Criteo AOV

aov

Overview

Criteo Average Order Value (Criteo AOV) calculates the average amount spent by customers per order, after shipping and taxes, as reported by Criteo.

๐Ÿ“˜

Criteo AOV = Criteo-Reported CV รท Criteo-Reported Conversions

The calculation is based on data from the Ads table.


Detailed Breakdown

The formula above is derived from the following components:

  • Conversion Value (CV) = SUM(conversion_value) --> Ads table

  • Conversions = SUM(conversions) --> Ads table

  • Criteo Ads = where channel = 'criteo'

Insights and Actions

Criteo Average Order Value (AOV) measures the average revenue generated per order from Criteo-driven sales, helping to evaluate customer purchasing behavior and campaign impact:

  • Analyze Purchase Trends: Track Criteo AOV over time to identify changes in customer spending habits, guiding pricing and promotional strategies.
  • Target High-Value Customers: Use AOV data to refine audience targeting, focusing on segments that generate higher-value orders.
  • Improve Campaign Profitability: Pair AOV with conversion data to assess the impact of upselling or bundling strategies, helping to increase the average order size and overall revenue.

Related Metrics

  • Criteo Conversion Value (CV): Shows the total revenue generated by Criteo campaigns, essential for calculating AOV and evaluating revenue impact.
  • Criteo Conversions: Tracks the total number of orders, providing context for interpreting AOV and understanding customer behavior.
  • Criteo ROAS: Complements AOV by revealing revenue efficiency relative to ad spend, offering insights into overall campaign profitability.

Example Use

Prompt

What's my Criteo-reported AOV for the last 90 days?

Response

Query

SELECT
  COALESCE(
    SUM(pjt.channel_reported_conversion_value) / NULLIF(SUM(pjt.channel_reported_conversions), 0),
    0
  ) AS channel_reported_aov
FROM
  pixel_joined_tvf () AS pjt
WHERE
  pjt.event_date BETWEEN CURRENT_DATE() - 90 AND CURRENT_DATE()  - 1
  AND pjt.channel = 'criteo'
  AND pjt.model = 'Triple Attribution'