Incrementality Is the Business Case: Measure What DTC Actually Changes
A direct channel is strategic only when the team can distinguish value created from transactions that merely changed route or timing.
The DTC business case should be built around a counterfactual: what would this player, cohort or market have done without the direct intervention?
A revenue line has no counterfactual
DTC revenue can rise while the underlying business creates little new value. Existing payers may move a purchase from the app store to the web. A promotion may pull next week’s demand into today. A discount may increase conversion while reducing total contribution. Each outcome can look positive in a channel dashboard.
The business question is different: what changed because the studio could communicate, recognise context and intervene directly? Revenue is evidence that a transaction occurred. It is not evidence of why the transaction occurred.
Classify value before calculating it
Match the metric to the mechanism
A channel-shift hypothesis needs payment-mix and contribution-margin measures. A reactivation hypothesis needs return rate, subsequent sessions and downstream retention. A first-purchase hypothesis needs payer conversion and later payer quality. Using the same KPI for every intervention hides the mechanism.
Two campaigns each generate $100,000 in web sales. Campaign A mostly shifts purchases from established payers who would have bought in-app. Campaign B reactivates dormant players, creates first purchases and produces repeat participation after the offer ends. The channel revenue is equal. The mechanisms, net value and strategic implications are not.
Use decision thresholds, not retrospective storytelling
Before launch, define what result would lead the team to keep, modify or stop the intervention. The threshold may combine incremental contribution, repeat behaviour and an operational cost ceiling. This prevents a campaign from being declared successful simply because one visible metric improved.
Guardrails prevent false wins
Useful guardrails include in-app purchase displacement, refund and chargeback rate, support contacts, offer fatigue, future payer frequency, fulfillment latency and player complaints. The goal is not to eliminate every trade-off — it is to make the trade-off visible.
Incrementality is never perfectly knowable. True. The standard should not be certainty — it should be decision-grade evidence that is explicit about assumptions and stronger than the evidence used for the previous decision. A useful model shows a range rather than a single heroic number.
Create a decision record
For each meaningful intervention, record the hypothesis, intended mechanism, audience, comparison, primary outcome, guardrails, result and next decision. Over time, this creates institutional memory of what changes player behaviour under which conditions. The business case for DTC is the studio’s growing ability to identify, test and repeat mechanisms that create net player value.
