Skip to content
  • There are no suggestions because the search field is empty.

AppLovin in Marketing Mix Modeling (MMM)

Which AppLovin campaign data Sellforte uses, at what grain, and how to extract it from the platform.

Introduction

AppLovin is a digital advertising platform where advertisers run performance campaigns. Sellforte uses AppLovin data to measure campaign delivery, spend and the purchases the platform reports, and models AppLovin as its own media channel. This article describes what the dataset covers, which fields Sellforte needs and how to extract them.

What this dataset covers

The dataset covers the advertiser side of AppLovin: spend, delivery and attributed outcomes of the campaigns you run for your own website. Revenue earned as a publisher on AppLovin is out of scope.

Grain and sample

Each row represents one campaign per date and country. Spend and delivery are booked to the date reported by the platform.

datecampaign idcampaign namecountrycurrency codespendimpressionsclicksconversionsconversion value
2025-03-0310001Prospecting AUSUSD1 250.00310 0002 100383 420.00
2025-03-0310002Prospecting BUSUSD640.00150 5001 050171 530.00
2025-03-0410001Prospecting AUSUSD1 310.00325 8002 260413 690.00

Dimensions

FieldImportanceRationale
DateMust-haveDate of the delivery. Sellforte needs the data on a daily level.
Campaign IDImportantStable identifier that stays the same when a campaign is renamed. Used to map campaigns.
Campaign nameOptionalShows campaign details in the Sellforte interface.
CountryOptionalRequired if each country is modeled separately.
Creative setOptionalShows creative level details in the interface.
PlatformOptionalThe platform the ad was served on, such as a mobile operating system or connected TV.
Audience strategyOptionalSeparates activity by the audience approach chosen for the campaign, useful when campaigns with different approaches run side by side.
Currency codeOptionalValidates investments. Needed for ROI calculation if several currencies are used.

Metrics

FieldImportanceRationale
SpendMust-haveAdvertiser spend. Modeling metric, and needed for ROI calculation.
ImpressionsMust-haveNumber of impressions delivered. Modeling metric.
ClicksMust-haveNumber of clicks. An alternative modeling metric.
ConversionsImportantPurchases reported by the platform. Used to validate a maximum cap for modeled results.
Conversion valueImportantCheckout revenue reported by the platform for those purchases. Used to validate a maximum cap for modeled results.
New customer conversionsOptionalPurchases by first time customers as reported by the platform. Useful when the model is built around new customer acquisition.

How Sellforte receives the data

AppLovin data is delivered through customer-managed delivery. See Sellforte supported data delivery methods for the available routes.

Customer-managed delivery

Extract the data with the platform's advertiser reporting for website campaigns. Request daily rows grouped by campaign and country, and add the other dimensions you want to deliver. Request spend, impressions, clicks, and the platform's reported purchases and checkout revenue. The platform offers click through attribution only or click through and view through attribution. Choose the setting your reporting is based on and keep it the same for the whole history. The platform returns a limited date range per request, so fetch longer history in several requests and page through the rows of each request. Dates are reported in UTC. Add currency code and account name as constant columns. The platform reports checkout revenue in US dollars, so state the currency of each monetary column.

Conversions and conversion values are the platform's reported conversions and can differ from the orders in your sales data. They are used for validation and not as the business outcome.