top of page

AI Case Study

Cosabella increased ROAS to 336% by segmenting target markets, and optimising channel selection, campaign content, and paid search using Adgorithm's deep learning

Cosabella uses an AI platform to optimise its online marketing mix by hyper-segmenting customers, choosing channels, optimising paid search, picking up on competitor trends and automating budget decisions, resulting in and ROAS increase to 336% and meeting all KPI targets.


Consumer Goods And Services

Retail General

Project Overview

From Campaign: "Adgorithm’s Albert was tasked with identifying and converting high-value audiences. It was given Cosabella’s paid search and social media marketing efforts to manage and was allowed to autonomously execute Cosabella’s digital marketing efforts using creative and KPIs provided by the brand. Albert is capable of moving budgets into any of the areas he manages, and if ROAS is routinely above target, he’ll recommend a budget increase. Albert will also analyse the keywords used by competitors and suggest if the brand needs to start targeting those areas. 'He’s notified us that all our competitors are running promotions at a time we weren’t, so we should probably get on that... One of the things Cosabella has benefited most from is Albert’s ability to detect "fatigue"... a creative concept is ‘fatiguing’, that is, its click rate is going down and interactions are dropping'." Alberta was used to hypersegment the customer base into hundreds of personas based on data patterns.

Apparel reports that "Albert tests creative content in seemingly endless combination in search, social and display channels, incorporating his learnings along the way and optimizing ad spend to deliver exponential results. Implementation took less than a week and Albert took a few weeks after "ingesting" all of Cosabella's creative to analyze the content and optimize it for new campaigns. Now Cosabella runs one "campaign" per country, so in the United States, for example, Connell will give Albert a budget, tell him who to target (new customers, lookalikes, etc.), and which channels: search, social and display."

Reported Results

Campaign claims that within the first month, "Albert increased Cosabella’s search and social return-on-ad-spend (ROAS) by 50% and decreased its adspend by 12%. On Facebook specifically, Albert produced a 565% increase in ROAS within his first month. By month three, Albert had increased Cosabella’s ROAS to 336% – a 155% increase over the previous quarter. Overall, Albert’s work in the last quarter of 2016 contributed strongly towards the company’s 37% increase in overall website sessions, a 30% increase in new users, and 1,500 more transactions."




Channel Marketing


Cosabella is an e-retailer in lingerie which operates in several countries. After a flat sales quarter the company decided to move from a traditional ad agency to an AI platform.



Unspecified: KPI data, "creative materials"

bottom of page