Skip to main content

Allocating Organics to Paid networks

Written by Kevin Jabbour

Summary

  • You can set paid, incremental blended, or full blended ROAS targets – see the Setting ROAS Targets section

  • You can include or exclude specific networks from receiving organic contribution via Settings

  • Blended targets allow you to increase spend on networks that drive organic uplifts

  • When using blended targets, organic installs are attributed differently depending on whether you want to account for incremental uplift from UA spend only or all organic impact

Why This Matters

  • Some ad networks or campaigns naturally generate more organic installs than others

  • An optimal UA strategy allocates more spend toward networks proven to lift organic installs

  • Ktrl lets UA managers set blended ROAS goals, while still giving ad networks paid-only targets that align with those blended goals

  • This prevents over-crediting paid channels with organic installs they did not drive

  • UA managers always retain the ability to override Ktrl's model with their own judgment

How It Works

Step 1: Configure Organic Allocation

  • Go to: Settings → Organics and Blended

  • Each ad network has a toggle to include or exclude organic allocations

    • Included:

      • Paid installs from this network are used in the organic factor calculation

      • Organic uplift from this network is allocated in blended ROAS calculations

    • Excluded:

      • No organic uplift is credited to this network

  • By default, all networks are included

  • This setting applies across all countries and platforms for that network

  • Changing this setting triggers a model update, which can take several minutes depending on game size

Step 2: KTRL Builds the Organic Prediction Model

  • For each country–platform segment, KTRL runs a linear regression on 90-days of historical paid:organic install data for each daily cohort

  • The regression produces two key values:

    • True Organics (y-intercept): the baseline organic installs that occur without paid influence

    • Organic Factor (slope): how many incremental organic installs are generated for every additional paid install, quantifying organic uplift driven by paid UA activity

Step 3: Allocating Organics Across Campaigns

  • Incremental organic installs are allocated only to campaigns within included networks, weighted by each campaign’s share of paid installs among those included campaigns

  • Baseline (“true”) organic installs are allocated across all campaigns, including those in excluded networks, weighted by each campaign’s share of paid installs across the full campaign set

Worked Example

E.g US Android - Jan 1st 2025

  • Total installs: 10,000 (actual, country-platform granularity)

  • Paid installs: 6,000 (actual, country-platform granularity)

  • Organic installs: 4,000 (actual, country-platform granularity)

    • True organic installs: 1,500 (linear regression)

    • Incremental organic installs = 4,000 - 1,500 = 2,500

All campaigns:

  1. Campaign A (Applovin)

  2. Campaign B (Google)

  3. Campaign C (Digital Turbine)

Included networks for incremental allocation: Applovin, Google

Excluded network: Digital Turbine

Incremental organic install allocation (by campaign)

  • Total incremental organic installs (US · Android): 2,500

  • Eligible networks: Applovin, Google

  • Incremental organic installs are distributed between Applovin and Google campaigns in proportion to their paid install volumes (known actuals)

  • This method credits incremental organic growth only to campaigns actively driving additional organic demand

True organic install allocation (by campaign)

  • Total true organic installs (US · Android): 1,500

  • All campaigns participate in the allocation

  • True organic installs are distributed across campaigns in proportion to their paid install volumes

  • Baseline organic installs reflect underlying brand demand and market presence, and are therefore distributed across all paid activity rather than attributed to specific incremental drivers

Full blended ROAS calculation (by campaign)

  • Organic revenue is allocated to each campaign using the same proportional allocation by share of paid installs to derive organic LTV per campaign

  • Applovin and Google are included, so they receive both incremental and true organic allocations

  • Digital Turbine network is excluded, so it does not benefit from incremental organic uplift

  • Result:

    • Applovin’s blended ROAS rises from 100% → 175%

    • Google’s blended ROAS rises from 100% → 175%

    • Digital Turbine blended ROAS risen from 100% → 113%, since no incremental organics are allocated

Incremental blended ROAS calculation (by campaign)

  • Organic revenue is allocated to each campaign using the same proportional allocation by share of paid installs to derive organic LTV per campaign

  • Applovin and Google are included, so they receive incremental organic allocations

  • Digital Turbine network is excluded, so it does not benefit from incremental organic uplift

  • Result:

    • Applovin’s blended ROAS rises from 100% → 150%

    • Google’s blended ROAS rises from 100% → 150%

    • Digital Turbine's blended ROAS remains at 100%, since no organics are allocated

Step 4: Converting a Blended ROAS Target into a Paid ROAS Target

  • Users can set a blended ROAS target (e.g. 100% at D180)

  • KTRL calculates the paid ROAS target to provide to the ad network so that, when organics are credited, the blended goal is met

  • This means:

    • The paid ROAS target will be < blended target

    • Networks accept a lower return per paid user, offset by organic uplift

Example:

  • Blended target = 100% at D180

  • Applovin’s paid ROAS target is lowered to 67%, but with organic allocation, blended ROAS returns to 100%

  • The adjustment requires an iterative calculation, since lowering paid ROAS → more spend → more installs → more organics → adjustment repeats until blended = target

Cases where blended calculations revert to paid:

  • Unrealistic fit: If a model produces a negative slope and a negative intercept

  • Data Insufficiency: If the average daily number of paid installs is smaller than 10

  • Not Enough Organics: If the cohort dates actual total number of organics is smaller than the model’s number of true organics, we assume there aren’t any incremental organics to allocate to the other paid segments

Did this answer your question?