.

In the chart below, the percentage increase is shown for major product quality issues that users complained about in Apple App Store reviews from the month before to the month after the release of iOS 15, by selected category:

*For this chart, unitQ surfaced product quality issues by ingesting Apple App Store reviews of more than 4,400 mobile applications. We then parsed those reviews with machine learning and proprietary unitQ algorithms to produce these results.

These destructive quality-of-life or customer-churning issues highlighted above occurred despite the release of public and developer betas ahead of the iOS 15 software update — and new betas are now being tested by developers and the public before the release of iOS 16.

All of which means the rollout of iOS 16 will likely cause headaches for end users and, by default, developers — despite their best intentions. But how long developers and their organizations allow product quality issues to linger in their apps is another story — a story whose ending depends on how and if organizations listen to what their end users are saying about them. 

Understand what your end users are saying

Organizations are throwing tremendous amounts of resources at building and maintaining their tech stacks. There are systems for auditing, for monitoring security and the performance of applications, and for microsystems and infrastructure as a whole.

CIOs and other company executives are providing these and a myriad of other tools to support their organizations’ continued success. However, there’s a goldmine of other data containing a wealth of actionable insights.

That data is the “Voice of the Customer.” This data is the single source of truth of product quality, complete with actionable insights with precise information about what users are saying about your product’s performance, features and capabilities. 

From reviews in the App Store or on Google Play, customer support tickets in Zendesk, incident response tickets in Jira, to social media and discussions on Discord, Reddit and Twitter — it’s possible to paint a clear picture of what features your customers enjoy, what they find hard to use or broken, and what bugs should be prioritized.

Detecting, investigating, prioritizing and fixing issues identified by users bolsters developer KPIs, enhances customer satisfaction and ratings in app stores, attracts new users and drives new revenue streams.

Parse user feedback data manually at your own risk

We all know that running a successful business relies on keeping customers happy. One way to keep them happy is to fix product quality issues that they identify. But many companies manually process customer reviews left on app stores, social media, Reddit, Twitter, Discord and you name it. The manual process to parse user feedback, sentiment, and product-related issues leaves the door open for much error, doesn’t account for languages, and takes way too long. 

And that’s a problem, especially when we all know that a brand new iPhone software version number update is going to cause havoc. To prevent a downturn in app store ratings, for example, why not discover these issues immediately, instead of in weeks or months?

How much of an effect do app store ratings have on your business? A lot more than you may think. In fact, half of users polled won’t do business with a company unless it has at least a four-star rating, and nearly three quarters of users won’t buy a product before they’ve read the reviews. 

When users’ experience with your app doesn’t meet their expectations, they tend to let you know right away, by filing more tickets for your support team and submitting lower app ratings and negative reviews — all of which damage your brand. When support tickets and negative reviews pile up and your app store ratings drop, your reputation suffers. New users are less likely to try your product, your existing users become less engaged, and churn results. 

In today’s crowded app landscape, quality is more than just something that’s nice to have. It’s a true market differentiator that can help you scale efficiently and beat your competition.

Embrace unitQ data intelligence  

unitQ Monitor provides organizations with the real-time insights and actionable intelligence necessary to identify and fix those product quality issues that are the top concerns for your users. 

unitQ captures user feedback in more than 70 languages from dozens of sources — including the Apple App Store, Discord, Google Play Store, Reddit, Twitter, Facebook, Instagram, and Youtube, among others. unitQ also integrates with productivity tools such as Slack, PagerDuty, Zendesk and Jira. 

With machine learning and AI, user feedback is bucketed into discernable product quality issues. unitQ customers like Spotify, Klarna, HelloFresh and Udemy harness this data to visualize how existing and new product features are impacting their users across regions, app versions, and operating systems — in real time. unitQ algorithms enable organizations to harness this real-time user feedback to take a measured, data-driven approach to their product quality efforts, fix issues faster, and leverage insights into roadmaps and product growth.

About the data for this study

For the purposes of this study, our charts are based on the ingestion of tens of thousands of iOS App Store reviews from more than 4,400 apps for the time period of one month before to one month after Apple’s release of iOS 15 on September 20, 2021. 

Overall, as illustrated in the chart below, social networking apps saw the biggest increase in users reporting product quality issues after Apple updated to iOS 15 last year. Following social media was the photo-video sector, gaming and so on.

The chart below shows the percentage increase in product quality issues from the month before to the month after the release of iOS 15, by selected industry:

While finance, travel and health and fitness apps did not see an overall increase in product quality issues identified by App Store reviews, they nevertheless succumbed to a steep increase in quality-of-life bugs, according to unitQ data and illustrated in the first chart found at the top of this report. 

Health and fitness apps, for example, saw a 462% increase in users complaining they couldn’t edit their body weight. Travel apps saw a 300% increase in users upset they couldn’t log in through their employer’s SSO. And there was a 67% increase in users worried they couldn’t make a “trade” in their finance apps after the update to iOS 15.

About unitQ

unitQ is an AI-enabled platform that listens to signals from an organization’s user base. unitQ is arming organizations with real-time actionable insights to improve product, reduce churn, boost star ratings and build great app experiences.

To learn more about these research findings or about unitQ, request a unitQ demo.

David Kravets is the unitQ Senior Content Marketing Manager

Customer Stories

Learn how HelloFresh improves product quality

HelloFresh, the world’s leading meal-kit delivery provider, is on a mission to change the way people eat. Founded in Berlin in 2011, the company delivers high-quality meal kits through a flexible, customizable subscription program. Its combination of affordability and convenience has turned it into a global powerhouse delivering 1 billion meals to more than 7 million active customers in 2021.

Running a successful subscription-based business relies on keeping customers happy, especially when it comes to dinner. For HelloFresh customers, having access to over 50 meal options each week satiates their hunger for meal customization — yet also brings more potential for a user experience glitch to drive them away. 

The company’s digital and physical product teams looked to user feedback for insights on delivery issues, cancellations, and user sentiment, but the process was entirely manual. With roughly 80,000 points of unique customer feedback flowing in each month from chat and phone support, surveys, social media channels, and app store reviews, HelloFresh was overwhelmed by so much customer feedback. HelloFresh also didn’t have the time or resources to manually turn that flow of user feedback into data-driven directions or prioritizations.

“We really care about what makes our customers happy, or unhappy, from the moment they convert to ongoing delivery, meal prep, loyalty rewards, and more. But our process for gathering user feedback relied on manual effort so we couldn’t quickly see sentiment, recognize trending issues, or track quality issues over time. That’s why we embraced unitQ.”

Jake Zukowski, Vice President, Global UX Research & Design at HelloFresh.

Continue reading the HelloFresh story on the unitQ Customer page.