Sneak Peak
What Marveltest can mean for the Fashion sector
a MarvelFashion Client Case
Marveltest combines different analyses to obtain the head start you're looking for!
Week 1
An insightful cooperation between data and fashion
Marveltest is a data driven marketing company that focuses on growth. Besides the experience and results in the Fashion sector, Marveltest accelerates online testing, optimisation and transitions.
Stadhouderskade 60A, 1072 AC, Amsterdam
AMSTERDAM
© 2020 Marveltest. All rights reserved
FOLLOW US ON
LinkedIn
ABOUT US
Marveltest has been guiding business decisions data-driven since our birth in the spring 2018.
Ons team werkend vanaf thuis, bereikbaar via 085-1302196.
Company Website
The second month of the HighCapacityTesting Algorithm for one of our clients
MarvelFashion
1. Historical Yardstick Data analysis
2. Developed and modelled 4 scenario's
-40% Marketing Cost
+98% revenue
The return on ad spend went from 3.2 to 8.5
1. Marveltest Scraper gathered competitor data
Speed went up, reaching the goal within 2 months
1. Our systems ran through historical marketing efforts on the channels
Average margin on orders went from 35% to 65%, with AOV: +32%
1. Machine learning algorithm discovered
Week 2
Week 3
Week 4
Marveltest's own created HighCapacityTesting algorithm finds your right channel, target audience and proposition to sell your products! Through 1.000 experiments Marveltest is able to validate different assumptions about your brand in a data driven way.
We lowered the cost-per-click by 41%, which resulted in reaching the target revenue with 41% less marketing spend.
The customer acquisition cost lowered drastically which resulted in a return on ad spend on ad sets that was a staggering 11,48.
We pushed the revenue in the first month to the target: doubling the revenue. This while marketing cost went down with 41%.
As our machine learning algorithm produced audiences that have higher AOV's, we reached an AOV increase of 38%
Online marketing is about data not about feeling!
The marketing department preferred visual 1, while creative went for visual 2. But our data suggested after the first validation round that visual 3 converted two times better!
That is the funny thing about data, you will be surprised every time.
Learnings: Visual 3 has more contrast than visual 2 and the model looks in the camera while the model in visual 1 does not.