Supercell's mo.co and a Soft Launch Strategy for Games

Supercell has spent this month moving mo.co through an invite-based rollout rather than a conventional open soft launch, according to reporting on the studio's approach. Whatever mo.co turns out to be at global scale, the rollout mechanics deserve study now, because they are a clean example of a soft launch strategy for mobile games built to protect signal quality rather than to produce a splashy download number.

Most studios treat soft launch as a smaller version of global launch. Pick two or three markets, buy some install volume through the usual networks, then watch the dashboards fill in. Supercell's invite mechanic does something else. It rations access, so the players inside the test have selected themselves for interest rather than arriving through a CPI-optimised campaign, which changes what the resulting data can tell a team.

Why gating access changes what you can measure

A soft launch fed by paid UA mixes two populations that nobody can cleanly separate afterwards: players who wanted the game, and players who clicked because a network optimised toward an event. Retention curves get diluted by the second group. So do session length and early monetisation, and studios routinely miss it until the numbers land worse at wider launch than the soft launch predicted.

An invite mechanic filters for intent before the first session starts. Someone has to want an invite, ask a friend for one, or sit in a sign-up queue until access arrives. That is a weaker filter than an existing waitlist of superfans and a far stronger one than an open app store listing. The cohort it produces is smaller. Its behaviour, though, sits much closer to what a studio should expect from the eventual core audience, which is the group that decides whether a live-service economy holds up.

What a soft launch strategy for mobile games should actually test

A soft launch strategy for mobile games earns its cost only if it answers specific questions before the studio commits to paid scale. At minimum, that means:

• Whether D1, D7 and D30 retention clear the studio's internal bar for the genre, not an industry average pulled from a benchmark report.
• Whether the core loop survives repeated exposure, meaning session count per player keeps climbing rather than flattening by week two.
• Whether the monetisation model produces payers at a rate that makes the unit economics plausible once CPMs rise at scale.
• Whether the creative concepts used in marketing tests to describe the game match what players actually experience once they get inside it.

An invite gate adds a fifth question that an open soft launch cannot answer as cleanly: does the access mechanic generate organic pull by itself. If invited players convert their allocation of invites at a healthy rate, that is a genuine acquisition signal rather than a retention one.

The invite mechanic as its own UA channel

Treat the gate as a channel in its own right rather than as friction to route around later. A referral-driven waitlist that grows without paid spend tells a UA team something no CPI curve can: word of mouth has a pulse before a single pound goes into Meta or Google. Studios that skip the step go straight to a paid soft launch. They meet the organic question only after wider launch, when answering it costs far more.

Invite-only isn't the right choice for every soft launch, though. Games with obvious genre-fit and a crowded competitive set may need broader early exposure, because their creative has to face real competition for attention rather than a queue of self-selected fans. The right call turns on category risk and on how confident the studio already is in the mechanic.

There is also a cost side to the gate that studios sometimes underweight. A waitlist page, a referral-code mechanic, a manual approval queue: whichever shape the invite system takes, somebody has to build it and then sit there moderating it, which costs engineering time and community time that a straightforward open listing never asks for. For a small team, that overhead eats into the very runway the gate protects by lowering wasted UA spend. The trade only makes sense where the studio genuinely lacks confidence in its retention numbers or its monetisation model at this stage. Where the mechanic is already well understood from a studio's prior titles, an invite gate mostly adds friction and little else.

A worked soft launch decision framework

Consider a hypothetical hybrid-casual title with a two-market open soft launch planned. A gated alternative might look like this, with figures used purely to illustrate the shape of the decision rather than as real benchmarks:

• Open soft launch: 40,000 installs over four weeks, blended D7 retention of 18 percent, CPI of 2.10 dollars, no clear read on how much of the retention curve is paid-traffic noise.
• Gated soft launch: 6,000 invited players over the same period, D7 retention of 27 percent, near-zero paid spend, plus a referral conversion rate that can be tracked as a standalone metric.

Neither number is inherently better. What matters is which question the studio needs answered first. Where the open economics already look shaky, an invite gate buys a cleaner read before paid spend scales the problem. If the mechanic is proven and the live question is competitive creative performance, an open soft launch answers it faster and the gate adds nothing the studio does not already know.

What to watch as mo.co scales

The test of Supercell's approach arrives when mo.co, or whatever the studio eventually takes wide, moves from invites to open registration. That transition is where the two data sets meet: does organic pull from the gated phase translate into a lower blended CPI once paid UA switches on, and does retention hold once the audience widens past the self-selected core. A studio reading its gated cohort honestly should be able to predict, within a reasonable band, where retention lands once the gate lifts. Surprise at that stage usually means the team read its gated data as validation rather than as a forecast, which is the most common way a promising soft launch turns into a disappointing global one.

The other thing worth watching is whether Supercell keeps any form of gating once paid UA does start, rather than switching to a fully open funnel overnight. A hybrid approach, where paid traffic still passes through some qualification before it counts as a real install in the studio's own reporting, would suggest the studio sees the gate as a permanent quality filter rather than a temporary soft launch device. As we argued in Portfolio Theory for Creative: How Many Concepts to Run, the value of any pre-launch test is only as good as how honestly a studio reads the gap between its test cohort and its eventual audience. Supercell's gate is a sharper filter than most studios use. The filter still has to come off eventually, and that is where the real UA plan starts.

Related archive reading

• What Gamescom Mobile Marketing Teaches Creative Teams
• What Big-Budget Launch Campaigns Teach Mobile UA Teams