BIG UPDATE! New Roblox Algorithm Changes June 2026

Big update: new Roblox algorithm changes, June 2026. In this video, I'm going to go over basically everything that I've come to understand about the new algorithm, the updates, summarize everything that's in the official documentation, and tell you guys what's important, what's not, and how this changes how you should develop games. Without further ado, let's just get into it, okay? I'm going to show you guys really quick. Let's go to... Anyways, [snorts] if I go here and I go to acquisition home metrics, okay? You guys will see we got some new stats where we're, uh, we're rocking with here, all right? We got some new stats. And I want to dispel some common myths because, god, dude, nothing like Roblox myths on Twitter, people spreading misinformation on Twitter, okay? So, first things first, we have play-through rate, okay? This used to be QPTR, and we have this new metric called first play bounce rate. Basically, what they did is they took QPTR and separated it into two metrics. Now, what's really important to note that I've seen a lot of misinformation on is bounce rate is not a good thing. You'll see this is the only stat where it says high rates lower your home recommendations exposure. Meaning this is a negative stat, meaning you want this number to be as low as possible. Because what is a bounce rate? Bounce rate is basically the percentage of people that, when they join your game, leave within the first 60 seconds. They also have bounce rate for 61 to 180 seconds, okay? So, you want this number to be low. Ours is low-key cooked. Low-key it's like 20%, that's too high, right? You want this number to be low, okay? Play-through rate is basically just what QPTR used to be, but they took the Q out of it and the qualified is kind of being more measured through bounce rate now. So, this is just sheer, um, click-through rate. Let's see what it says. The percentage of home impressions converting to play sessions. Okay? Now, something I want to mention here that a lot of devs also overlook. Shout out to Derp, hit me up about this. And, um, I want to show you guys this. Okay? You think about your game, okay? There's the initial impression, right? Okay, we all know what we're talking about here, right? Impression. Your thumbnail, your game shows on home metrics, right? Okay. Someone clicks the game, there's the experience detail page right here. I can't tell you how many devs I've seen completely ignore this page. This page is very, very important when it comes to your overall play-through rate, okay? If you go to your stats, you go to acquisition here, and you look at user detail page CTR. This is users who played from detail page slash users who viewed the detail page. Okay? So, when we're talking about PTR, we're talking about the initial impression, okay? They click to see detail page, okay? And then from the detail page, they decide to either play or not, okay? Play or not. Now, this is important because if this number... I want you guys all to, while you're watching this right now, go look at what this number is. User detail page CTR. I've seen a lot of devs completely neglect this page. They don't even put thumbnails or anything. Simple one here is to take your best performing thumbnails, either from ads or from home metrics, and just put them here on the user detail page. Now, you want to kind of describe what the gameplay is going to be like. Like you can see I'm not just spamming the same thumbnail again and again. I kind of give an idea of like, "Hey, look at this guy stealing the data. He's hacking. Okay, there's some tycoon elements, blah, blah, blah." All right, that's really important when it comes to overall play-through rate. That's a lot of... yeah, but let's go back to the new stats. I'm going to show you guys what it says in official documentation, which of the stats are most important, and how to bounce, how to improve them from my understanding. Now, bounce rate, I don't know if you guys have noticed this, but I've seen some top devs experimenting with cut scenes. Something to think about. You play Grove Garden 2, there's a 15-second cut scene, okay? Bounce rate really has to do with is your game clickbait or not. All right? Before I keep yapping, let's just consolidate this, okay? We have these new metrics, which you can see if you go acquisition home metrics, you scroll down and we see play-through rate, we see bounce rate, okay? We see play days per user. This is not new, but what is new is that D28 is now being measured, okay? Which we'll also get into. But Roblox has never been more clear on which metrics are most important when it comes to getting more distribution in the algo. And these are engagement signals, okay? Play-through rate, first play bounce rate. These are highly, [snorts] highly important. So, if you have templates or you're running like X999 modded clones, like you're kind of cooked. Or I don't know if you guys have seen those clickbait games where it'd be like it would say like mysterious game, right? [laughter] [snorts] It would say like very mysterious game like this. Look at these, all have zero CCU. And you'd click and it's just a steal a brain rot clone. This shit's cooked. You can't do that anymore. People were like kind of exploiting how QPTR worked, and you can't do that with first play bounce rate anymore. Okay? Play days per user, play time per user, this is just average play time, and this has to do with retention, okay? D1, D7, D28 is play days, and the play time is just basically how much they play on each of those play days. Okay? So, really not much has changed other than QPTR deprecated, it's now play-through rate, first play bounce rate, which again is a negative stat, you want it to be low. I've seen a lot of, uh, misinformation on this. You want bounce rate to be low. I can probably do a separate video on how to lower this once I... I it our game, 'cause it's low-key cooked. I think our title is a little misleading. So, some things are changing. All right, and I'm, I'm learning and I'm figuring it out. Some things are changing. I'm not saying that I'm an expert in all of this, but I'm just learning and sharing my learnings along the way. Okay. Let's keep going. What does this mean? So, engagement is more important than mono, mostly. Okay? Don't get mad at me. That's literally how they spell it out right here. Clickbait is cooked. So, these games that say like mysterious game, yeah, you can't do that anymore, bro. That's cooked. If you do template games or clones, in my opinion, this has always not been the best way to learn game design and get good at development. Okay? Tell template games, clones, I wouldn't do it, bro. I wouldn't do it. Now, I want to make something very clear. Is obviously Hack a Business is inspired by Oil Empire, right? If you do do a quote-unquote clone, you need to make a meaningful twist to the core loop. If you play Oil Empire, you play Hack a Business, you'll see what I'm talking about. You'll see the twist that we did. All right, that's [snorts] that. Now, I'm going to link these resources below, 'cause I mean, it's just crazy that a lot of devs don't even read or watch any of Roblox learn content, but they spell it out for us. This guy right here, which I believe is, um, who is he? He is the principal... Yeah, product manager, Sandy. He says in that video, they want high-quality games on the platform. Okay? They want high-quality games on the platform. So, let's just summarize. What's with this new algo? Everything is the same as it was before. Ranking the algo takes the same time. Okay? From my understanding, based on what they say in this video, D28 retention is now tracked. All right? QPTR got deprecated. It's now replaced with PTR and bounce rate. And this is all in an effort to get more visibility into higher quality games. Okay? Clickbait's cooked. You can't say mysterious game and then, yeah, you'll see. Those are all cooked. So, how to win with this new algo? I'm currently in the proce...

important, what's not, how this changes how you should develop games. Without further ado, let's just get into it, okay? I'm going to show you guys really quick. Let's go to Anyways, [snorts] if I go here and I go to acquisition home wrecks,

okay? You guys will see we got some new stats where we're uh we're rocking with here, all right? We got some new stats.

And I want to dispel some common myths cuz god, dude, nothing like Roblox myths on Twitter, people spreading it misinformation on Twitter, okay? So, first things first, we have play through rate okay?

This used to be QPTR and we have this new metric called first play bounce rate. Basically, what they did is they took QPTR and separated into two metrics. Now, what's really important to note that I've seen a lot of

misinformation on is bounce rate is not a good thing. You'll see this is the only stat where it says high rates lower your home recommendations exposure. Meaning this is a negative stat, meaning you want this number to be

as low as possible. Because what is a bounce rate? Bounce rate is basically what is the percentage of people that when they join your game leave within the first 60 seconds. They also have bounce rate for 61 to 180 seconds,

okay? So, you want this number to be low. Ours is low-key cooked. Low-key it's like 20%, that's too high, right?

You want this number to be low, okay? Play through rate is basically just what QPTR used to be, but they took the Q out of it and the qualified is kind of being more measured through bounce rate now. So, this is just sheer

um click through rate. Let's see what it says. The percentage of home impressions converting to play sessions. Okay?

Now, something I want to mention here that a lot of devs also overlook. Shout out to Derp, hit me up about this. And um I want to show you guys this. Okay? You think about your game, okay? There's the

initial impression, right? Okay, we all know what we're talking about here, right? Impression. Your thumbnail, your game shows on home wreck, right? Okay. Someone clicks the game, there's the experience detail page right here. I can't tell you how many

devs I've seen completely ignore this page. This page is very very important when it comes to your overall play-through rate, okay? If you go to your stats, you go to acquisition here, and you look at user detail page CTR.

This is users who played from detail page {slash} users who viewed the detail page. Okay? So, when we're talking about PTR, we're talking about the initial impression, okay? They click to see detail page, okay? And then from the

detail page, they decide to either play or not, okay? Play or not. Now, this is important because if this number I want you guys all to while you're watching this right now, go look at what this number is. User detail page CTR.

I've seen a lot of devs completely neglect this page. They don't even put thumbnails or anything. Simple one here is to take your best performing thumbnails, either from ads or from home wrecks, and just put them here on the

user detail page. Now, you want to kind of describe what the game play is going to be like. Like you can see I'm not just spamming the same thumbnail again and again. I kind of give an idea of

like, "Hey, look at this guy stealing the data. He's hacking. Okay, there's some tycoon elements, blah blah blah." All right, that's really important when it comes to overall play-through rate.

That's a lot of yeah, but let's go back to the new stats. I'm going to show you guys what it says in official documentation, which of the stats are most important, and how to bounce how to improve them from

my understanding. Now, bounce rate, I don't know if you guys have noticed this, but I've seen some top devs experimenting with cut scenes. Something to think about. You play Grove Garden 2, there's a 15-second cut scene, okay?

Bounce rate is really has to do with is your game clickbait or not. All right?

Before I keep yapping, let's just consolidate this, okay? We have these new metrics, which you can see if you go acquisition home metrics, you scroll down and we see play through rate, we see bounce rate, okay? We see

play days per user. This is not new, but what is new is that it's D28 is now being measured, okay? Which we'll also get into. But Roblox has never been more clear on which metrics are most important when it comes to

getting more distribution in the algo. And these are engagement signals, okay? Play through rate, first play bounce rate.

These are highly, [snorts] highly important. So, if you have templates or you're running like X999 modded clones, like you're kind of cooked. Or I don't know if you guys have seen those clickbait games where it'd be like it would say like mysterious game,

right? [laughter] [snorts] It would say like very mysterious game like this. Look at these, all have zero CCU. And you'd click and it's just a steal a brain rot clone.

This shit's cooked. You can't do that anymore. People were like kind of exploiting how QPTR worked, and you can't do that with first play bounce rate anymore. Okay? Play days per user, play time per user, this is just average

play time, and this has to do with retention, okay? D1, D7, D28 is play days, and the play time is just basically how much they play on each of those play days. Okay? So, really not much has changed other than QPTR

deprecated, it's now play through rate, first play bounce rate, which again is a negative stat, you want it to be low. I've seen a lot of uh misinformation on this. You want bounce rate to be low. I can probably do a

separate video on how to lower this once I I it our game, cuz it's low-key cooked. I think our title is a little misleading. So, some things are changing. All right, and I'm I'm learning and I'm figuring it out. Some

things are changing. I'm not saying that I'm an expert in all of this, but I'm just learning and sharing my learnings along the way.

Okay. Let's keep going. What does this mean? So, engagement is more important than mono, mostly. Okay?

Don't get mad at me. That's literally how they spell it out right here. Clickbait is cooked. So, these games that say like mysterious game, yeah, you can't do that anymore, bro. That's cooked. If you do template games or

clones, in my opinion, this has always not been the best way to learn game design and get good at development.

Okay? Tell template games, clones, I wouldn't do it, bro. I wouldn't do it. Now, I want to make something very clear. Is obviously Hack a Business is inspired by Oil Empire, right?

If you do do a quote unquote clone, you need to make a meaningful twist to the core loop. If you play Oil Empire, you play Hack a Business, you'll see what I'm talking about. You'll see the twist that we did. All right, that's

[snorts] that. Now, I'm going to link these resources below, cuz I mean, it's just crazy that a lot of devs don't even read or watch any of Roblox learn content, but they spell it out for us. This guy right

here, which I believe is um who is he? He is the principal Yeah, product manager, Sandy.

He says in that video, they want high-quality games on the platform. Okay? They want high-quality games on the platform. So, let's let's just summarize. What's with this new algo?

Everything is the same as it was before. Ranking the algo takes the same time.

Okay? From my understanding, based on what they say in this video, D28 retention is now tracked.

All right? Q PTR got deprecated. It's now replaced with PTR and bounce rate. And this is all in an effort to get more visibility into higher quality games.

Okay? Clickbait's cooked. You can't say mysterious game and then, yeah, you'll see. Those are all cooked. So, how to win with this new algo? I'm currently in the process of figuring this out, right? Cuz we have Hack a

Business. I got a few other games that I'm working on. But as I've always said, read data, iterative optimization, watch my previous videos on how to actually improve your stats fast.

Okay? I share my thought process, how we bumped the stats of Hack a Business.

Updates should be improving a specific stat. Okay, I just said that. Just said it twice. You want to read data, analyze competitors, blah blah blah. I've said this in like most of my videos. Go watch them. Now, what I want you guys to think about, and

this is what Sandy described in this video is your game is basically a funnel. All right, let me let me explain this to you guys.

Because your game is really just a funnel. All right? And if if you don't know what a funnel is, it's a very common term to use in internet marketing. But let's say up here, these are all the players on Roblox.

Okay? And this funnel represents your game. All right? So, we have these players on Roblox, and at the very top of the funnel is a home wreck impression, which means if I go on home right here, I see a game. Look at Farmer Digger. RIP

to it. New algo absolutely destroyed this game cuz it has horrible retention. Learning lessons, right?

Top of funnel, we have the impression. Okay? They see the thumbnail. Do they click? Yes or no? If they do, boom, we have play through rate calculated. Now, we're going to lose a significant portion of these players because they saw the game,

but they didn't click. Okay, there's a leak in our funnel, right? We need to get a better thumbnail, maybe the packaging is messed up. We need to optimize it better. Now, what's going to happen is a significant portion of these will

also bounce. So, we're going to get players that go actually play the game after they click, and in the first 60 seconds, a small chunk of them are going to bounce.

Okay? So, these guys are gone. They've bounced. Now, you guys are starting to see how we're losing players in our game based on each of these metrics.

Okay? [snorts] So, good amount of them play. Now, a even smaller portion of them are going to play the next day, which is D1. Okay?

So, we have these guys, they're they enjoyed our game. They're playing the next day. Boom, D1. But, again, another portion of them didn't enjoy the game enough, or [snorts] onboarding was confusing, or the tutorial's cooked, and you're leaking players in your D1.

Okay? Now, what's going to happen? I mean, you can see the pattern, right? We have D7.

So, a even smaller portion of them are going to play the game again a week from now, or over the next course of the next few days, days two to seven.

All right? Now, an even smaller portion of them them are going to play from days 8 to 28.

All right? And you can see how we started with all of these guys at the top on Roblox who initially saw our game, and a lot of them were filtered out because of how the funnel actually works. Now,

how do we win in this new algorithm? Well, we want to figure out where is the biggest bottleneck in our funnel?

Is it play-through rate? Do we have a crappy play-through rate? How do we know if we have a crappy play-through rate?

We just look right here. How we measuring up against other games in the in the genre?

Okay, that's the first place you're going to want to look. Do we have a high bounce rate? If yes, we want to optimize that. And what we want to do is come up with a hypothesis on how we can actually improve these

stats. So, play-through rate's very simple, guys. It's usually just packaging, which is thumbnail and title.

Okay? If you already have a good title, just test new thumbnails. Bounce rate has to do with is there a mismatch expectation between the title and thumbnails and what the actual first 60 seconds of gameplay show.

All right. D1 has entirely to do with tutorial and onboarding. A lot of devs I've seen and talked to, yes, they use funnels, but what they don't understand is that you can actually filter by device type here.

So, I have a 77% tutorial completion rate, which is pretty good. It could be a little bit better.

But, what was happening for us is I noticed I went filter by console. I noticed that our console completion rate was a lot lower. And it's starting to go up. This used to be like 30%, and that's how I found a

console bug. All right. And this being lower than the 77% average amongst all users is bringing down our average. So, we can start to use data to interpret, okay, exactly where is the drop-off? I know that I have a high bounce rate,

but on a which device type? On which traffic source? I can start to break down these metrics to really drill down, where's the exact problem? Now, D1, again, tutorial, onboarding. D7 has everything to do with, in my opinion, from what I've come to

understand, is weekly events, so you're admin abuse, but also just having enough content in the game. And not just content, but mechanics in the game that offer deeper progression systems to the players. Same thing with D28. So, something we're doing for Hack a

Business is we're going to be implementing an index, for example. We're going to be implementing certain mechanics in the game that allow the player to have a much deeper progressive experience. [laughter] I don't know how else to explain it, but

like just more content, but more meaningful content. Not just spamming more models in the game, but ways to interact with the models in different ways. Okay? So, I hope this makes sense to you guys.

This is really just a funnel. You have to think about your game as a funnel, and you want to identify where is the biggest bottleneck in your funnel. And you can do that by looking at your stats. If it's PTR, I already went over

that. If it's bounce rate, right? And you want to start to think about how can I actually optimize each of these stats. And there's different ways to optimize each stat. D1 is usually everything to do with first time user

experience, which I've gone over in other videos. D7 almost always has everything to do with how much content's actually in the game, how quick the player can actually complete the game. I found with Hack a Business it takes

about 2 and 1/2 hours to complete the game, so that's why we're launching a huge update to have more content. And then D28, this is still a question mark in my head, but I imagine that it's just deeper progression mechanics. All right,

this is a lot of yap, but this is how you can win. And one thing I do want to mention is the VP of engineering, Benam right here, he himself suggests to launch AB tests. I'm going to wrap up

this video. AB tests are something I haven't even launched or tested yet in Roblox, but I want to show this to you guys cuz a lot of devs actually don't know what this is. You can go here, go

to experiments, and you can actually create an update for your game and have it only show to a certain percentage of your users and see the impact that update is having on a specific stats.

So, that's how you optimize your entire entire funnel iteratively. Cuz you can AB test thumbnails, but you can also AB test the updates and target specific metrics that you want to improve over time. So, a lot of yap.

Hope this was helpful, guys. Peace and love.