This case is about our MLPLA software. The Machine Learning Case Study gives you an overview of the benefits that MLPLA has given to the Garden Orchid UK website operated by Phytesia.

About Phytesia

With approximately 500k € in annual online revenue, 10 employees, one breeding lab, 3 distribution platforms and 3 transit points, Phytesia is the leading producer and distributor of hardy orchids in Europe.

eCommerce Merchandising Facts

  • Due to the significant range of orchids accessories, it creates inventory risk;
  • To update the (enriched) content of the extensive product catalog is very labor intensive;
  • How to find the right keywords in an ever changing online market becomes challenging;
  • Despite all efforts, current site traffic continues to be flat and too low..

Project and Objectives a Commerce Machine Learning project

UK and Germany are the top markets for flowers purchasing in Europe. Though they are very large, they are also difficult to enter. Prospect visitors expect a large choice of products as well as high informational quality content. Besides this, they won’t visit a web site unless it’s visible on either Google, Bing or Yahoo preferably on the first page.

Phytesia had grown in recent years, mainly through an aggressive commercial policy towards wholesalers, which resulted in a significant increase in the number of products to manage. Each product had a different set of logistics, requiring Phytesia to reorganize its network and processes. At first, a half time staff was dedicated to enrich content data as well as organizing PPC ads and weekly newsletters.

However, Phytesia’s competitive advantage comes from having the best array of rare orchids. The complexity of Phytesia lies in the fact that its brand is relatively unknown and most of the gardening hobbyists only do know about tulips, roses and begonias. Therefore, Phytesia decided to request LR Physics’ help in order to build an innovative and highly reliable aggregated website (i.e. pumper site), which should enforce the local UK identity and also should be able to present as many orchids growing products as possible.

Day-to-Day Commerce Machine Learning

Phytesia’s daily marketing tasks are mostly about gathering and analysing consumer behaviour data on its website by using Google Analytics. Until recently, all products showing traffic traction were closely looked at. Every day, a copywriter spent some hours to improve the current product content. The remaining part of the time was spent on publishing the targeted product into Social networks and implement PPC campaigns to aim for better site traffic and related purchase conversion.

In addition, Phytesia had to provide product information to hundreds of retailers. The Demand Planning process includes collecting information from these field operations, including promotions, assortments and seasonality to define a demand plan which converts dynamically into a production and corresponding distribution plan.

Today Phytesia manages this process with the Mash’n Learn MLPLA (Machine Learning Product Listing Advertising), which is capable of identifying, analyzing and enriching product content based on gathered information over the internet. To manage the keyword analysis, MLPLA uses the latest IBM Watson built in technologies.
The set of APIs are analysing users feeling about the product keywords and then mining the web and Phytesia’s library for text. The goal is to make sure all key products are ranking as high as possible on Google search engine.

Results and Benefits in a Machine Learning case

  • Mash’n Learn was able to identify the  top keywords based on Product information;
  • The MLPLA tool set  based on machine learning technology translates keywords data into usable quality information product content;
  • Phytesia brought its UK organic daily visits from 20 to 300, tripled the sales revenues and introduced 26 orchids accessories and bundles;
  • All Phytesia products reached page one of Google and reached twice the top 10 Home & Garden sales on Amazon UK;
  • The Product content management workload was reduced from 1.5 to .5 FTE, creating the opportunity for the Marketing staff to focus on A/B testing and new advertising opportunities;
  • A significant improvement on the conversion rate were realised as the content quality reached up to 10/10 on Google Adwords. Therefore more and more qualified organic traffic caused a doubling of the conversion rate which resulted in revenue increases up to 450% in the UK only;

MLPLA software is used for managing Product information and to provide content to wholesalers. Online direct sales are having an indirect effect on wholesale as Gardening retailers request for Phytesia product is now rising as it outbids the biggest Home and Garden retail chains in the UK. Since growth is tangible, MLPLA proved to be a fundamental tool to support decision-making at the strategic, tactical and operational levels, all with the same base data.

MLPLA is a turnkey software that can easily be implemented within a maximum of 10 days depending on the company’s eCommerce and / or ERP environment. The MLPLA’s ‘Pop Up’ functionality can also create an instant store from scratch on a WooCommerce (WordPress backbone) instance.

About Mash’n Learn

Mash’n Learn (ML) is a Tool Suite based on machine learning designed and developed by LR Physics (see also

The ML Tool Suite uses all of the latest science and innovations on Machine Learning and is taking advantage on the R&D capabilities of LR Physics. An extensive range of repetitive tasks could be optimized by a powerful Machine Learning set of algorithm. The innovative and advanced technologies of ML enable you to daily brass thousands of extensive Product Data and even fully automates your end-to-end online marketing processes.

Mash’n Learn’s solutions span key marketing platforms areas such as but not limited to Google PLA, Amazon Seller Central, Facebook Products and Pinterest integration. Since April 2016, Mash’n Learn gained the trust of various online retailers with whom we’re partnering today in order to get more qualified traffic and related sales revenue. The machine learning Tool Suite also includes Artificial Intelligence integration for Supply Chain Demand Management, Sourcing and ERP Product Catalog.