Our team's campaigns · Subscriptions & utilities

Phone cleaner app · US/UK/AU/IE/PH

Our team compared five acquisition sources for a subscription phone cleaner. The campaign recorded 9,180 trials started and 132,450 installs on $215,000 spend.

Business model
Subscription with trial
Markets
United States, United Kingdom, Australia, Ireland, Philippines
Channels
Mintegral; Unity Ads; Google Ads; TikTok Ads; Meta Ads
Services
User acquisition / Creative studio / Measurement & attribution / Traffic quality

The challenge

The growth constraint.

The cleaner needed a source mix that balanced install volume with trial and purchase quality. User quality differed sharply between networks, and trial costs rose when individual markets or Tier-1 spend scaled aggressively.

The approach

Decisions behind the delivery.

  1. Separate funnel goals

    Our team split campaigns into Install, Trial Started and Trial Converted goals across five sources.

  2. Review bids and placements

    Weekly bid optimisation and placement pruning in Mintegral and Unity Ads addressed differences in downstream quality.

  3. Shift towards trial depth

    Budget moved towards sources with better trial quality and funnel depth. Regular creative refreshes kept new tests available.

Attribution

Our team’s campaign work.

Channels

Where the plan ran.

Channel scope and rationale
Mintegral became the main source for trials and purchases. Unity Ads supplied low-cost installs. Google Ads supplied high-quality search traffic. TikTok Ads tested new markets and creative; Meta Ads added performance volume.
Delivery controls
Weekly source, bid and placement reviews informed budget allocation. The team balanced install cost with trial quality, rather than treating low CPI as the sole success measure.
  • Mintegral
  • Unity Ads
  • Google Ads
  • TikTok Ads
  • Meta Ads

Creative

The product promise, tested.

Formats and tests

Regular creative refreshes supported the five-source comparison. TikTok Ads supplied a route for new creative and market hypotheses; the source does not give a creative count or isolated uplift.

Readout

The campaign’s creative read sits alongside source quality and funnel depth. No measured creative-only effect is reported.

Measurement

The basis behind the results.

Source
Team campaign data.
Platform
Android.
Event basis
Installs and trials started are reported across all five sources. Purchases are reported only for Mintegral, Meta Ads and Unity Ads; TikTok Ads and Google Ads purchase cells are unreported, not zero.
Cost basis
USD. All-source spend $215,000. The 3,930 reported purchases use $178,800 spend from Mintegral, Meta Ads and Unity Ads only, giving the source-reported blended CPA purchase of $45.50.
Reporting scope
Campaign totals and source-reported metrics. Attribution windows and cohort maturity are not supplied; no causal lift or measured payback is inferred.

Results

What the campaign reported.

  • Trials started

    9,180

    Reported total across all five sources.

    Source: Team campaign data

  • Installs

    132,450

    Reported total across all five sources.

    Source: Team campaign data

  • Media spend

    $215,000

    Reported total across all five sources; purchase CPA uses a narrower three-source scope.

    Source: Team campaign data

CPA purchase: sources with reported purchases

USD per purchase

Mintegral $44.00Mintegral · CPA purchase: $44.00
Meta Ads $48.93Meta Ads · CPA purchase: $48.93
Unity Ads $44.51Unity Ads · CPA purchase: $44.51

Mintegral reported CPA purchase of $44.00. The three-source blend was $45.50 on $178,800 spend and 3,930 purchases; Google Ads and TikTok Ads have no reported purchase data.

Team campaign data

View data table
CPA purchase: sources with reported purchases
MeasureCPA purchase (USD per purchase)
Mintegral$44.00
Meta Ads$48.93
Unity Ads$44.51

Interpretation

The five-source campaign recorded 9,180 trial starts. Within the three sources reporting purchases, Mintegral had the lowest reported CPA purchase at $44.00 and the largest purchase count at 2,100. The three-source purchase blend must not be presented as a five-source result.

Source table

Spend and event counts share each row’s source scope. CTR = clicks ÷ impressions; CPI = spend ÷ installs; CPA purchase = spend ÷ reported purchases. “—” means unreported, never zero.

Source Spend Impressions Clicks CTR CPC Installs CPI Trials started Purchases CPA purchase
Mintegral $92,400 18,600,000 576,000 3.10% $0.16 62,000 $1.49 4,020 2,100 $44.00
Meta Ads $54,800 10,250,000 256,250 2.50% $0.21 28,700 $1.91 2,180 1,120 $48.93
Unity Ads $31,600 7,840,000 282,240 3.60% $0.11 25,300 $1.25 1,430 710 $44.51
TikTok Ads $21,200 4,980,000 109,560 2.20% $0.19 11,400 $1.86 910 — (unreported) — (unreported)
Google Ads $15,000 1,200,000 36,000 3.00% $0.42 5,050 $2.97 640 — (unreported) — (unreported)

Source: Team campaign data

Additional reported figures

All-source totals
  • $215,000 spend
  • 132,450 installs
  • 9,180 trials started
Three-source purchase scope only
  • 3,930 purchases on $178,800 spend
  • blended CPA purchase $45.50

Source: Team campaign data

Metric basis explorer

Trials started: 9,180. Basis: Reported total across all five sources. Source: Team campaign data.

What comes next

The next question.

Which source can add trial volume while preserving paid-event quality? Keep purchase coverage visible before making an all-source allocation decision.

Detailed case studies, including programmes above $1M per month, are shared under NDA.

Working through a related growth question?

Tell us about your app, markets and the paid event you need to understand.