The truth about app development and optimization with Juliana Jackson
AI-Generated Summary
Media.Monk's Juliana Jackson shares the truth about app development, optimization, and how web and app should work together.
Audio
AI-Generated Transcript
Juliana Jackson 0:00
Come on LinkedIn and say yes, but the motivation is not if the motivation is high and the friction is low, or if the motivation is that much the friction doesn't matter. It does.
Rommil Santiago 0:10
Hi, my name is Rommil Santiago, and I'm the founder of experiment nation. On today's episode, Gerda interviews, media amongst Giuliana Jackson, they speak about the hard truths about working and optimizing for mobile, how web and apps should work together, and how much it costs to build a mobile app and all sorts of other things. We hope you enjoyed the episode.
Gerda Vogt-Thomas 0:32
Thanks for joining everyone watching this. I'm geta. I'm here today with Giuliana who is an absolute AV testing influencer.
Juliana Jackson 0:48
Marketing goddess, I hate to put that on your your, your your you're playing dangerous games because I have no field. What's your assignment, maybe this thing in less time.
Gerda Vogt-Thomas 1:02
And we're going to talk about mobile apps. And I wanted to do this because my own experience as an optimizer is mostly on websites and web apps. Yeah. And you are obviously now super into apps, mobile apps and talking a lot about that tracking strategy, everything. And you recently did this blog article that was really good about the team fragmentation and how even within mobile app teams like people can get it together to kind of work together and all that. And I think that if you take a step back with larger companies who have web experiences and apps, it's even worse, right? Like the web teams are so far removed from what is happening within apps and vice versa. So I thought it would be interesting to kind of ask you questions as a web optimizer about how apps actually work and just basically ask a lot of dumb questions about how to bridge the gap there. But okay. Okay, cool. So in case somebody doesn't know you can your intro like, can you introduce yourself? How did you get into apps and what it is that you do these days?
Juliana Jackson 2:21
Are they find new hope? To answer to that question, I will be in probably in a very good space of my life, Amelia and I just do stuff on the internet. It's kind of what I, I don't like a lot of people might think because I post on social media that I'm like this super extroverted person. I'm so not like everyone that kind of like knows me. Like in person or like spend some time with me knows that I'm like, very introverted. So like, when people ask me to talk about myself, I'm like, but no, I just I actually, if there's one thing that I can say about myself, I'm a salesperson, bro. Like you. I'm honest, like I started in sales. That's how I, you know, I think we met when I was at Omniconvert, I met you and Ryan. And I was sales vendor side. Like for the best part of my career, I started in sales. But I kind of like bridge sales with product back in 2016, when I had my first product job as a product owner for cluster CSR, Romanian software company, so awesome. And that was like, kind of the first time that I had to optimize it. I didn't even I wasn't even using the word optimization in 2014, it was just starting to be a thing, thanks to smarter people. So I was always kind of trying to figure out how back then I was saying, How can you make it easy for people on the internet, you know, very simple, very pragmatic. And I've been in product for a long time, going from marketing to sales, Product Owner manager, like, um, oh, salesperson, a product person, that's kind of like the thing that I did the most. But when I was at CSL in 2020 ish, something like that I took all the courses because what can you do you know, you're sick so you have to take the courses. And I remember pipe was telling me all the time that I should take the Technical Marketing me degree, I actually blame pep for a lot of stuff like good stuff. And I ended up doing the technical marketing me degree, which was one of my favorites. I didn't like the CRO one that much. Sorry, I love some horses from it. But it's just like for me, I always never I never wanted to call myself a CRO and Romeo and she won't know about this stuff. I don't think I'm a CRO, personally, I'm a product person that has, you know, a lot of experience in different things. So I'm like CRO Jason. But yeah, when I was a CSO, I took the Technical Marketing mean degree. And I remember cmo saying that you need to join the conversation, you need to know how to speak with developers. But also you need to not speak with marketers and data people and so on. So it was like, I did everything. But I don't know how to implement data collection, I got into this panic mode that I needed no. But I was like, I need to do something to learn how to, you know, set up a tag on a website, I need to know, what is the tracking code? What's a web browser like? And you know, I wrote the technical marketing media degree afterwards myself. So I left CXO. And I said, I'm going to become a digital journalist, which I did. In a way I like that's I ended up with media monks. And I did it for six months. And then I was like, you know, this is cool. But can we just like focus on making it easier for people on the internet? Like, I kind of get back to the same predicament that was in 2016. So that's a very long winded answer to tell you that I don't know what I'm doing yet. But I am a salesperson. I'm a product person that has experience with data and optimization and different things. But yeah, I don't consider myself an AV testing. Not at all, like I don't, if I cannot not touch AV testing, I gladly will do it.
Gerda Vogt-Thomas 6:13
But are you exclusively in mobile apps now? Or are you still kind of dabble with all this other stuff?
Juliana Jackson 6:19
It's actually cool, because nobody asked me this stuff. So when how I ended up with mobile apps. Because I'm a product person. In my agency, we have two pillars is we actually have like 10 pillars, but we have two pillars that work together as platforms, which build apps and build websites. And like we are the devs. And then we have another pillar called data. In the data team. I have a small startup in this one 1000 people, which is you have a small startup, which is called digital experience optimization. I don't name it CRO or experimentation, because you know, like, everyone likes to name it differently.
Gerda Vogt-Thomas 6:57
was dead anyways,
Juliana Jackson 6:58
is that yeah, exactly. This cure Oh, now what this is Rama
Rommil Santiago 7:03
Santiago from experiment nation. Every week we share interviews with and conference sessions by our favorite conversion rate optimizers from around the world. So if you liked this video, smash that like button and consider subscribing, it helps us a bunch. Now back to the episode, Jamie
Juliana Jackson 7:16
the lungs, I was put on a client that is a retailer and also like a restaurant chain that has web app. And as an excited person that read case studies about this, you know, this specific brand, I was like, Oh my God, you know I did. But it's a very big difference from the US part of the brand, and the one in Europe, and how they're building stuff. So we were building their apps, we were building their website. So I ended up on this plane, we were just I was just doing J for implementation, like very basic. And time passed. And obviously, knowing how I am very curious, I said like, well, you know, this, this part of the customer journey in the app doesn't work that well, because I was testing actually. So when you test tracking on the app, you have to download the app, the UAV version from the App Center, or whatever you're using. And then you shake your phone to have tweaks. And then you can see in real time how the events are popping on your screen based on what you click. So, you know, I was like, well, this doesn't seem right. Like we should do this, we should do this. So I ended up telling my manager like, Hey, can we do a UX evaluation of the app? She was like, Yeah, sure. So I do the UX evaluation. I was like, Can we do also some, you know, user testing and review analysis? Can we do that? So from that, it ended up being like a whole practice. So I started working with other clients with mobile apps, like I don't work 100% exclusively with mobile labs. But because I have product experience, I kind of like gravitated towards those. Because I'm more pragmatic, like I'm not the type of person that is going to come in nearly two because the I don't know the CRM or DMD is not something like I couldn't give less. I just want us to have the mindset of testing and to do better for the clients. So sometimes for the users and the clients. And sometimes that doesn't, I guess, I don't know show as how good you are with Matt, Matt. So that's another long winded answer. Like you should, you should have known that when you asked me to do this, I will give
Gerda Vogt-Thomas 9:27
so no, I love it. Yeah, good. I'm
Juliana Jackson 9:29
also like unfilter because it's late for me. So maybe you're not but yeah, so I mostly exclusively what I was by also the web. I have a lot of like ecommerce because you know, I I love ecommerce and I worked in this space.
Gerda Vogt-Thomas 9:47
I remember when you used to call yourself the CLB lady. I
Juliana Jackson 9:51
wasn't calling myself people were calling me that and then I took it because it was cool. It was better than like some others. Find it brace that we should bring this, you're really
Gerda Vogt-Thomas 10:02
the resurrection of this young lady got nice 2020 24? So do you have any, like favorite industries or type of apps that you like to work with? Either categories?
Juliana Jackson 10:19
I like For that select? You know why? Because it's like, it's when you're ordering food is very primary needs that you have, like, there's no need for all the I mean, yes, you can make some interesting experience. But like, it's a very, I posted this on LinkedIn A while ago, if your app says that you're showing me parking spots around me, I want to open the app and find my parking spot. Like, I feel like food apps, because it's a primary need of feeding yourself, right? It's not too much. You don't have outcomes you don't do just to be done to feed yourself, right. The core experience is very important. And I am not too much of a big fan of new feature development when it comes to mobile apps, because the core experience is much more important. Like if you're selling that you're doing something your app needs to do that stuff. So I love to optimize for that amazing core experience. So I like food apps. I also like gaming apps, but with gaming, it's a whole different like game food from like food is very primal, primal primal needs, but gaming. Man like I remember we had that.
Gerda Vogt-Thomas 11:29
It's also primal, just getting dug down basically know what do you know
Juliana Jackson 11:33
how many data points are in a game? Like I had, we had a client that is a cookie, besides the other things, and they had the game. And one of my colleagues there was supposed to tag it. I was like he was telling me, You Can Do you know how many journeys there are so amazing. So I do like gaming apps. I want to test on dating apps as well. I think dating apps are more interesting in terms of what you can do from an experimentation perspective. But yeah, like I primarily, like very basic stuff that are focused on court experience. Because again, like, if there's one thing that I can say that I am is I'm very pragmatic. So yeah, I don't know if fast food apps, probably also because you know, it's not easy to look like this. Like, you want to put food in there.
Gerda Vogt-Thomas 12:27
Food is so good. Oh, yeah. But, but like that was one of my tracking questions for you actually, like, let's say the gaming gaming apps compared to food apps, right. So there's no in app purchase? And you just said that there's so many touchpoints and journeys. There's a lot
Juliana Jackson 12:44
of female purchases in games. Oh my god. Depends on the game, but okay, I'll shut up. Okay.
Gerda Vogt-Thomas 12:51
Okay, but let's say an app that doesn't have an in app purchase. And it's mostly about engagement. So how would you approach like tracking that or the success of the app?
Juliana Jackson 13:02
Yeah. So obviously, it's gonna be engagement metrics. user engagement open, like, basic thing is like this, you'll download the app, that's a metric app downloads, you remove the app, which means deleting it. But it also doesn't mean uninstall it like two different things. Then you have obviously daily active users, you know, very product metrics. They live active users, monthly active users, average revenue per user, like very simple product, like if it's exactly like a software company, if like you would. So SAS is kind of the same principle, even if you don't have mobile order and payment features in the app. So you're focused on engagement, you're focused on functionality, error prevention, or how do you recover from error error is like, so underrated. Like, that's one of like, the most prolific data points in an app errors, like the messaging from the errors, the recovery? Or what happens after people see specific things in the app? Like, do they adopt them? Do they turn them off? So kind of like, this things should be in all apps, I think in all apps, you should look at engagement and purchase is kind of like one of those, you know, like, if you do this part good, is back to that core experience, then people might purchase or, you know, refer a friend or do other, you know, engagement that you can put in an app. Yeah. Yeah.
Gerda Vogt-Thomas 14:27
Makes sense. I read hacking growth recently. And this is like a marketing book from 2017, like seven years ago, and obviously seven years is like eons in digital marketing. Yeah. But it's, it's said that, at the time, at least Apple users need to opt in to receive notifications, but Android users are opted in by default. Is that like, still true? And how does the whole like opting into notifications even influence doing experiments and all that kind of stuff?
Juliana Jackson 14:58
Yes, it does. So basically, We hear Apple is, you know, has their own approach to privacy. And they are, in their own way, their own interpretation of privacy, right. So they're trying to give the user control over how they want to be engaged with, right. So if you want to, you know, be notified by certain things, you have to opt in for those things. So Apple is kind of like doing it in its own way. On Android is a bit different. I have an Android phone like for me, I don't mind it. But I do turn off notifications. I know, yeah, that was not the end of the world. But Apple does their own thing and has their own approach. And they're always like, very user centric, when it comes to anything. And of course, that poses a lot of that my hand is very big. That poses a lot of issues for tracking. And for instance, I remember one of these releases, maybe it was 14, or 15, one of the releases where you couldn't, I think it was 14, five, when you couldn't track too much anymore, if people opted in completely opted out completely. You couldn't see that data. So you were missing, like 25 35% of the data from the Apple store of downloads, and I don't know one installs and sales. So what we did at that point, we actually try to estimate how much 35% would be to have at least some sort of direction or not imagine that people are smarter than that. But that's how we we dealt with. And yes, it can impact of course, a B testing and personalization, because it's not that the audience is smaller, but it might be not representative necessarily. Because if you don't want notifications is clearly you know, that you don't can be clipped or are swept. So I would track if people turn on notifications, like that's an interesting data point, because that shows more interest and intent and stuff like that. But it does Yeah, affect, of course, the sample size. But again, that's not my biggest problem. My biggest problem is the representative because you can have a lot of data and, you know, is it useful? Yeah,
Gerda Vogt-Thomas 17:19
it's kind of interesting that they haven't changed that. And past like, seven years.
Juliana Jackson 17:24
I know they didn't vote, I might be wrong, I should have researched this, but I didn't have time. I might be wrong. But as far as I know, it's you have to like when you install anything you have, you can choose the way you want to be bothered or not with it. But again, like this is very on brand for them. And everything that they're doing is to keep people locked into their ecosystem. And they have kind of their own ways to interpret privacy and engagement and stuff like that, like even now, for instance, with PWA. Initially, they said that PWA is progressive web apps, which are like, an, they're like a nap type that you can download on your homescreen. And you can use it but it's actually a web page. It's nothing, you know, a native app. So they were saying initially, because of the digital marketing act, you know, they will stop the PWA in Europe, because for PW what the MA wanted was basically to let developers have up who to work with other browser engines like brave or chrome and so on. And that's disaster for Apple, because they want people to use WebKit and Safari, the browser engine, because they want to keep people in their ecosystem is not because they're like evil giants is because it's very expensive and very risky. To allow people to come with make believe, you know, like apps that work on I don't know, opera, our brave or whatnot, and put them in the ecosystem where they can put at risk the security of their users. So as far as I know, this didn't change, because it's very on brand for with the notification is very on brand for Apple, but who knows. I didn't have time to check to be honest at this change.
Gerda Vogt-Thomas 19:06
Okay, but if you look at like iOS and Android experiences, in your work, like do you see like massive differences in the user journey? Or how do you even like approach the strategy and research? Do you have to do like, two sets basically? Or how do you approach that?
Juliana Jackson 19:25
It's such a good question. That experience when I always sound right is very different because the users of Android and iOS are very different. They're different types of people. So what what we did on a client was actually do research in surveys and interviews to understand what's the Android user versus the iOS users in the context of that client. Like we actually did surveys and interviews. And basically what we noticed is that people that are on iOS are mostly funny enough. They were younger people. They were people that were much more focused on life. style and fashion and photography like more. I don't know, whimsical people that we found those types of people also on Android, but people on Android are kind of like more focused on functionality focused on making sure things are working fine. Which is funny, because the reviews and ratings on Android or worse than the ones on iOS, I don't know, how is that possible. But basically, on iOS, people were rating the app higher. And then on Android, it was lower all the time. Because this one app that I'm thinking about the experience was a bit different. One fun analysis with it is like when you purchase something through the through the app, you can customize it. So customization is one of the most, this is one of the biggest levers to mobile weather and payment, because it shows obviously intent, but it also shows that, you know, you want to actually go to the end and to the to the purchase. But the funny part is on iOS, it was taking people between eight and 15 screens to customize that specific product. And on Android, there was just like, five to eight. And we were asking like, what, why, but then we realized that people are an iOS, what am I more undecisive? Because the experience was different. So we were still working to do some maybe testing because then is when you actually want to a B test. Like that's an interesting challenge like to do it's not kind of like what I do appreciate about AP and I want to say this here, record that then on the record is that I do appreciate that. People in the teams are not as precious, precious. I what I do like about mobile apps is that people in teams and developers and you know, the whole squad the product, are not as precious about being so to the tee with statistical significance and calculating that you take, it's mostly about your prototype a future, you flag it you feature for feature flags, is doesn't have crushes, it doesn't have rage clicks, it doesn't have you know, like, uninstalls. That's enough. So like, we're looking more at signals, then if there's one or last, if that makes sense. Like, it's it's kind of like it's exactly like in SAS, right. You're focused mostly on do we did we correct the experience in a positive way? Do we? Because we don't look at the revenue. We don't look at cov I know like you know, Hartford Giuliana say that, like we don't have those type of goals, the goals that we have, which I read. That's why I really like the space is mostly like the, you know, that people uninstalled the app? Plus, do we have such a big drop off? Since we fix the onboarding? Or since we fixed the customization? Then people are like, no actually does better? Oh, well, then let's open the feature in another market. Let's open it in the monitor market. Let's assess. The thing is that to do a B testing in mobile apps, you need an app release for each test. That's a lot of technical. Yeah, I
Gerda Vogt-Thomas 23:07
actually wanted to ask you about that. Because you just posted about like, if you do any implementation changes, you need to basically release a new version of the app, right? Yes. Like
Juliana Jackson 23:18
developers? That's crazy.
Gerda Vogt-Thomas 23:21
Yeah, right. And then but you said it's like not always necessary or something.
Juliana Jackson 23:27
Lacking? No, yeah, for tracking is not always necessary. Because you can still do some things. If it's a PWA, or a web app, or hybrid app, there's, there might be some things you can do in the tracking tool in the interface, G for amplitude, whatever you're using. But to do like actual tracking measurement stuff, you need to do it in the app code base. So developers need to do it, you write that nicely, the measurement plan, and then you give it to the people to do it. So for AV testing is the same like for each AB test that you do, you need to do an app release and there is feature flagging, or you do also a pre and post. So you enable the feature, like not everyone is doing like show like to the t a b testing. Like it's foreign. Like the way you do it on web when you have to see Oh, my God, how many weeks do I need to run it? Well, I look at 80% confidence or 85 or 90 like that. I can I'm sure if people will listen to cycle. No, that's that's that's a catastrophe. That's blasphemy. I use might calculate. Calculate. No, in the real world. You don't go. It's basically again, like the goal. I'm not saying all teams are like that, but what I've experienced, this is me, right? I'm not speaking for people that work at Yeah, I'm gonna shut up but that's my experience, and I stick to it. People that work in this space are more interested in functionality, experience, making sure that it doesn't Crash making sure you don't have low performance because performance is a bit different for engineers and developers than it is for us. And marketers like they couldn't give a about, I don't know, some sort of campaign and promotion, they care if the user can log in log out can be in the app, the app doesn't crash, you don't have bugs, like the battles that you fight are a bit different. So I found it very hard to sell experimentation and like beautifying things and, you know, random things that you would do in Web. Because it doesn't matter for them that much their
Gerda Vogt-Thomas 25:38
biggest. So it's more like robust or
Juliana Jackson 25:41
is very different. And there is no such, like, when I wrote that article that you mentioned that you participated in. I like Jim Gordon said this, not all silos are built equally. And some silos are actually centers of excellence on their own and not in the classical way of how in the CRO community are defining centers of excellence. Sometimes basically, you just have to let people do what they do. And try to come and support versus bulldoze with, if I would go right now to some of the developers that I work with and tell them about, sorry, but we need to run this for nine weeks, this feature needs to be in the market for nine weeks to get confidence and to get power to the test. And they will be like, what? Why? And then it's like, well, because that's how, you know, people on LinkedIn, right? That we should run this. Like I cannot be? How can I beat maybe this thing influencer this year? If I don't do that? I'm sorry, I'm gonna force you. Why did you invite me This is why you invited? No, that's sorry. Sorry, Romeo, you can delete this whole part. So
Gerda Vogt-Thomas 26:52
one of the differences in terms of teams are team like structure? That seems to me and correct me if I'm wrong, but like in web, you know, you have web developers, and they very rarely do like the tracking implementation, like sometimes they help or whatever. But usually, we CROs kind of put together the strategy for tracking and implemented as well. Yeah. So in mobile, it seems that there isn't like that type of position. Like the developers are the main people that do that work, right?
Juliana Jackson 27:22
Yes. But they don't know how to do it unless you told them how to. So it's a strategy is still you need you need a dagger, the thing that connects everything together, like you do need it. Of course, they don't know, what's the latest, you know, best practices on a j for data layer, or some stuff like that. So you need to build this measurement plans. Specifically, though, don't like once the event was the purpose of the event, where should it fire? Why should they fire? What are the parameters? What's the value of the parameters, you do a measurement plan? You give it to them? And then they write it in the codebase. And that's how it's done. So you can, again, there's some tools and configurations. And obviously, it's It depends, right? But primarily, you do need the developer to do tracking in native native mobile apps, because it's a difference. If you do web apps is the same. You can do it too, I can do it. Maybe it's not the best idea for me to do it. I'm looking for my water. I'm sorry, I'm still here.
Gerda Vogt-Thomas 28:22
No worries. So I'm going to throw like a case study or use case you from a web experimenter. So let's say I'm working on a website, where one of the main goals is to get people to download the app. And then basically, I'm running tests to on landing pages to get them to do that, and whatever. And then I want to see if those people that converted through the test, like what are they doing in the app? Is that even possible? And how do I combine kind of the web engagement and all that together with the experience in the app? Like, is that now simpler with Firebase in GA four, because like some years ago, it was like, so difficult still,
Juliana Jackson 29:07
I haven't tested that. But if your G for implementation is correct, you're tracking webinar. And you use the audiences in GA four that contain app data to run tests if you're using J four with an A B testing tool, and most people do that by now. So if you can you if you collect user ID and client ID and you have the audience built in J four that you're using for the AV testing, which contains webinar. It's it all depends on how it's kind of like your data quality depends on your data collection mechanism, because you consume more from years ago, and I heard you say that, so that was very sure.
Gerda Vogt-Thomas 29:52
Well, yeah, I wasn't expecting it to be that easy. But yeah,
Juliana Jackson 29:55
like I mean, we were easy, easy. You know how easy it is so huh? There's always problems. Yeah. But in theory, in theory, if you do proper data collection in G for your audiences leaving G for you having J for data from your ads from your Firebase from Search Console, it's kind of the same. If you, you, you, I urge people to collect user ID, it's very important to have user ID and client ID depending on what you're using in your test. Then it's kind of like just another audience in GA For if, of course, it was fine. I feel like you know, it was too short.
Gerda Vogt-Thomas 30:34
But that's over overly optimistic. Yeah,
Juliana Jackson 30:38
I mean, I'm trying, man. It's like the end of the day for me. I'm trying to. Yeah, I'm trying to be cool right now.
Gerda Vogt-Thomas 30:45
Yeah. Nice. Also, so we touched on saws, and what are kind of like the similarities there. But basically, everyone's saying that, you know, the size app development is being sped up because of AI and whatever. Are you seeing that in the mobile app industry as well, that people are releasing stuff faster, or whatever, or the development is still like too different for that?
Juliana Jackson 31:11
I mean, given that in, you know, the development sits in media amongst, you know, I talk to these guys, like, yeah, we use AI for experiences improving the experience improving, you know, the delivery of content, or copy or whatever, but we don't, to actual to the actual aspect of app development writing code. Now. It's still people doing it, thankfully. Okay.
Gerda Vogt-Thomas 31:39
And like, this is probably an impossible question. But like, if somebody wants to build a mobile app, how much does it cost? Like, do you have any ranges? Or how does it compare to like, web development or anything like that, then,
Juliana Jackson 31:53
bro? Yeah, I mean, it depends, right? Do you want to build an app, it's like, let's talk about the politics, let's imagine a Chrome extension, you can build a Chrome extension on your own, and if you know how to code or you can figure it out. So with apps, for instance, if you build an app that just shows you the weather, you need the interface, you need the I don't know, I pick the location type of thing to track, you know, the weather, some privacy stuff, and you can probably do it for I don't know, 20k 15k, if it's a basic, basic app. But if you want to build a video game for the ordering app, if you want to build the Uber app, or booking app or whatever. Then clearly, you know, you need more developers you need in developer resources, very expensive, as you know. So you need developers, you need UI UX, you need data, it's obviously but if you just if you just want to build a basic app. I mean, you also have to think about how you're gonna host your app into App Store and play store because you know, it's not free.
Gerda Vogt-Thomas 33:06
Yeah, let's talk about the app stores. Yeah. So basically, you have to pay them for your in app purchases. My amazing googling skills, said that it's like 30%.
Juliana Jackson 33:19
Yeah, I make sure yeah.
Gerda Vogt-Thomas 33:23
And so does that only apply to like, digital products? Or like, let's say you have like an E commerce app? Do they then take something from those purchases? Or how does that work?
Juliana Jackson 33:34
It's for digital services and for goods, per se two, it's, it's very similar to the Shopify model. So when you build an app on Shopify store, you need to pay I think, when I did it last time with the reveal, when I was at Valentin at Omni convert, it was 30% that you need to pay. So a lot of people not as, not as we were behaving, but a lot of people create like to try to game the system would build like a proxy to send you to another website. So the transaction doesn't happen in that context. So well, that's not illegal anymore. With Shopify. I checked some months ago, because
Gerda Vogt-Thomas 34:17
that because the first kind of example that came to my mind was Kindle, where, you know, they show you the books in the app, but you can't buy them like they make you go
Juliana Jackson 34:30
on Amazon. I hate it. I don't know how they're doing this. Because like, I cannot read any more books like at all like, you know, I read on my phone because of my ADHD. I have no patience. And I hate it when they did that. Like two years ago, probably they started doing that because in the beginning, you could buy the phones from the app like I bought so many, even now like I have books that I bought because I haven't bought it ever since. So I think the last one I was buying was which is mean don't Don't ask what humains I read that, like, he's no Rick Ross's biography. That's the last book I bought. Nice. Yeah, no, I'm weird. Did you expect that? Probably not? No, that
Gerda Vogt-Thomas 35:14
was sad surprise.
Juliana Jackson 35:16
So, yeah, there you go. I have even the Britney Spears Memorial.
Gerda Vogt-Thomas 35:22
Oh, did you read that already?
Juliana Jackson 35:24
I have Elon Musk's. Why No, I have the sample. Because what my behavior, the reason why I'm opening the app to show you is that you can only get the sample on the phone. Right? So I read that I get satisfied, because it's a lot. So I don't go and buy it. Because like nobody trying to go to the website to do the purchase, I want to do it on my phone. So they might think I wouldn't recommend doing this from any point of view. Like yes, of course, it sucks when you give 25% or 15% or something like that. I what I know, because I asked people is that now there's like, how do you say it in English? I forgot. Oh, sure. It's like you can make a certain amount of money. Oh, it's kept. That's the word. It's a calf right now on how much you make. And if you're under a certain like threshold, you don't have to pay 30% is less or something like that. Like there were some adjustments and actually asked one of the developers to prepare for this interview. And he told me this. And then I said, Can you give me a link and he was like, I don't have time, but I still love them. But it's I know that the one of the developers said that they made these changes, because a lot of people were trying to game the system just like they're doing with Kindle. But I wouldn't recommend doing this. Because if you interrupt the experience that people have on the app, especially if they're reading Come on, bro, like I want to read something I read the sample I'm so excited. Like, yes, for sure people will come on LinkedIn and see us but the motivation is not if the motivation is high and the friction is low. Or if the motivation is that much the friction doesn't matter. It does when I was a CX, so me and Kyle, actually were thinking about the cancellation process. And we tried to simplify it and make it easier because we realized that if people want to come back, they will come back. But if you make it hard for people to cancel their subscription, they're going to be so full of resentment there, they might never come back. So it's kind of like a long term type of loss.
Gerda Vogt-Thomas 37:28
Wanting to go back to the EECOM apps to like really big E comps like Shara a sauce, whatever. They have, you know, regular web mobile experience, but they have the purchasing and apps as well. So do you have any like advice on how big a company has to be in terms of when is it the right time to launch mobile native app alongside your regular purchasing experience or whatever?
Juliana Jackson 37:59
Just look at your traffic sources in if you have 90% of people coming from mobile to your website, and look at what is the revenue from desktop versus mobile? If mobile Trump said it's time for you to you should? Yeah, you should already start and like invest. Yeah, yeah, we're just the PW starts with the PWA. Start with the web app first and test the waters and then build a native app like you don't have to start the native apps are more expensive. But you can build a PWA or a web app or a hybrid app. But I think this is also in general, a lot of traffic to websites recently, from all the websites I've seen and gotten those. I've seen so many websites in the last one year and a half. Most of it is like 70 80% mobile traffic.
Gerda Vogt-Thomas 38:46
Yeah, especially in EECOM. We're seeing the same thing. And oftentimes we don't even like test on desktop. Yeah. Because if 90% is on mobile, it doesn't make sense to invest in testing on desktop. Yes.
Juliana Jackson 39:01
No, I agree with you. And you shouldn't test on web before your visitors or mobile makes no sense. It's like,
Gerda Vogt-Thomas 39:07
but I think it's still like a hard balance. If you're saying that, you know, even a simple app build costs, like 20k. And, you know, if you have a pretty Yeah, and probably for any comments even more, right, so it's like, you know, when is the right time? Or what type of revenue do you have to have? Yeah, I
Juliana Jackson 39:26
mean, of course, if you're just starting out and you're just hitting your million or two millions and in E commerce, that's nothing considering how much it costs to, you know, have the business sell online and stuff? Probably not. But it depends on the goals of the company, how much money you're making, how many customers like traffic, it's a sign conversion on mobile, it's a sign. people accessing your website, mostly from mobile, it's a sign so you will do a bit of research will ask people like hey, You know, what do you think about an app? What do you think about it? So it's a natural extension to, to the experience, because I'm going to share this link with you after like the mobile app report from 2024. And mobile apps is a $6 trillion type of thing.
Gerda Vogt-Thomas 40:18
Yeah, I saw that in your blog post actually. It was like, what? That's crazy number.
Juliana Jackson 40:23
But it's where most people end up the research is so good. Like I've I don't really look at research that much, but that was so well done. And it's, it has like a lot of you know, it's statistically significant, I guess. So like, if everyone uses their phones all the time, like if you you advertise for them to them on Instagram On Facebook And uses Facebook or Instagram on this stuff. My auntie, point my auntie, she's playing. Exactly, but if your market is, I don't know. Also an interesting thing that I want to share here that I learned recently. It's who are your buyers right? Because if most of your buyers are people that are Gen Z, or like close to millennials, the classic advertising and sales channels don't work anymore like right now for a client we're actually building a Roblox because most of their users and potential buyers are actually teenagers 20 year olds that play Roblox and you will think that Roblox is just a game for kids so my son plays Roblox while one of my kids plays Roblox. Oh, there's like grown ass people on Roblox. And Allianz actually had a great one. They built a Roblox game where they were teaching kids how to ensure their goods because in Roblox, you get a lot of like drops in, in game. So then people trade those for other types of items. So a lot of people were scammed in the game. My son too, like he was camp of his items, because he was doing like this back pocket change. Yeah. So I can start, how can I teach younger people younger generation about the importance of insurances? So they created the Roblox game with this, like, you know, super. What do you call them, like, superheroes with superheroes in the game, and it was basically educating kids on how important it is to ensure your goods before you exchange them, or you put them on the internet. And I thought that was really awesome. I was like, Whoa, that's that's so more and more people create endgame experiences, because that's the only channel that they can reach out to the younger generations and my son, like his 12, right, but he sees Roblox we're in fortnight. And he says, I don't know, I want the new Travis Scott, you know, shoes or stuff like that. So that's another like, I feel. And I was taking all responsibility for this is like, yes. You know, like, it's still good to, you know, do the common sense things on webinar app? For sure. You see, I'm not saying best practices, because Yikes. But what are your channels? Like? That's interesting, like, how do you AB test that? How do you push a videogame? And you get people from the video game? Like, how do you, you know, quantify that, like, that's an interesting way. For me like, it's something that's really obsessing me lately. And that's what we're trying right now to do with a client to create like a skin in robots that you can buy and try to educate people. So that's kind of like where my interest is lately, especially if you want to reach certain audiences. So that's also very important. And coming back to your question. Figure out, you know, like, if you make enough if you're profitable enough, and all your traffic and it's coming from mobile, do a bit of research in demography, demographics, also, like interests and stuff like that, to see what are ways where you can reach these people that are not so you know, classic, like, I'm very curious to see like an industry report on desktop versus mobile use for a lot of stuff, because I, and that's why I get the bid. Eilish. When I see best practices for web that showed this stuff.
Gerda Vogt-Thomas 44:22
You know, like, man, like, yeah,
Juliana Jackson 44:24
of course, if you're a bank, for sure, if I'm gonna listen. So that is the mentality for me as a millennial, if I need to do some bank that I cannot do in the bank app, I gotta go on desktop. So when I go on the desktop is when I'm booking flights. When I'm doing some, like important, that actually requires a lot of mental effort, right? I'm not gonna shop on my desktop for shoes or lipstick, or, I don't know, random, but if I need to book a flight, if I need to do some financial, or like an application like that, that is what I'm going to use at this time. Anything else?
Gerda Vogt-Thomas 45:00
Yeah, it's weird. It's like almost desktop as this serious adult bureaucratic live person.
Juliana Jackson 45:07
Yeah, just realize this talking to you like, this stuff is real.
Gerda Vogt-Thomas 45:13
Yeah, totally. Yeah, another, I guess optimization question for you was about the early on triggers, and up user experience and like, I've read that a bunch of companies, when they, you know, start an app, they kind of amp up the triggers in the beginning to show that there's a lot, a lot of engagement, and then that like, puts up the stats, but then that actually ends up backfiring, because people get so annoyed. So it's like, Do you have any advice on how to balance the short term engagement with long term retention and usage and all that kind of stuff? Marketing
Juliana Jackson 45:55
response? It definitely, you know, depends on there. But so like, we can take an example, if it's a food app. Of course, you know, you want to send notifications about sales and promos and the sandwich of the day, like that. And I think those are fine. But if your app is an E commerce app, let's say it's so I don't know, do this. No, not at this. I can talk about, let's say it's a Zara. I can talk about, let's say Zara. Oh, actually, I know what that's annoying about you. So about you. It's a fashion. Bro, they kill me with the notifications. Like, I do love to browse. Because I do buy I'm a heavy purchaser of goats. Like I buy random. I do I have I do retail therapy. But I, you have to like pick your battles with notifications and getting people in the app like, you have to ask yourself like, is this worth stopping this person from their day to share this type of notification, you gotta like, be strategic, I guess with the type of stuff you want to get to get people to open the app, you're also always going to use when you just build an app, you can always you look at first open, which is the first time they open the app after they download it. And then you can look at very basic metrics, like seven days, activity, 14 days, 30 days. And yes, you can focus much I think, if my advice would be to focus more on onboarding people into the app versus trying to constantly get them into the app, because I always go back this, like you want to build a habit. Like if you don't build a habit from the first use of the app, it's very hard to get people to use it again. So like to fashion brands, like, if you build an app, you know, you have an app and you get somebody that downloads just from the first time, show them around, show them where like, you know, you they can go to see promos, and whatnot, like do a nice, you know, like onboarding with, you know, great like screens and show them where it is. And then I would ask, and I actually did this, you ask people like, what do you want to ask to notify your boss, so then people choose what they want to see. So if you kind of like build an experience that the user is in control in the app, which is very important, since it's your phone, like there's nothing more private than your phone, in my opinion, like, you know, it's the most private thing you have, I would say and advise people to let the user have the sensation that they're in control. And let them pick what they want to see. But focus on the onboarding experience. So important. So many people like Oh, under invest into things like change password, or login, you will be surprised how many people bounce in the uninstall apps, because they just cannot even log in into the app, like one login. Or to see your welcome screen. Whereas those you click here to see the paper, click here to see this. Notification. Yeah, we see worried, you know, like, you know,
Gerda Vogt-Thomas 49:07
even like basic kind of flows where even like, on web, if you sign up for something. The like confirmation link is like broken, or something that hasn't been in
Juliana Jackson 49:17
the cognitive dissonance. Yeah.
Gerda Vogt-Thomas 49:23
Yeah, totally. Well, I feel like I could talk you talk to you forever, but don't know probably. Probably doesn't want us to have like a three hour at some point. Oh, man. So yeah. Any last thoughts you want to share about like web and apps working together? What What should people pay attention to? What's the what's the big problem we need to solve?
Juliana Jackson 49:51
I would say that there's no competition between web and app and it shouldn't be it shouldn't be a cohesive journey, because people will use Both types of you know, channels to reach to your brand. I would say that I prefer apps and product in general, because I always felt that then I always felt that in the community, in the experimentation community. Sometimes there's so much focus on literature and science papers and all these best practices things and like I do see their place in time and purpose. But I do appreciate that, in this context of mobile apps and the same product, things are more pragmatic, and they're more like, a lot of people say that they're using all this literature to make better decisions. But actually, it's kind of like gatekeeping a lot of people from trying to at the store trying to run just a feature flag, I feel like if you want to help the industry in the community and push it further, people should get keep plus, and stop being so precious about every little thing, because people get scared and fair enough for because of that. So that's why I felt like I found my place in mobile and product. Because people don't make me feel stupid, because I haven't read some science paper on Google Scholar about statistical significance. So yeah, that's kind of my last words.
Gerda Vogt-Thomas 51:35
Nice. Well, I think, yeah, a lot of CROs and white people also can learn a lot from like how people do things and how to coexist, especially in big organizations. So I really appreciate your time. Thanks for answering all my questions. was fine. Yeah, me too. So good luck. rawmill editing this, and I'm sure we'll talk soon again.
Juliana Jackson 52:02
Would this optimized is the best tool I ever used. Love Optimizely. Yeah, both of us please.
Rommil Santiago 52:10
This is Rommil Santiago from experiment nation. Every week we share interviews with and conference sessions by our favorite conversion rate optimizers from around the world. So if you liked this video, smash that like button and consider subscribing it helps us a bunch. Now back to the episode.
If you liked this post, sign up for Experiment Nation's newsletter to receive more great interviews like this, memes, editorials, and conference sessions in your inbox: https://bit.ly/3HOKCTK