By clicking “Accept”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.
July 7, 2021

Isahit supports ManoMano in obtaining quality product data

July 7, 2021

Isahit supports ManoMano in obtaining quality product data

Europe's leader DIY marketplace ManoMano and isahit have joined forces to address the following issue: How to get quality product data to ensure the best possible user experience ?

ManoMano in a few words:

  • European Leader specialized in DIY & Gardening
  • 3600 merchants around the world
  • 10 million references
  • Markets: France, Belgium, Spain, Italy, UK and Germany
  • 50 million visitors per month

Isahit in a nutshell:

  • First ethical data labeling platform for AI and data processing
  • 1400 projects with more than 1500 HITers
  • HITers coming from 37 countries
  • 250+ satisfied customers

Identification of the project main phases

In order to respond to the identified problem: " How to get quality product data to ensure the best possible user experience ? ManoMano has identified four main processes:

  • Matching (identifying similar products)
  • Categorization (grouping by product family)
  • Enrichment (finding additional product information)
  • Moderation (compliance with the editorial charter for product reviews and comments)

Before the collaboration with isahit, ManoMano was only using automated processes such as machine learning and RPA to improve the quality of their data. However, these processes did not allow them to achieve 100% data completion on the defined perimeters.

To achieve their goals, ManoMano called on the services of isahit.

Indeed,Thanks to the human intervention inherent to its solution, isahit is able to complete the data provided by the automated processes and guarantee 100% data completion on the defined perimeter. Indeed, only human intervention can tackle the following missions : train the algorithms, validate the predictions of the algorithms, and clean up/enrich unprocessed data.


Unfortunately, ManoMano's automated solutions could not handle all the tasks for the following reasons:

  • High subjectivity
  • Lack of homogeneity
  • Multitude of data sources (+3000)
  • Linguistic issues (+5 languages)
  • Product codification not standardized

By offering a trained and selected workforce based on the required skills (language, experience, market knowledge, performance...) for a given project, isahit had the most appropriate solution to support ManoMano.

Processes identify and set up for the successful completion of the project

In order to ensure that the project was carried out correctly, ManoMano's and isahit's teams set up specific and precise processes. ManoMano made sure to provide isahit with the following elements before the launch:

  • SOP Writing (Standards Operation Guidelines)
  • Preparation of the data not processed through automated solutions (batches or on the flow)
  • Objectives definition: timelines & quality

The isahit team assigned the project to a dedicated manager and implemented the following processes:

  • HITer selection based on the project’s requirements
  • Project Manager training with ManoMano and finalisation of the SOP
  • Training of HITers
  • Test phases to validate quality & targeted timelines

Project monitoring specific to ManoMano's needs

In order to ensure the most efficient project monitoring, isahit has developed a monitoring strategy specific to ManoMano needs:

  • Test phases to define quality objectives and deadlines
  • Monitoring of financial costs from isahit platform
  • Weekly meeting with the project manager
  • Quarterly review with all the interlocutors (Sales, ops, community..)

This strategy allowed to continuously monitor the following KPIs: Project progress, quality rate, average time per task, volumes, performance, financial ROI.

A double victory: Operational & Social

One and a half year after the beginning of our collaboration, we are mutually very happy to make the following assessment:

  • Several million tasks completed
  • 5 languages covered
  • A scalability ratio of 1.7 per month

... and a positive and lasting impact on nearly 50 of our HITers!

You might also like
this new related posts

Want to scale up your data labeling projects
and do it ethically? 

We have a wide range of solutions and tools that will help you train your algorithms. Click below to learn more!