Method

The CREPE method.

Every engagement runs the same decision system. The depth changes by buyer and scope. The reasoning standard does not.

The five stages

Classify, Reposition, Evidence-test, Plan, Execute.

C — Classify

Establish the real Steam category, buyer, comparables, and demand tier, and whether the game sells on looks, on play, or on both.

R — Reposition

Clarify the commercial promise: familiar category, meaningful difference, capsule, trailer, screenshots, copy, tags, and price framing.

E — Evidence-test

Challenge every assumption with Steamworks data, comparables, builds, playtests, creator response, festival results, reviews, and refunds.

P — Plan the beats

Sequence Coming Soon, playtest, public demo, festivals, creators, Next Fest, release date, launch, and catalog activity.

E — Execute and evolve

Implement approved work with named owners, monitor the evidence, and decide: continue, fix, reposition, delay, launch, reduce, or stop.

The evidence standard

Nine fields. Every recommendation.

So you can always distinguish observation from interpretation, and current evidence from inherited doctrine. Documented platform behavior and reliable client data outrank heuristics.

  1. Observation
    What is happening?
  2. Evidence
    What client data, comparable, player feedback, policy, or platform behavior supports it?
  3. Evidence type
    Platform-documented, client data, external research, case, heuristic, inference, or unknown.
  4. Confidence
    High, medium, low, or unsettled.
  5. Alternative explanation
    What else could plausibly explain the result?
  6. Commercial implication
    Why it matters to demand, conversion, execution, compliance, or risk.
  7. Action
    What exactly should change.
  8. Owner and deadline
    Who is accountable, and by when.
  9. Success test
    What evidence would support, reject, or leave the recommendation inconclusive.

Adversarial quality review

Before a recommendation reaches you, it has to survive questions like these:

  • Are correlation and causation being confused?
  • Is an outdated threshold or platform behavior being presented as current?
  • Is this primarily a demand, product, page, traffic, timing, or implementation problem?
  • Are the comparables genuinely comparable in category, scope, price, and audience?
  • Does the page promise an experience the game cannot deliver?
  • Will the activity create qualified player behavior, or only vanity metrics?
  • Is additional spending rational given the evidence and the revenue ceiling?
  • Would a skeptical developer or publisher successfully challenge this recommendation?

What we refuse to sell

Strategy and implementation stay tied to commercial decisions with measurable consequences. That rules out:

  • Daily social media or Discord management as the lead service.
  • Mass press-release distribution or unqualified bulk creator outreach.
  • Advertising as a rescue strategy for weak product-market fit.
  • Guaranteed placement, viral marketing, or fixed wishlist packages.
  • Unlimited access, revisions, title coverage, or undefined consulting hours.
  • Standalone analysis with no implementation or decision system.

How an engagement runs

Eight steps, start to decision.

Production days pause while we wait for access, assets, decisions, or approval, so schedules stay honest in both directions.

  1. Application
    You provide buyer type, game, stage, dates, budget, wishlists, demo status, history, team, and the decision you need to make.
  2. Qualification
    A 20-minute fit call confirms the buyer path and scope. It is not a free teardown or strategy session.
  3. Data room
    You supply Steamworks data, builds, assets, prior plans, creator and festival history, playtest notes, and metrics.
  4. Diagnose
    We classify the market, demand, product, page, evidence, and the principal commercial constraint.
  5. Decide
    Together we confirm positioning, priorities, exclusions, owners, deadlines, success indicators, and the next checkpoint.
  6. Implement
    We and your team execute the approved page, asset, Steamworks, vendor, or milestone work.
  7. Measure
    Before/after traffic, wishlists, demo behavior, coverage, sales, reviews, and refunds are captured where available.
  8. Review
    The commercial review decides: continue, intensify, fix, reposition, test, delay, launch, harvest, or move on.
The boundary we work inside

ContentCrepe guarantees the defined scope, evidence standard, implementation responsibilities, and delivery process. It does not guarantee wishlists, revenue, reviews, coverage, virality, creator acceptance, festival acceptance, or Steam placement.

Commercial honesty cuts both ways: where the evidence points that way, we will recommend repositioning, delaying, reducing scope, stopping spend, or moving on to the next game.

Put the method to work

Bring a decision worth testing.

The application asks for the same evidence the method runs on: page, build, metrics, timing, budget, and the decision in front of you.

Applications are reviewed for fit before any engagement is proposed.