Using iROI to optimize retail media alongside promotions

Retail media spending is almost always run alongside promotions (either as a straight discount or buy 1 get one free, etc.) or subscriptions such as Amazon’s Subscribe & Save. Oftentimes individuals both click through on a Sponsored Product or Sponsored Brand ad AND convert on a product that has a promotion running on it. Given this overlap in tactics, how do we understand how the two can work together to unlock incremental growth?

When the two are run together effectively, they can be a force multiplier by using advertising to bring the product to new potential buyers and using promotion to entice them into a trial.

When they are not harmonized, it can lead to inflated ROAS as advertising takes credit for conversions being driven across the line by the promotion. This in turn can lead to wasting budget and subsidizing what would be organic sales. We need to disentangle the influence on driving incremental sales growth to understand how these can be used together effectively.

Here’s a real-world example.

To bring this to life, let’s look at some real data for a mid-sized CPG brand. They were selling through Amazon, Walmart.com, and their own DTC site with the bulk of their sales coming from Amazon.

The brand had a consistent promotional spend on Amazon advertising efforts except for November. They ran heavy discounts across numerous SKUs to try to entice new trials going into the holidays. During this period the cost of promotions (discounted revenue) was nearly 3x of the monthly average. In parallel, they were running evergreen Sponsored Product and Brand Ads on Amazon along with some off-platform digital spending on YouTube and Meta to build awareness and drive traffic to these ecommerce platforms.

Looking at reported ad-attributed sales and the corresponding ROAS of their spending on Amazon would suggest that running the promotions increased the efficiency of their advertising spend with overall ROAS nearly doubling during the heavily promoted month of November.

As we’ve noted before, retail media spend is almost exclusively measured through last touch attribution which doesn’t control for other factors such as promotion (or off-platform ad-spend, seasonality, etc. for that matter). This makes retail media particularly susceptible to false signals of incrementality by mistaking correlation for causality. Other factors such as promotion or seasonality may be driving demand and conversion at higher rates but retail media just happens to be reaching these same users.

Measuring Retail Media & Promotions for Incremental Growth

We put the same brand’s data through our Incrementality model, which controlled for promotional spending, their off-platform investments in YouTube and Meta as well as macro factors like seasonality and holidays. The results suggest a very different view of performance during this period.

During November where last touch attribution indicated a more than doubling of ROAS, iROI (incremental return on investment) actually declined. This would suggest that the type of advertising efforts during this period were non-incremental, they were reaching users who were already very likely to convert on their own due to outside factors like promotion.

This can easily be seen in the increase in the proportion of incremental sales being driven by promotion relative to ads, with the number of incremental sales coming from promotions nearly tripling in November.

Why this happens, shouldn’t be too surprising given the flywheel nature of ecommerce. The combination of discounted pricing coming from the promotion as well as the increased velocity improved the brand’s organic search rank bringing it above the fold on numerous products and lessening the need for advertising to capture that same digital shelf space.

Planning Retail Media WITH Promotions to Drive More Growth

Had the brand had visibility into this dynamic, how could they have invested in their campaigns differently to take advantage of these promotions? By measuring and isolating the incremental growth that comes from retail media alongside promotional activity, retail media can be planned in a way that takes advantage of this flywheel effect to further accelerate growth. Let's take a closer look at how iROI changed for a few ever-green campaigns during that heavy promotional period of November.

‍

When comparing the iROI of November (when the promotions were running at their highest) to the iROI of the rest of the year, a few opportunities jump out.

• The iROI of mid and upper-funnel campaigns like their off-platform spend on Amazon and Meta and their on-platform spend on Sponsored Brand improved during the promotional period. This should intuitively make sense as these tactics which would not typically drive immediate conversion, perform better in the presence of promotions that can take that awareness or interest in the brand and immediately convert it.
• On the other side, campaigns targeting branded keywords performed much worse during the high promotional period. This would indicate that the promotions were likely doing the heavy lifting. These users were already likely to convert, and promotions pushed them over the line. Additionally, the combination of discounted pricing coming from the promotion as well as the increased velocity improved the brand’s organic search rank against these branded keywords bringing them well above the fold on numerous products and lessening the need for advertising to protect their digital shelf during this period.

With this knowledge, the brand could have reallocated its budgets between campaigns to drive 2-3x as many incremental sales from their advertising while improving the efficiency of this spend. While this is just one example, it clearly illustrates the importance of more robust measurement in retail media and the competitive advantage that better measurement can unlock for brands.

‍