Is the relentless pursuit of measurable ROAS fundamentally at odds with building long-term customer trust in a privacy-first world?
Agility requires marketers to move beyond legacy attribution models and embrace a more dynamic approach to measurement and monetization. This is especially true in the rapidly evolving mobile ecosystem, where the rules of engagement are constantly being rewritten.
Today, we’re going to talk about the new playbook for performance marketing in the mobile app ecosystem. We’ll explore how to drive growth and measure return on ad spend in an environment defined by signal loss, look at the innovative ad formats that are capturing user attention, and discuss the role of AI in balancing automation with creative effectiveness.
To help me discuss this topic, I’d like to welcome, Phoena Pang, Vice President of Sales and Global Partnerships at Mintegral.
About Phoena Pang
Phoena Pang is the Vice President of Sales and Global Partnerships at Mintegral, a leading global mobile advertising platform. Based in the US, Phoena brings deep expertise in mobile advertising, strategic partnerships, and business development to drive Mintegral’s go-to-market growth and operational excellence across global markets.
Phoena has held senior roles at top-tier technology companies including Google, Moloco, Vungle, and Chartboost. At Moloco, she led product go-to-market strategy and partnerships for mobile performance marketing. During her time at Google, she served as Global Product Lead and Strategic Partner Lead, where she spearheaded global gaming ads solutions and scaled high-impact partnerships worldwide. At Mintegral, Phoena focuses on optimizing operational performance and cultivating strong, strategic partnerships with advertisers and publishers, reinforcing Mintegral’s leadership in programmatic mobile advertising solutions.
Phoena Pang on LinkedIn: https://www.linkedin.com/in/phpang/
Resources
Mintegral: https://www.mintegral.com
The Agile Brand podcast is brought to you by TEKsystems. Learn more here: https://aglbrnd.co/r/2868abd8085a9703
We’re proud to be a media partner for #MAICON26 – Oct. 13-15! Learn how AI can power your marketing and business and help you grow smarter. Use code AGILE150 to save! https://aglbrnd.co/r/7fe458ced0f04658
Reach your customers with Reddit. Spend $500 in ad spend, get $500 back in ad credit! Learn more: https://advertalize.com/r/491818c79fb1873f
Don’t miss We Make Future – the International Festival of Innovation in AI, Tech, and Digital Marketing, June 24-26 in Bologna. Learn more: https://aglbrnd.co/r/c80991afff416bb2
The most influential minds in software, AI, and engineering leadership will be at WeAreDevelopers World Congress North America, September 23-25 in San Jose. Learn more: https://aglbrnd.co/r/60a7299222a7bcf1
Enjoyed the show? Tell us more at and give us a rating so others can find the show at: https://aglbrnd.co/r/faaed112fc9887f3
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter and more: https://aglbrnd.co/r/35ded3ccfb6716ba
Check out The Agile Brand Guide website with articles, insights, and Martechipedia, the wiki for marketing technology: https://www.agilebrandguide.com
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content. https://www.missinglink.company
Transcript
[00:01:12] Greg Kihlström: Hi, I’m Greg Kihlström, host of The Agile Brand. And here’s a question
[00:01:15] Greg Kihlström: for you. Is the relentless pursuit of measurable ROAS fundamentally at odds with building long-term customer trust in a privacy first world? Agility requires marketers to move beyond legacy attribution models and embrace a more dynamic approach to measurement and monetization. This is especially true in the rapidly evolving mobile ecosystem, where the rules of engagement are constantly being rewritten. Today we’re going to talk about the new playbook for performance marketing in the mobile app ecosystem. We’re gonna explore how to drive growth and measure return on ad spend in an environment defined by signal loss, look at the innovative ad formats that are capturing user attention, and discuss the role of AI in balancing automation with creative effectiveness.
[00:02:01] Greg Kihlström: To help me discuss this topic, I’d like to welcome Phoena Pang, vice president of sales and global partnerships at Mintegral.
[00:02:47] Greg Kihlström: Phoena, welcome to the show.
[00:02:48] Phoena Pang: Hi, Greg. I’m happy to be on this show with you guys.
[00:02:52] Greg Kihlström: Yeah, looking forward to, to talking about this. Before we dive in though, why don’t you give a little background on yourself and your role at Mintegral.
[00:02:58] Phoena Pang: Sounds good. so hello, I’m Phoena. so as you know, Greg just introduced me, I’m VP of Global Sales and Partnership. so I’m managing, the global sales team, and also the partnership relationship for, Mintegral. and if you don’t know Mintegral, we are like a performance platform, like especially in mobile. so we work with more than 100,000, app developers, helping them to, one, monetize their inventory, and second, to help them to do, like, very effective user acquisition.
[00:03:33] Greg Kihlström: Great. Great. Well, yeah, let’s, let’s dive in here. And we’re gonna, we’re gonna talk about a few things, but I’m gonna start with one of the things that I highlighted in the intro, and it’s really this idea of redefining performance in a privacy-centric era. So, you know, with the decline of deterministic attribution, a lot of marketers are leaning on probabilistic models, media mix modeling. You know, from your perspective, what’s the most significant mindset shift that leaders need to make when evaluating campaign performance and, and ROAS today?
[00:04:05] Phoena Pang: Yeah. so I’ll say that it would come down to two things. You know, like, like, honestly speaking, I got this question a lot. and also the first thing I will tell them is that, is the acceptance of the new reality.
[00:04:19] Phoena Pang: So I think that a lot of, like, marketing leaders, they need to get comfortable with the confidence intervals rather than the absolute in, like, ab- absolute nu- num- nu- numbers they’re getting from their ad. So, because evaluating the ROAS today, you need to require looking at, like, aggregated trends. and also, you know, the, like, we also, like, you know, mentioned MMN to understand, you know, how the different channels influence each other, because you cannot trace every individual user. and also, the second also I come, I, I come a lot to, this discussion is about the long term incrementality. so instead of, like, reacting to the daily fluctuation, in, like, very subtle data, so a lot of sophisticated, like, marketing
[00:05:04] Phoena Pang: team, so they’re looking at incrementality testing. So, sometimes, you know, people will do, so for example, if I, turn this spend off, like turn this, like, channel off, so would the re- would, like, would the revenue actually disappear or decrease, or for how many percent? I think for those, you know, can also go into, the analysis. so there are … So, you know, like, in summary, so there are, like, you know, different things you need to all, like, going to look at when you’re looking at the new reality for the performance.
[00:05:35] Greg Kihlström: Yeah. And beyond modeling, you know, you’ve, you’ve noted the importance of creative in boosting performance.
[00:05:41] Phoena Pang: Yeah.
[00:05:41] Greg Kihlström: Certainly that’s, that’s an important aspect.
[00:05:43] Phoena Pang: Mm-hmm.
[00:05:43] Greg Kihlström: How are the most sophisticated app marketers using creative testing and optimization as a primary lever to drive results when user level data is, you know, potentially less available?
[00:05:56] Phoena Pang: … yeah. so also as a user level data, they become, like, less available, so the algorithm, so ………………………. like, the machine learning, the black box, so they will have less information to find the perfect user for you.
[00:06:08] Greg Kihlström: Right.
[00:06:08] Phoena Pang: So, then in response to that, so creative has a more important role, so they’re now, like, you know, as rep of the campaign, but they should be, like, the primary targeting. So, a lot of sophisticated ad marketers from what we work with, so they are using, like, the creative ta- like, the- the- the- the- a lot of the creative tests in the three different, specific ways. so I’ll say, like, the first one, so which also the most common we see, is the creative line signal generation, because when you cannot target by specific interests, so your creatives then will do the filtering.
[00:06:40] Greg Kihlström: Yeah.
[00:06:40] Phoena Pang: So, the marketers usually they’re running, like, broad campaigns, so where the visual, you know, hook to find the audience. but if a creative feature, say, like, they’re, like, more, targeting of the hardcore, game strategy, so the algorithms will also learn, like, you know, for them what type of user they’re engaging with these creatives.
[00:07:00] Greg Kihlström: Mm-hmm.
[00:07:00] Phoena Pang: So then to drive, you know, the, the users that, you know, fit to their game and… because of, you know, the creative they’re generating. and also I would say, like, the second big piece is on the, you know, the concept testing. So, basically, you know, rather than testing say, oh, what, like, what button color I’m using or, you know, like, where I’m gonna put, like, this button, that button, so it’s about the macro concept. So you’re looking, like, about a different emotional hook. So for example, you want people to feel, you want people to, do, like, the stress relief, you want people to feel competitive, so you have, like, those type of, like, element, like the big rock testing or we call, like, as a hook to test, like, on different, like, different, like, concept and to attract the
[00:07:45] Phoena Pang: users and to see, you know, how they perform.
[00:07:48] Greg Kihlström: Yeah.
[00:07:48] Phoena Pang: and the third big piece, which is also, like, really big on our platform, is to diversify the ad format. as you know, the performance testing is not only on video, so, we also, like, put a lot of focus on the playable. because the playable actually, they actually provide, like, a very, you know, high intense signals r- rather than, you know, the user, like, they just watch that, so whether they, you know, finish the mini game or whether they replay the mini game, like, a couple times, so this all provide, like, a really good signal to indicate whether the user had, like, a, you know, like, a long LTV.
[00:08:23] Greg Kihlström: Yeah.
[00:08:23] Phoena Pang: So for the… Yeah.
[00:08:25] Greg Kihlström: Well, and, I mean, to that point, you know, it- it does seem like there’s a renewed focus on, things like you just mentioned, you know, in-app advertising or playable ads, rewarded video. You know, what’s driving kind of this- this, either resurgence or just, you know, the- the- kind of this diverse, these diverse methods working and, you know, how do these formats change the value exchange between-
[00:08:50] Phoena Pang: Mm-hmm.
[00:08:50] Greg Kihlström: … user, publisher, and advertiser?
[00:08:52] Phoena Pang: Yeah. Yeah, this is, this is, well, I- I would say, like, in general, when I see, like, I wouldn’t even call, like, it’s a resurgence, because, like, I think, you know, from what I see, the rewarding video has always been a, like, rising trend.
[00:09:06] Greg Kihlström: Mm-hmm.
[00:09:07] Phoena Pang: similar to playable. So playable, like, you know, like, right now, it’s kind of become, like, a norm for people to use.
[00:09:12] Greg Kihlström: Yeah.
[00:09:12] Phoena Pang: and the reason for that is because, there’s, like, a shift toward, like, ROI-centric advertising, and also the rise of the open internet. because, you know, a user acquisition, the, you know, on the traditional, like, social platform, so usually, like, you just have, like, one static, format, versus, you know, for, like, you know, the rewarding video or the playable, you’re actually creating the value exchange for, the user, for the players. So they are more likely so to engage, with the ad because the reward that you are providing to them. and also similarly, you know, for the publisher, because the rewarded ad and also the playable, you’re actually giving, like, more things to the users and the users feel like
[00:09:58] Phoena Pang: so they are being considered rather than just, “Oh, I just been, you know, thrown to an ad,” and, you know, actually a lot users that will engage, with the playable, like, in the publisher app. So for the publisher, they also see higher retention. So it’s kind of like win-win for them. And also similar for the advertiser, right? So rather than just, you know, showing, like, a, you know, ad, you know, like, asking for click, so you actually getting a lot of high-intent lead.
[00:10:22] Greg Kihlström: Mm-hmm.
[00:10:22] Phoena Pang: So the advertiser, they actually pay for the user who ha- experienced the app before the app- be- before the user, you know, come into the game.
[00:10:30] Greg Kihlström: Yeah. Yeah. And- and your role gives you a- a global view on this as well. You know, are- are you seeing any regional differences? You know, do different, you know, users in different regions respond to different types of these ad formats and-
[00:10:46] Phoena Pang: Mm-hmm.
[00:10:47] Greg Kihlström: What can brands learn when- when planning their- their strategies, particularly global strategies?
[00:10:52] Phoena Pang: Yeah, totally. So I would say, like, from across all the regions I’m working with is very, very different.
[00:10:59] Greg Kihlström: Yeah.
[00:11:00] Phoena Pang: So I will start with APAC, because in many APAC market is digital-first. So particularly in the Southeast Asia and also, you know, greater China.
[00:11:09] Greg Kihlström: Yeah.
[00:11:10] Phoena Pang: So users, they’re more receptive to those high frequency rewarded format, because the rewards economy also, like, very deeply integrated into their daily life. So if I use like a, you know, a app in China, like, you open the app, there are, like, so many button to collect the daily, like, re- reward.
[00:11:29] Greg Kihlström: Yeah.
[00:11:29] Phoena Pang: You will never see that in US or Europe. And because of that, so the- all the APAC customers, they’re also being educated that way.
[00:11:38] Greg Kihlström: Mm-hmm.
[00:11:38] Phoena Pang: So, there is, like, a very high engagement rate with all this, like, reward-based app, and also, there’s also, like, very re- re- resilient. If you put, like, say, uh…… 30 ads, like, for them a day. They, they wouldn’t turn.
[00:11:53] Greg Kihlström: Hmm.
[00:11:54] Phoena Pang: It’s because the nature of the user.
[00:11:55] Greg Kihlström: Yeah.
[00:11:56] Phoena Pang: And it’s very different in North America, because in North America, so the US and Canadian users, so they will prioritize more, like, you know, the brand familiarity and also the convenience rather than, you know, like, showing them a lot of, like, a lot, like, showing them a lot of ads. So, usually, so, in the US, that’s why, like, rewarded ads are, like, more competitive because, like, you know, for, from the publisher perspective, they don’t want to expose the users to too many rewarded ads. So, but then for every rewarded ad you show to the user, so the US user, they will tend to finish it, so we see very high completion rate, but then also because of that, there is also, like, very low patience for … there’s, like, you know, like, long-loading playable ad, because they are now, like, less resilient to the,
[00:12:41] Phoena Pang: you know, to those, like, rewarded format.
[00:12:43] Greg Kihlström: Right.
[00:12:43] Phoena Pang: but, like, people are still gonna, like, finish, for the ad. and also, for the North American user, we see them, convert really well, for the free trials. so that’s also, like, a little bit different from, I would say, like, other regions.
[00:12:59] Phoena Pang: and, and also in Europe, so in Europe, it’s also, like, like, very different, because in Europe, we see, like, especially for, like, especially for some users, like, we see they’re surpassing US in, like, I would say, like, the, median LTV, like, in some category.
[00:13:17] Greg Kihlström: Hmm.
[00:13:18] Phoena Pang: so the users are, like, more selective. So, they have, like, a very strong preference, like, for quality over the quantity. So, usually, like, you know, when we put the ads in front of the European user, so we want to make sure, so they’re, like, very, you know, high production quality.
[00:13:35] Phoena Pang: Otherwise, it will cause, like, a higher turn rate. and also in, and aga- it’s different from US, because, you know, we just said US user, they prefer to convert on free trial. And actually in EU- in Europe, so European user, they’re more willing to pay for premium ad-free experience-
[00:13:51] Phoena Pang: … once they’re hooked. So, what we see, like, a very interesting trend in Europe, so if you just, like, set up, like, a $0.99 to remove the ad, it’s usually converted really well-
[00:14:03] Phoena Pang: … in, Europe, comparing to other countries. Yeah.
[00:14:07] Greg Kihlström: Yeah, yeah, it’s, yeah, it’s fascinating to see the- the- those differences.
[00:16:08] Greg Kihlström: Let’s talk a little bit, I know you touched a little bit on AI when, you know, toward- towards the beginning, but I wanna talk a little bit more about that, and, in- in two ways, but I wanna start from the user acquisition part of this. And, you know, I- I know a lot of people, many are already using different AI tools to, you know, to help in- in user acquisition, but sometimes it can feel like a black box, you know, you, you know, i- if it works, great, but you don’t necessarily (laughs) always know why. You know, can you maybe talk through what are some of the most impactful practical applications of AI that you’re seeing that are helping, you know, helping brands identify and target those high-LTV users that you, that you’ve mentioned?
[00:16:50] Phoena Pang: Mm-hmm. Yeah, totally. so yes, of course, I mean, like, in our, like, especially in open internet, like, the, our, like, machine learning engine, every time that we tell people is, like, “How you identify how a user is like, oh, yeah, because it’s our machine learning algorithm,” it’s like, you know, like, “Our AI works.”
[00:17:08] Phoena Pang: So, how to demystify the AI, so in this UI, in the, in the UI work, so I would say, like, it’s less about guessing, but it’s more about, like, you know, recognize the right pattern. so, like, when I say I recognize the right pattern, it’s basically, like, we train to identify what’s a high-value user, high L- high-LTV user for you. and, there are, like, you know, different perspective, so I would say, like, the most important, you know, how, you know, AI, like, how we utilize AI, in our, like, ad machine learning engine is about the value-based bidding. so now that we actually, you know, like, doing everything is, like, automatically bidding, it’s not like, you know, the old word of, like, you know, running, like, a CPI campaign.
[00:17:53] Phoena Pang: so we are trying to identify the users based on all the post-back data event. So, when the user, when they behave, so for them who, like, they make a purchase, or, you know, they interact with an ad, how all of those signals, so the machine learning engine, they will collect, and then we will learn. So what are the, how the valuable users are gonna perform? And then, we’re gonna predict the users who are kinda similar to those, I would say, like, the whale behavior, and try to find those users for you. So, that is, like, the value-based bidding, and how is, how we use AI. and also, another thing, so which we also talk about is about, like, the prediction. So, the prediction is, like, how you predict, based on, so because you want to predict for the long-term value, so every,
[00:18:38] Phoena Pang: you know, if you learn, if you, you know, heard about, like, you know, the Day 28 Model, like, from, like, you know, all this, like, you know, the big ad tech company.
[00:18:45] Phoena Pang: so basically, you’re trying to predict a lifetime value, using the first few hours’ signals. So, uh-… as I said, I, you know, all this, you know, the, for example, the tutorial, how the user finish the tutorial, how the user finish, like say, the game level, or, you know, how many sessions they play. So all of those data, so it can be used to analyze, like how valuable the user is. So those, like, prediction for the user LTV, so those are another, like, you know, I would say, like, a big thing that we use in the, like, the AI machine learning, um-
[00:19:19] Phoena Pang: … you know, for basically, like, you know, there’s, like, a lot of things are black box, but those are, I would say, the two main things that we look at, yeah.
[00:19:26] Greg Kihlström: Yeah, yeah. Well, and, and of course, the, the other co-component that we touched a little bit on, regarding AI is, is, the creative aspect, so, you know, creative automation at scale. Where do you see the ideal balance between, you know, human-led creative strategy and machine-driven iteration?
[00:19:45] Phoena Pang: Yeah. I think that’s also, like, a common question because a lot of, team that when it came to us, they, like, worry about their creative team is gonna be gone because it’s gonna, like, all replaced-
[00:19:55] Phoena Pang: … by AI, which I don’t think so because I think there’s still, like, ideal balance, you know, in between, like, the human, and AI, flow. so from what I, from what I see or from what I see, like, the best practice, is that human should define the soul, so which is the emotional hook and also, how you gonna, like, set your brand strategy. Like, what’s your brand? Like, you know, what’s the story? All of those things, it should be defined by human. But then the AI should be handling the skill. So you can generate thousands of the variations and also all the, like, you know, the adoration, all the AB testing. All of those, you can set up based on AI, or you set it, like, in, like, a automation. Um,
[00:20:40] Phoena Pang: and all this automation, they can enhance based on the result automatically. So all of this testing can be done by AI, but then the human will still come in to define the core.
[00:20:54] Greg Kihlström: Yeah. Yeah. So to, to make best use of, of all of the, you know, kind of everything that we’ve talked about, I guess, you know, for, for marketing leaders out there that want to maximize, really across the board, what, what’s, what’s a capability that they really need to be investing in to make sure that their, their mobile strategy is not only effective today, but really positioned for growth in the coming years?
[00:21:19] Phoena Pang: Yeah. Totally. so I would say, like, the most critical inv- y- y- investment, is probably gonna be, like, open internet diversification. So I’m saying this, you know, also coming back, you know, coming from, like, a Google background. So I used to work-
[00:21:34] Phoena Pang: … in the walled garden, but then I’m working in the open internet. and I see the importance, of, you know, basically, like, engaging users in different perspective. So you want to be able to diversify your budget, to bridge the gap between the walled garden and the open internet, and especially, like, in 2026, so we also see a lot of report, the user attention, they have been, like, highly fragmented across a lot of high engagement open internet sectors. For example, people use more show drama, people use a lot of, use, people use a lot of AR-driven apps or other utility apps. So those are all there to play games, right? So those are not, like, walled garden. So, and also to capture those, and the leaders, they also need
[00:22:19] Phoena Pang: to invest in more smart bidding, in, like, a smart bidding, like a strategy. So how you gonna be, looking at a data? How you gonna be set up your value-based bidding strategy? So outside the walled garden, because people are, like, more used to, like, oh, you know, doing user acquisition in Google Meta, but then outside the Google Meta, like, outside the, the walled garden, you know, open internet, you need to, like, set up, like, a new strategy of, you know, like, how do you, like, how you measure, like, the user value, and also, you know, in general, like, how you’re gonna be measured to total, like, the channel efficiency. Yeah.
[00:22:54] Greg Kihlström: Yeah. Yeah. Well, Phoena, thanks so much for joining today. I’ve got two, two last questions as we wrap up here. the first one, if we were having this interview one year from today, what is one thing that we would definitely be talking about?
[00:23:07] Phoena Pang: yeah. I think there’s also, there’s, like, something, you know, our, company we’re working on a lot, I think we’ll be talking about, like, maybe in one year, is about agentic flow.
[00:23:17] Phoena Pang: … so we want to be, like, shifting, like, the company, rather, you know, like, rather than just having AI as a tool, but, you know, also having AI as a operator. So for them in the context of, like, our company, so I want to be able to transform, say, from our AM to do all the campaign management, but then, you know, having the AI agent to really manage the campaign, but then manage the campaign from end to end, not, like, in the single task, but do more, like, in the, like, as I said, like, agentic flow.
[00:23:51] Phoena Pang: I think that would be, like, the future of ad tech, and hopefully, you know, we can talk about more in one year that we’ll be have it more, more, mature product.
[00:24:00] Greg Kihlström: Yeah. Love it. Yeah. Would l- love to have you back on. We can, we can chat about it. Th- that’s great.
[00:24:05] Greg Kihlström: And, last question for you. what do you do to stay agile in your role, and how do you find a way to do it consistently?
[00:24:12] Phoena Pang: Yeah. I think for me, I think there are two things. I was like, one is, like, to, stay open, but then the second thing’s also, like, very relative to the first thing, is to g- get as many feedback as possible because especially in my role, like, I work with all functions. It’s like, I work with product, I works with marketing, like, sales, account manager, you know, from different regions. And everyone, they will have their own roles, you know, based on, you know, their, like, interaction with the clients and also meeting with the clients and hearing their feedback about our, you know, product. So usually, like, when I get those feedback, how can I leveraging all the data signals I get into my feedback loop to, you know, make my thinking better. and also, you know, again, you know, like, I don’t have, like, a, you know, just, like, one set thing I need to follow this rule 100%. Things changing so fast these days. So keep hearing, you know, what’s coming in, you know, have, like, a very, like, open, like, curiosity mindset is, very helpful for me to stay agile.
