When marketers activate augmented reality advertising in their media plans, we often hear that they’re unsure how to assess the performance of this newer experience relative to more traditional media channels like television and online video. While Snap offers a robust suite of measurement capabilities across our ad products, we understand the value that cross-media measurement provides marketers.
Enter marketing mix models (MMM) — an important tool to properly evaluate a wide range of marketing activations. Snap is highly supportive of brands’ MMM efforts, and we partner with key third-party MMM suppliers to ensure campaign outputs include the granularity and specificity to properly measure Snapchat’s advertising contribution to sales and return on investment.
Snap commissioned Nielsen to identify whether isolating AR as its own channel within marketing mix models could provide valuable insight to marketers. If Nielsen did in fact find that evaluating AR on its own was warranted, we also tasked them with identifying best practices for accurately measuring AR within an MMM. To achieve this, Nielsen evaluated the performance of Snap Lenses AR experiences from fifteen brands across three CPG subcategories and three countries: the US, the UK, and Saudi Arabia. Nielsen built large training data sets for each brand, containing three years of sales, media spend, brand-health tracking, and other data, along with individual advanced econometric models.
Across countries, brands, and subcategories, compelling insights speak to the full-funnel opportunity augmented reality offers brands. When modeling to YouGov BrandIndex 1 tracking data, Snap AR Lenses showed high awareness effectiveness, consistently outperforming total media. Snap AR’s return on ad spend (ROAS) averaged across the fifteen brands was $1.67 — notably higher than TV, Digital, or Total Media performance averaged across these same brands.
Nielsen identified best practices for how to include AR as its own touchpoint within an MMM. Through a variety of transformations, KPIs, and tests, Nielsen determined that AR executions should be modeled in paid impressions as their own entity in MMM and should not be combined within other digital or social tactics. Paid impressions showed high model explanatory power in the models Nielsen built — however, they suggested that if a modeler is experiencing poor model fit or there is no observed relationship with sales, a transformation could improve model fit. This could be paid impressions multiplied by share rate, or share rate or save rate itself. Nielsen also found that earned AR impressions should be measured separately to gain the most accurate view of AR advertising’s impact.
For Nielsen’s full paper, click the link below to download. Please reach out to your Snap marketing science partner to request the proper data about how to include Snap AR in your marketing mix models.