Hey! I’m about to run a short-term 20k budget campaign (using skyscraper ads). I got some decent results in tests in the past, so I’ll use the next strategy: 8.5k (Thursday 11PM EST - Friday 11PM) + 8.5k (Friday 11PM - Saturday 11PM) + 3k (Saturday 11PM - Sunday 11PM) as I’d get more non-school nights.
Do you suggest to start it at 6AM (but on Friday instead of Thursday)?
Also, let’s say I get, for example, 100 concurrent players. How can I keep those? Will I have to keep running ads constantly everyday or keep doing short-term campaigns? Don’t advertise at all? Maybe sponsoring? What worked for you?
Heya! So this website allows you to see previous trends in advertisement rates.
Here’s the last month:
Thumbing through it, in the last month it seems the most expensive days are Saturday-Sunday. On those days you’re looking at around 35 impressions per R$. Advertising Wed-Friday however seems to consistently provide rates above 60 per R$, aka a 70% increase in impressions. As for the specific time of day, I’d recommend doing it in the early morning or late at night. If you have R-Track Pro, you unlock an awesome concurrents chart:
From it you can see a clear trend for the last few day, which is that between the hours of 7AM to 7PM EST you get the majority of play sessions. Since the cost per-add fluctuates by day, the time of day doesn’t really impact cost, but it can directly impact how many players are online. To best test a new game you want at least one full server, and that’s something much easier done during the day.
If you can release in time, I’d recommend Thursday before 7AM, then the same time Friday, then use the two consecutive days of player interactions to determine your day 1 retention rate. If the retention is above 10-15%, go ahead with weekend stuff, otherwise I wouldn’t recommend spending the money.
Let’s see if that’s realistic, below I’m going to solve for what peak concurrents to expect.
Assumptions:
Time period (hours): 24
CTR: ~1.25%
Click-to-Play Rate: ~50%
Avg Session Duration: ~10 minutes (might be lower, might be higher, this will skew a lot of stuff)
Provided:
Impressions per R$: ~60
R$ per day: $8500
Calculated:
Impressions: R$ per day x impression per R$ = 510,000
Clicks: CTR x Impressions = 6375
New players: Clicks x Click-to-Play Rate = ~3188
Players per hour: New players / time period = 133
Sessions per hour: 60 / Avg Session Duration = ~6
Average Concurrents: Players per hour / Sessions per hour = 22 players
Soooo, yeah - you probably don’t need to worry about 100 concurrents. This is the simplified version of the equation though, if you factor in stuff like peak hours you can likely hit around 50% more, or in this case 22 * 1.5 = 33 concurrents.
That’s an important question, but the metric you should be looking at is retention. If a player comes back, you don’t need to keep buying new players. In my experience the three best ways you can achieve this is to make the game social, to have replayable game mechanics, and to provide longer term goals for the player to achieve over multiple play sessions. Those are all easier said than done though.
You’ll likely have to try a few things, so by watching how your retention rate changes over the updates you can see if you’re going in the right direction. In general, a game with >10% D1 retention will have a permanent audience, hence why I recommend only advertising 2 days in a row until you know if your retention is doing fine.
I would not recommend longer term campaigns unless your game is super well monetized and even a single player who doesn’t come back spends more than the cost to bring them in - this is quite rare though.
So, this is actually an interesting strategy lol. My gut reaction was to say this wouldn’t work, but the more I think about it there’s no technical reason it shouldn’t. Now, it’s definitely doing game dev on hard mode, but it’s not impossible. You’d need to have a game that can be fun with only 1-2 people, enough so that a person and their friend could have a good time and a reason to come back.
It’d also have to be discoverable through things like the search bar, so be sure to include relevant tags in the description. If the game’s good enough, word of mouth is certainly strong enough to lift a game with zero advertising. As you’ve tested in the past though, I’m not sure your game currently fits this criteria.
Ehhhh, I’d only recommend sponsoring if you want to target mobile or a specific age group, otherwise advertisements are usually the superior deal.
My Super Hero Life games actually succeeded through word of mouth. Originally I planned to advertise the first after a renovation, however as the renovation continued, people just started joining the game and having fun. In fact the game was super broken, and yet people still played it (ironically, not much has changed in this sense).
In more recent times though, it’s mostly been through updating the game with the express goal of improving metrics like retention and solving for their underlying influencers through analytics. It’s what I recommend to anyone who has the patience for it.
At the end of the day though, sometimes it’s just easier to make a new game than to fix an old one, so give this project all you have but if it doesn’t work out don’t sweat it - you’ll approach the next one all the more prepared.
Hey! I’m pretty disappointed because of that expected result (spending all of my budget for some 20-30 concurrent players), there isn’t any way to enhance this? How can big games make it? They just spend hundred of thousand Robux in advertisements? Should I look for an investor if I really want to get more than 20 concurrent players?
I mean, it’s certainly not going to get you to the front page but as I said prior - the goal for advertising should be to fill a server with players so that they can experience what your game is like when playing with other people. A game that fails to take off after a full server of people for a day, is a game that’s not ready to take off.
Some do advertise like this, but only if their ARPU is above ~7-8 R$. It doesn’t really make economic sense for anyone else.
Instead, the secret to their success is retention - in a top 25 game, 40% of players after their first visit return the following day, and 25% return on the 7th day. These players are recommending the games to their friends, and even indirectly advertising them with things like the “Friends are Playing” sort. A game with a good retention rate actually has a negative cost per player, meaning you gain money whenever a player is recommended the game because you spent 0R$ on getting them in there.
But, as we discussed earlier, your game likely has retention issues if not others - if you want people to return to the game, you need to be smart with your budget and improve the game and only advertise on updates to test for higher retention.
If you can’t see a path to success no smart investor will give you money, and if one does you’ll still be just as stuck as before, just under a lot more pressure. I definitely don’t recommend seeking funding unless you’re seeing steady improvements in your retention pointing towards an eventual success.
I don’t know how else to say this though - your game is the reason you succeed or fail, not your advertisements. There’s no advertising magic bullet that will fix an uncompelling game. Use advertisements as a tool to improve the game - that’s really the best advice I can offer.
Hope this helps, and best of luck on the upcoming release!
-CJ
Hey! I thought about what you told me about advertising on weekdays since I can get 70% more impressions, would you suggest it?
I’ve been thinking about retention: how can my game succeed with a DDR1 >10% if my concurrent playerbase is 20 (10% out of 20 = 2; that means that the next day only 2 players will come back)? Also, (I think I already told you this once) I find pretty difficult getting DDR# in my game, specially if I’m not familiarized with analytics (and my game isn’t popular).
I’m pretty upset right now since I can’t find any way to make my game popular or anything like that, and I’ve been working on this since September 2020.
I do suggest advertising on weekdays, but do so to test your updates for retention improvements. Think about it like you’re having to pay to take a test, such as a driver’s exam. If you fail the test you go back home, practice, and then pay the fee again and retake it. You keep doing so until you pass.
So, concurrents = players in the game atm, what D1RR cares about is new users across the entire day - which in this case is 3188. That means ~319 players who like the game enough to return. With that many people returning to your game, the Roblox algorithm + just people inviting their friends, seems to be enough to get a game rolling.
Unfortunately, that’s usually what holds people back - the amount of math and patience needed to learn and set up and refine your skills in analytics can be intimidating. After an expensive project of mine faceplanted though, I found analytics to be less intimidating than being powerless to fix a game.
I’m sorry to hear you’re feeling upset - game dev’s definitely not a career everyone finds rewarding enough to pursue. I find the challenge of troubleshooting a game to be engaging, and I’ve also been making games since 2008 - at this point I couldn’t imagine doing anything else. I believe anyone can develop the skills to make a living off Roblox, but that doesn’t mean it’s the job for everyone. There’s no shame in just using Roblox as a fun hobby to express yourself.
Retention is going to be something that all developers struggle on. We’re all going to have bias towards our game and think it’s fun, however the player may think otherwise.
Put yourself in the shoes of a player, think of a game you’ve played (not your own game, cause dev bias) and ask yourself why do I come back and play this game? Is it because you’re high level on some leaderboard? Is it because the game has/had unique gameplay? (Had as in it was one of the first to implement said unique gameplay) Is it because of cosmetics? Is it because your friends play it? Daily rewards? Trends?
Why should a player play your game over another game? If you give them a good reason, they might come back
So you suggest advertising on weekdays for testing retention, but not for an advertising campaign (like the one I want to do), right?
Still, these are only 319 players joining through the day, which would only improve my concurrents by 2 or 3.
Would you suggest doing a more equilibrated 6k (Friday) + 6k (Saturday) + 6k (Sunday) campaign or keep the one I’m using (more aggresive on the first 2 days)?
I think I have some mechanics that are unique in the type of game I’m making (as I said replied you privately) but now I feel really “forced” to make bots/NPC’s to fight off for rewards in game (to improve retention/playtime), do you agree?
Mhm, 2 days in a row (preferably starting Wed-Thurs) is the max you should be doing until the D1 retention is above 10%.
So, one undervalued metric is natural discovery - aka how many people just stumble into your game despite not paying for it. They’re typically the result of joining friends, finding on the discover page, or even seeing youtube videos about it. Roblox doesn’t necessarily tell us where they come from, but what we do know is that retention is heavily correlated with higher natural discovery. For example, a game of mine SHL2 had a D1RR of around 14% at its peak, and it’d received almost 15-20k new users for free each day. That 10% mark I keep mentioning just happens to be the tipping point for most games where the natural discovery is boosted to the point of having a permanent audience where there’s always at least one server going.
While I’m sure there may be some benefits to weekend advertising long term (maybe they have longer play durations / spend more), for the short term your main focus is troubleshooting and testing the game, and since you don’t necessarily know how many tests you’ll need imo the best strategy is to keep costs low by advertising on cheaper weekdays.
Didn’t you say that I should run long-term campaigns only if my long-term retention is bad?
I really appreciate all of those advices, but I think I’m not being clear enough: I don’t want to keep costs low to run tests or troubleshooting (since then I won’t have enough budget to run an actual campaign, plus I can do that on DevForum or Reddit), I want to run the definitive short-term campaign to release my game.
I’d recommend running 1k every day (except Saturday, as there’s usually always a problem then), and I would try to run them at a time you are able to run them at and at that same time every time you run an ad. It will take time to achieve players and popularity, so theres no need to worry if you only get ~500 visits in your first week or so, plus ROBLOX has been recently not doing so good in terms of stock and servers, so that definitely has an impact on smaller developers.
Hey! Following to your strategy I would be able to run the campaign for 20 days, but it is pretty the opposite to what @CJ_Oyer (and other popular developers) told me, so I’m a bit confused right now.
At the moment, I’ll focus on improving playtime and retention so I’ll be running these ads on the next weekend (6th - 8th), any advices on that?
Yep, D1 retention = short term retention. Longer campaigns are useful for troubleshooting longer term retention.
Then release your game - the cheapest days to do so are Wed-Friday. I can’t promise people will come back, or that it will earn a profit - those things typically require testing and iteration. I’ve yet to find a magic-bullet perfect ad campaign that can around that.
I wish you luck, and if you change your mind on doing testing and iteration I’ll be happy to continue this convo
-CJ
As you suggested, I’ll add some features to improve retention and playtime and then run the campaign.
So on these days I should get more players for the same amount of Robux, right? It’s just I think that if you come back from school/job (weekdays) and want to play Roblox then you’ll play a specific game (which you were thinking about playing it before) and, instead, if you have a lot of free time (weekend) you’ll explore and look for new games. What do you think about that “theory”?
It seems a reasonable theory, and I believe average play duration is slightly longer on weekends, but I personally don’t have any concrete data to suggest advertising on the weekends would save you money / benefit the experience - all I can really say for sure is it’ll be more expensive per impression.
I’d still say stick with the 1k every day, even though those devs may be more popular, it doesn’t mean that it’s because of ads; often times things such as a popular youtuber visiting (or draining their mom’s bank account) cause this popularity, like with Adopt Me. The 1k every day helps because bigger ads tend to draw a lot of people in at once yet aren’t as good for acquiring player retention over the course of a period of time. Mastering the skill of player retention is hard, unless you get lucky with youtubers such as Flamingo and Jayingee.
