Intelligent Budget User Guide & Best Practices

General Process

Intelligent Budget distributes your overall monthly marketing budget across channels based on projected long-term performance. The resulting budgets are then executed either through Intelligent Automation (for AlgoLift-managed channels) or by the client for non-managed channels. This guide highlights the process and best practices for clients using IB to distribute their spend.

Data requirements

To effectively deploy Intelligent Budget, we need to ingest your user data and your ad network data.

Client Inputs

1. Monthly Budget Goal
   The total monthly budget to distribute across all eligible channels. This budget should exclude any reserved spend for testing or retargeting.
2. Performance Goal
   Business goal to meet. Can either be a target, e.g. "break even on advertising spend by 90 days post install", or simply to maximize a KPI whilst spending the budget in full.
3. Portfolio Definition
   A list of all eligible channels.

Scheduling

• Initial Allocation
  New monthly budgets & client configured business logic should be delivered to AlgoLift 7 days before EoM or as early as possible.
• Intramonth Re-allocation
  After the first 7 days, budgets can be recalculated every day with the additional data from each new daily cohort. The frequency of delivery of the new budgets is client configurable, but AlgoLift’s recommended best practice is once per week to avoid daily volatility.

Best Practices

Updating Client Inputs

The client should update the monthly budget goal at most once per week for budget changes < 20%. Larger budget goal changes may not be possible to meet, especially late in the month.

Data Formatting - Standardize Channel Names

The names of each channel should match in both the user level attribution data and the ad network metrics data, if both of these are being provided by the client to AlgoLift. For example, if the ad network metrics data uses 'Facebook' and the attribution data uses 'Facebook Ads', then the client should add an ETL step to map the user data ad fields as 'Facebook'.

Execution of Allocated Budgets

AlgoLift will execute on budgets on channels for which we are leveraging Intelligent Automation (IA) to manage spend. Please note that budget allocation between AlgoLift-managed channels may be very dynamic as IA will respond quickly to changes in performance and is free to move spend between such channels. The allocations provided for AlgoLift-managed channels should be treated as early estimates of how IA might allocate spend between them; the resulting spend distribution may differ significantly depending on future campaign performance.

For IA ad networks, AlgoLift will only count spend from campaigns ingested into our system toward the allocated budgets. Spend from retargeting, test, or other campaigns kept outside of automated management will not be considered part of the performance budget.

For non-AlgoLift managed ad networks, clients should aim to pace toward the platform/channel allocations as close as possible. Of course, there will be cases where this is not possible to achieve - in these scenarios, reoptimized budget allocations will factor in the spend pacing to date for each client-managed channel.