Starting as a Manga Chapter 84: Closed Beta

Gas aren't physical products.With physical goods, once they’re produced, there's no going back—updates can only happen with a new release.

But gas can be updated continuously—old content can be optimized, new content added, chanics adjusted.

That’s why there’s a saying in the software world:A product is always in beta.It’s always being tested.

This is also why many online gas can stay in “open beta” forever.

Though, to be honest, the biggest reason is just that it helps companies save on operational costs.

Ga testing typically falls into two key phases:CB (Closed Beta) and OB (Open Beta).

Closed Beta refers to small-scale, limited testing—usually invite-only.Open Beta is large-scale public testing—no restrictions, and no data wipes.

Right now, Avalon Studio's ga was clearly still in the CB phase.

Tang Yao had already run a very small test once—just a handful of people, all people she knew. That helped confirm things like ga stability, UI usability, and basic functionality.

But the test Kang Ming brought up today?This would be the second round—with a focus on player retention.

Generally speaking, new gas go through multiple rounds of closed testing. So gas do three or more.

And each test serves a different purpose.The first test usually focuses on tech performance and retention data.The second often tests monetization potential.

If any round fails to et its goals, more tests are added.

Because no ga is ever perfect.No matter how prepared you are, closed testing is your chance to let the problems surface—so they can be fixed before launch.

Skipping testing and going straight to release?

That’s basically asking to blow yourself up.

"Exactly… I recomnd we don’t just test the technical side this ti,”Kang Ming said, glancing at Tang Yao, who was thinking aloud.“Let’s also test monetization. Retention, conversion rates, ARPPU—these key trics will help us refine everything.”

“You don’t trust the pricing model, huh.”Tang Yao imdiately picked up on the subtext.

“…”

Kang Ming let out a sheepish laugh. “Just playing it safe… you know, just in case.”

“Alright. Fair enough.”Tang Yao didn’t push back.“Let’s start preparing then. First, post a survey on the website to screen interested readers. Then pick a group of them to participate. That’ll also serve as pre-launch hype—get the word out that we’re making a Fate mobile ga.”

“Also, we’ll need to change how the website is positioned. It’s basically our only official user channel.”

Launching a ga is complicated.

If this were her previous life, Tang Yao would’ve had a headache already—just applying for the necessary licenses would be a nightmare on its own.

But here?Much easier, thanks to being a first-mover.

Of course, “easier” is relative.Being first ans less regulation and fewer roadblocks—but it also ans no established infrastructure. Things like multi-platform distribution? Forget about it.

In her previous life, closed betas were usually run through distribution partners.The biggest na for ani-style gas was Bilibili—it just had the perfect user base.

But in this world?

No such partners existed yet.

Tang Yao couldn’t even find a proper publisher. The mobile ecosystem was still in its infancy.

So for now, she had to do everything herself.

And her best shot at recruiting beta testers was through her own website.At least it ant she could target precisely the right users.

Kang Ming nodded. Then hesitated.“Alright, I’ll start getting things ready. But the questionnaire, testing terms and conditions…”

“I’ll handle it.”

Tang Yao cut him off without hesitation.

Letting a programr handle that kind of user-facing stuff? A bit much.

Even if it was just a test—It would still be the first ti the ga ets real players.

Honestly, Tang Yao was feeling the pressure.

Mainly because—Over the last few months, their bank account had been draining like a leaky faucet.There was no way she wasn’t feeling it.

She might be confident—but she wasn’t made of stone.

After all… that was Miss Li’s dowry she was spending.

Still, this much pressure? She could handle it.

After returning from Kang Ming’s office, Tang Yao sat down at her desk and got to work.

“Alright… for a data-wiped test…”

She picked up a pen and tapped it against the desk, muttering to herself.

“How many resources should we give out?”

In closed betas, it’s common to give players free resources.Since everything’s getting wiped later, you might as well go generous—it helps players enjoy the test and keeps morale up.

But from a product design standpoint, giving out too much can skew results.

For example—if you hand out hundreds of gacha pulls?Players will be too busy enjoying the flashy cards to notice real ga issues.

The result?Fake data.

On the flip side, giving out nothing also doesn’t work.It’s unrealistic to expect players to spend money during a test if everything’s going to be wiped.

Especially since this round was also ant to gauge monetization behavior, striking the right balance was crucial.

The core idea:What influences your data most should be the ga itself, not the rewards you give out. You can’t let bonuses distort the picture.

Fortunately, Tang Yao had experience to draw from.

She furrowed her brows slightly and began recalling how previous mobile ga tests were structured.

Then quickly put down her pen and started typing.

Line after line of content appeared on the screen.

After locking in the test reward structure, she leaned back in her chair.

“If we’re testing monetization too…then we need a solution for post-wipe paynts.”

Because no player wants to see their progress erased.

And that goes double for people who spent money.

Paid-wipe testing required a clear, thoughtful refund policy.

In her past life, the common practice was:Refund all spending from the beta—plus a bonus—when the ga officially launches.

But simply giving cash back felt too impersonal.

There wouldn’t be many paying testers anyway…But this was the first-ever ani-style gacha ga in this world, and the first ga from her studio. Tang Yao wanted to give those early supporters sothing special.

It would an a lot—both to her and to the players.

And those willing to spend during closed testing?They were guaranteed to spend again after launch.So rewarding them was just smart.

Maybe…

Give them an exclusive card?

The idea popped into her head, and her pale, pretty face took on a thoughtful look.

Actually… that could work.A unique card would feel aningful—sothing commorative.

The stats couldn’t be too strong, though—a 4-star rarity would be fine.Anything higher would be unfair to other players at launch.

But she’d draw a custom illustration just for it.

Now the question was—who should the card depict?

Tang Yao started brainstorming.

But just then—

Buzz buzz.

Her phone vibrated on the desk.

Tang Yao glanced over, pausing her train of thought.

She picked up the phone… then blinked in surprise.

“Hi, so sorry… I just saw your DM now.”

After more than three months…

Rumi-sensei had finally replied.