
Fashion eCommerce
Validating Marketing Mix Model with Meta Conversion Lift Studies

Meta MMM ROI
ROAS Lift During CLS
Incr. ROI is between 2.91 and 5.09
Client Context
BrandAlley is a UK-based online fashion eCommerce specializing in flash sales of premium and luxury brands.
Established in 2008 as a joint venture between News International and BrandAlley France, the company offers members exclusive, time-limited discounts on a wide range of products, including womenswear, menswear, children’s collections, homeware, and beauty items.
The platform launches multiple flash sales daily, each typically lasting a few days, allowing customers to access high-end products at significantly reduced prices.

Client Requirements
As BrandAlley’s marketing team launches multiple flash sales on a daily basis, the yearly total sums up to over 1000 campaigns. At the same time, the campaigns are very short-lived, putting extra emphasis on real-time insights.
BrandAlley’s team was especially interested in:
- Making sure the model captures the incrementality of different media channels accurately amidst all the noise in the data
- Building a holistic, real-time insights that help the team to manage the myriad of campaigns

Meta Conversion Lift Study
To confirm that BrandAlley’s MMM was capturing the incremental uplifts from Meta accurately, the team decided to run a Conversion Lift Study (CLS) on Meta.
In a CLS, people who matches the ad audience is randomly split into two groups, where one is intentionally withheld from seeing the ads and another is exposed to the ads.
Comparing the outcome between the two groups reveals the true value of Meta’s incremental impact, hence allowing us to validate if MMM is accurately capturing the impact from Meta ads.

BrandAlley's Meta Conversion Lift Study
Specifications of the test:
- Duration: 4 weeks
- Included Advertising Channels: All
- Included Campaigns: All
- Primary objective: Purchases
Comparing the sales between these two groups, the CLS found an incremental ROAS lift for Meta Ads to be 4.00x.

Validating MMM Results with a Lift Test
- Given that 3.91 is well within the 90% confidence interval, the test then confirms that Sellforte’s MMM results are in the correct ballpark and can be used with confidence in media planning

Tailoring MMM insights for flash sales
BrandAlley’s campaign frequency and quantity is very high due to the nature of it’s business.
Ongoing analyses and feedback on optimization efforts is a necessity for BrandAlley and Sellforte’s continuous MMM provides just that. With daily updates, the team is able to make real-time adjustments to their marketing tactics, based on the most current data.
In addition, BrandAlley was able to add custom dimensions (e.g. Brand tiers, Campaign types: Category-focused, Brand-focused…), on top of the many key default dimensions, to personalize and further solidify the insights they were getting from Sellforte.

It’s been great how the Sellforte team has been able to adapt to our business model and help us to validate the insights through the Lift Test. We have now a solid foundation to optimise our marketing campaigns and make decisions about the budget

Richard Lucas
Head of Performance Marketing @ BrandAlley