Follow

Games marketing

Follow how we test ideas for finding players, read the evidence, and decide what to try next.

Organic campaigns live · measurement gaps remain · Updated

Organic campaigns are live. The agreed target week has closed, but comparable downloads, purchases, eligible game sessions and a complete demand census remain unavailable. November 6 delivery is reported separately from that cutoff.

Latest learning and decision

Weekly review

Finish the story, then close the measurement loop

What we saw
Owner feedback caught repeated footage, stale A-10 sounds and static captions. Saved November 6 receipts record corrected A-10 and Omega releases, but weekly downloads, purchases, eligible sessions and qualified demand remain unverified.
What we decided
Verify current source and runtime picture/audio before editing, freeze the player-visible proof early, and give GrowthOps dedicated collection time. Keep two active packages and independent GPT review.
What happens next
Reconcile the closed KPI window from compatible per-game reports and record actual post observations when available. Keep Civilization capture held while engineering investigates simulator startup; no cause has been established.

This retrospective uses saved production and publication receipts, owner feedback and the November 3 interim scorecard. Public links identify the creative; they were not freshly reverified in this review. Delivery and owner preference do not establish audience lift.

Earlier marketing updates

Planned experiments

These follow-ups are in preparation. Each needs verified game proof, reporting and release checks before launch.

Danger Close: WW1

Proposed · Checks pending

Does explaining the free ending and optional unlock help interested players visit the store and buy?

Planned approach

Show the free campaign ending and optional unlock with actual gameplay proof.

Game page

Hawgs over Iraq

Proposed · Checks pending

Does showing a complete gameplay decision bring first-time downloads?

Planned approach

Follow the player’s first-session support unlock, positioning and deployment.

Game page

Omega War: Unthinkable

Readiness pending · Checks pending

Can a clear gameplay change hold attention and prompt interest in playing?

Planned approach

Choose three new, source-qualified visible game changes. Preserve the before-and-after consequence and test one Facebook opening variable; collect comparable observations before scaling effort.

Game page

ComboThrone games

Proposed · Checks pending

Do players ask to try a game after seeing how a decision changes the board or workforce?

Planned approach

Explain board-piece boons in Puzzle Quest and workforce decisions in Civilization. Each game keeps its own availability information.

Game page

What we’re working toward

Build an Arcade people return to, a reputation for distinctive indie games, and a company that learns from the work.

Games are our focus now. The wider company and software portfolio remain on the record.

Goals and working process

Next 4–6 weeks

Make games easy to discover and access. Check which posts bring visits, downloads and play, and begin measuring whether players return where supported.

Next 90 days

Find repeatable ways to bring players to multiple games without paid promotion, record specific game interest, and improve production and review.

Next 6–12 months

Grow a recurring Arcade audience, a trusted indie-developer reputation, repeatable launches and sustainable commercial results.

Changelog highlights, design decisions and player-feedback follow-ups continue alongside product campaigns. Weekly games roundups, fiscal-quarter operations reviews and next-operation introductions make up the wider company story.

The review loop

  1. Choose a player question and a specific piece of gameplay proof.
  2. Review the claims and footage, then adapt the post for each platform.
  3. Publish when the release checks pass, and record the exact links and dates.
  4. Read the results at comparable ages and keep missing evidence visible.
  5. Decide what to repeat, change or stop, and publish what we learned.

Weekly reviews also track production time, repeated factual corrections and missing observations. We use those records to improve how the company works.

How we measure

The owner agreed these starting goals for October 30–November 5, 2026. They are management ambitions with limited baseline evidence. Visits, downloads and play stay separate; unavailable figures stay blank. Total sales and sales linked to a campaign are reported separately.

Player acquisition

People who start playing a game, once real gameplay can be measured. App downloads and page visits are counted separately.

50 first-time native downloads (stretch 100), plus 100 measured visits to playable game destinations (stretch 200). These are separate outcomes, not a combined player total.

The November 3 audit found an App Units report with a different definition and window from the agreed first-time-download goal. A matching weekly download report and real eligible game-destination session source remain unverified; no acquisition actual is scored.

Who is counted
Released native games and publicly playable browser games, reported separately by title.
When we measure
Weekly timestamped events; native downloads use complete source report days in their own timezone.
Players who return

People who play again. Check whether new players return seven days after first playing; returning to a webpage alone does not count.

No numerical retention target yet. Verify gameplay measurement and establish a valid baseline.

Actual gameplay-return events and mature day-seven cohorts remain uncollected in the latest saved audit. Webpage returns do not establish repeat play.

Who is counted
Per-game player cohorts. Arcade totals require verified identity and deduplication support.
When we measure
Mature day-seven cohorts and weekly repeat-play counts, with the return event and denominator stated.
Discovery across games

People who start a second game in the library. Clicking another game’s link is reported separately until actual play can be verified.

Prepare one relevant recommendation experiment; a numerical second-game-start target awaits valid measurement.

We have not verified a way to connect play across games. Recommendation clicks stay a separate intent signal.

Who is counted
Eligible Arcade players with authorized cross-game measurement; no assumed cross-device identity.
When we measure
A defined seven-day period once cross-game measurement is available.
Qualified game demand

Distinct audience accounts explicitly asking to play, test or access a named game. Generic praise, craft-only discussion and owner replies do not count. Creator trust and returning content audiences are supporting measures.

10 distinct platform accounts with qualified named-game interest; stretch 20.

The latest saved audit covered only selected author threads. A complete weekly named-game access census remains unavailable; general praise and creator replies do not satisfy this goal.

Who is counted
Non-owner accounts, reported by game and platform. Cross-platform deduplication requires verified identity; one inquiry is counted once.
When we measure
Weekly received contributions and compatible 28-day trends.
Commercial results

Purchases, refunds and the amount received for each game with a paid offer. Sales linked to a campaign stay separate from total sales.

5 verified purchase or unlock transactions; stretch 10. Free downloads do not satisfy this goal.

The November 3 native Sales/IAP audit did not provide a scoreable weekly result or content attribution. Purchases, refunds, proceeds and actual marketing costs remain unavailable; no sales or ROI result is inferred.

Who is counted
Released games with a verified paid offer. Free downloads are not paid sales.
When we measure
Complete native report days and compatible 28-day trends. Delayed reports retain their actual cutoff.
Portfolio support

Every eligible active game receives a meaningful weekly public feature and feedback triage. Held games keep an explicit status; a mention or internal assignment is not a published feature.