// Channel · Google App Campaigns

Google App Campaigns tuned to iROAS or tCPA — not to Google's default.

Universal App Campaigns rotate assets on our cadence, bid to your economic target, and map each ad group to a Play Store Listing Experiment so paid traffic never lands on stale creative.

4% of managed spendBelow $12K/mo → $479/mo floor · Above $12K/mo → flat 4%
// Google App Campaigns · asset rotation
// asset · rotation
video · portrait · v31.42 iROAS · promoted
video · landscape · v10.71 iROAS · killing
image · lifestyle · v51.11 iROAS · hold
image · UI · v21.28 iROAS · scale
headline set · benefit+18% CTR · winner
↳ Winners → Store Listing Experiment · D7 iROAS target 1.15
20+
Assets per group
video, image, text
≤14d
Asset review
against control
iROAS
Bid strategy
or tCPA · not tCPI
0%
Media markup
billed to you by Google
// Campaign structures

What we actually run on this channel.

Every structure below maps to a decision your finance team should have a view on — not a template we run because the platform recommends it.

Structure · 01

Install (iROAS)

For subscription and in-app-purchase apps — bid to a return target on Day 7 or Day 30 events.

Structure · 02

Install (tCPA)

For utility and free apps where install-quality proxies for revenue. Event schema tuned around meaningful actions.

Structure · 03

In-app action

Post-install optimisation — trial start, subscription, purchase. Requires a clean event schema before we launch.

Structure · 04

Pre-registration

Pre-launch demand capture — mapped to a launch-day metadata plan, not a standalone tactic.

// The playbook

How we take this channel from zero to compound.

Same four-beat wave as our organic practice. Nothing about paid earns a different rhythm.

Wk 1

Event schema

Audit and rebuild your post-install event schema so the bid model has meaningful signal on day one.

Wk 2

Asset build

20+ assets per group — headlines, descriptions, portrait/landscape video, HTML5 where relevant.

Wk 3–6

Learn phase

Explicit learn-budget scoped. No optimisation actions during learn — we watch it, not steer it.

Wk 7+

Compound

Winning assets promoted to Store Listing Experiments; losing assets killed at 14 days regardless of Google's recommendation.

// Creative cadence
  • Portrait video + landscape video + at least 5 image variants per ad group at launch.
  • Asset rotation on a 14-day review cadence — Google's 'learning' status is not an excuse to keep dead creative alive.
  • Store Listing Experiments run in parallel — winning UAC creative becomes the new organic control.
// Measurement
  • iROAS or tCPA reported against your written target, not against Google's recommended bid.
  • Firebase / GA4 event schema documented and re-audited quarterly.
  • Monthly board-ready deck with a written note on which decisions are safe from the data.
// Guardrails
  • We optimise to your economic target, not Google's tCPI recommendation.
  • Every ad group maps to a Play Store Listing Experiment. Paid never lands on stale organic creative.
  • Media spend billed to you directly by Google — we never touch client cash.
// Questions

Questions we get about this channel.

Why not just optimise for installs?+

Because tCPI optimises for the cheapest install, not the most valuable one. For any app monetising on subscription, IAP, or ads, tCPI leaves 30–60% of value on the table within 90 days.

Do you handle Firebase setup?+

Yes — event schema, conversion linking, and audience audits. We treat the tracking stack as scope, not a prerequisite the client has to solve.

What about Play Store Listing Experiments?+

Included. Winning UAC creative gets promoted to a Listing Experiment; winning experiments push back into the default listing. Paid feeds organic.

Can you run Google Ads only?+

See Apple Search Ads FAQ — same rule. Paid without ASO burns cash. We scope with a written economic forecast.

// Continue reading
// Free audit

Want us to scope paid UA on this channel?

45-minute audit. You leave with a written unit-economics read on whether this channel is worth your next dollar of spend.

30 seconds. No commitment.