AppSight™ Introduction
We’re kicking off a new video series focused on AppSight™, our AI-powered app analysis tool.
Watch below to find out what generated this idea and how our team brought it to life.
Long Live UX!
00:00:01:16 - 00:00:23:14
UX is dead. Long live UX! The advent of generative AI and the hyperventilating around the end of user experience design seems like a good time to talk about AppSight™, our internal framework for AI driven quantitative user research.
Our Process
00:00:23:16 - 00:00:50:10
So let's talk about UX research and the evolution of the post-app launch qualitative user study. So we've had this process that we've developed, I don’t know, over the last 15 years that usually has a certain amount of UX work on the front end, where we're trying to understand the customer requirements and user needs and all that stuff. And then towards the end of the process, when there's enough of the app running where you can either launch it as an MVP or you can release it to a closed audience.
Qualitative User Studies
00:00:50:11 - 00:01:10:22
We would then run usually some kind of close qualitative, user test or user research or user study, whatever you want to call it. And that's very time consuming. And it takes, a certain amount of, you know, expertise. I'll say, I mean, there's some subtleties to it to do a good job with it. You're doing a lot of interviews.
Roadmaps
00:01:10:22 - 00:01:32:03
You're recording both screens and you're recording user expressions, user interactions and whatnot. You're creating highlight reels, you're analyzing, and then you're kind of reporting maybe back to a client with a sort of highlight reel. You know, we saw this. We observe that, that kind of thing to make last minute corrections to the app and or to adjust their future release roadmap.
The User Journey
00:01:32:05 - 00:01:57:02
So we had this idea that, what if we were to design a data model around the key user stories from the app design process? We call them app stories. They're like the 4 or 5 like main success scenarios, if you will, that are important for this app. And what if we were to instrument the code a little bit, to drop breadcrumbs into a database somewhere on the back end to map the user journey?
Data Sync Patterns
00:01:57:05 - 00:02:20:18
And then what if we were to replicate a lot of our favorite data sync patterns, between the mobile apps in the back end services, we're kind of big fans of Google Firebase, Google Firestore or Google Cloud Platform. And for years now we've been building a bunch of apps with using that, mechanism to sync data from backend services, up onto the apps, etc..
Training Time
00:02:20:20 - 00:02:44:01
And we could just reverse that. Right? We'll collect some data in a local database. And then when it's, you know, when's the right time, we'll send it back. And then what if we were to take this new Firebase, Firestore and connect it to Google Gemini AI and train it on a data model that's related to the, the app stories.
The Results
00:02:44:03 - 00:03:00:07
And it was awesome. And the results were surprising. I mean, they were jaw dropping, to be honest with you. We it took us a while to kind of get all this hooked up and get it all working right. And we've been playing around with this internally for months. And we looked at this and went, wow, that's really interesting.
App Stories
00:03:00:09 - 00:03:22:03
And then we found that not only could we get summary information of every user session, we could ask for specifics. And then we had summaries of summaries if you will. So you could see in broad sort of in broad brush strokes, you know, how is, you know, app story one doing, how is app story two doing, what percentage of users are completing that app story.
Going Live
00:03:22:11 - 00:03:39:01
What percentage of users are going into app story three and then maybe bouncing back out. And why is that? You know, and that sort of thing. And so that's what we did. And so then the next step for us was we went live with, a real customer a week ago or two weeks ago, I guess, now.
What’s Next
00:03:39:04 - 00:04:12:12
And we, instrumented a code for the Parade of Homes, the Minnesota Parade of Homes, many thousands of users on that. And, we made sure that there was no impact to using the app and that sort of thing. And there wasn't. It's very lightweight. And, so we've been collecting lots and lots of data now. And so what we're going to do here in, the weeks ahead is we're going to kind of create we're going to create a series of videos where we're going to kind of walk through kind of in a very transparent way, kind of how we built, you know, what drove this, how we built this.
App Story Concepts
00:04:12:12 - 00:04:30:04
So we're going to talk about the app story concepts that we use here internally to design mobile apps and web based apps, and how that fit into this generative AI model. And we're going to bring our director of design, Rachelle, in to join us on these videos and talk about that. We're going to then talk about how we design the app dashboard.
The Dashboard & Back-End
00:04:30:04 - 00:04:54:16
And we'll bring a couple of members of the design team in to talk about the AppSight™ dashboard and what value that brings to our clients. And we're going to talk about the AppSight™ generative AI. And back end will bring, a couple of the developers in who have been working on that to talk about the software stack that they use and some things they learned along the way to connect all this stuff together, and then we'll kind of bring it all, you know, in for a landing at the end.
More to Come
00:04:54:16 - 00:05:12:03
We'll bring Rachel back in, who is the lead designer for the new homes project for us and kind of, you know, show you the end result of a real world example of doing this and kind of talk about what this looks like for next steps. So, like us follow, follow us. Look for the upcoming videos and remember long live UX.
End Note
Follow us on LinkedIn to see the next article in your feed.
Ken Krutsch is Founder & President of KRUTSCH, a digital product design firm. From concept to delivery, KRUTSCH specializes in designing consumer and commercial applications. We generate and execute ideas, finding opportunities for our clients to innovate. Because solving the right problem builds careers, organizations, and professional relationships.
Follow KRUTSCH on LinkedIn to read the follow-up posts in the series.
