Comparison Shopping Platforms: Brand and Category Demand Tracking and Prediction Tool

Data about consumer search behavior from comparison shopping applications have the highest validity for predicting eventual consumer shopping behavior. It is available even before real-time sales actually happen, have a high correlation with actual POS selling data and it is as deep as the SKU level.

CENEJE is part of Heureka Group, which is Europe’s largest price comparison website and online shopping advisor, has a presence in 9 countries, and provides 20M monthly users with access to product offers advertised by 55K merchants. Websites from the Heureka Group, including CENEJE, list 30M products organized in 2K categories. It addresses 2M daily unique visitors in the pre-purchase and purchase stage deep into the funnel, collecting 2-3B data points from the funnel yearly. 

“Brand Performance Insights” is the solution developed by CENEJE, in collaboration with EW-Shopp, to help shoppers (Shopper’s Insights), sellers (Seller’s Insights) and vendors (Vendor’s Insights) to use short term product search insights and behavior predictions to support their smart purchase or business decisions. The tool is built on top of years of users’ uniquely integrated behavior data about searching products, and different types of business events.

WATCH THE VIDEO: Brand Performance Insights: predicting shopper interaction with advanced data management

KEY ASSETS

  • Shoppers are able to use relevant purchase information and real-time predictions to support their purchase decisions.
  • Sellers manage better pricing, category, and campaign management.
  • Vendors get brand and SKU level purchase information in the shortest possible time to react and improve pricing, sales, and campaign efficiency.

THE APPLICATION

Brand Performance Insights collects information about users’ searches and clicks on vendors’ offers – indicators of the performance of the advertised products – and analyze their evolution at different levels of aggregation (SKU, brand, category, vendor, price). It also derives event information by spotting sales and price changes, so as to make it possible to analyze the impact on these events on the market. A predictive module also estimates how the market reacts to certain business events like price changes, marketing campaigns, new product introductions, leading towards what-if analyses. All the analytics are available through a web app (a screenshot is reported in the picture) and via APIs for data exchange with third parties like vendors and sellers. In the web app, analytics are enhanced by responsive visualizations that support users in quickly get insights into ongoing trends with possibilities to drill down the events to relevant product depths. The application is fed daily with fresh data fetched directly from CENEJE’s website. But not just that, the product expanded and we included the product/category demand data for eight Central and Eastern European countries (Czechia, Slovakia, Hungaria, Rumania, Slovenia, Croatia, Serbia, and Bulgaria).    

 

The key business areas covered by Brand Performance Insights are:

  • Brand performance management & benchmarking, where vendors and merchants deep dive into the brand position and shares from week to week or month to month
  • Assortment & category management, where the key is to identify – almost in real-time – top-selling products in the market or product features that lead the demand (especially to manage highly demanded but missing products within the assortment)
  • Category seasonality and impact of marketing events, which allows you to preplan activities that would allow you to harness all the available event-based demand on the market efficiently
  • Marketing efficiency tracking and evaluation, a unique and prime opportunity for vendors and merchants to track the performance of their campaign through user response with top one day delay
  • Price positioning and elasticity to manage the competitive position of the merchants in the omnichannel market, where setting up responsive processes and dynamic pricing to get most of the demand is a crucial ability of top players on the market
  • Prediction of short term search demand is the most difficult task in complex marketplace environments, therefore being able to estimate the direction and impact of actions on demand could give you key efficiency gains in the business

The service has been built by CENEJE with the support or EW-Shopp partners, in particular for event-based prediction components. 

ACHIEVED IMPACT

Brand performance Insights supports better tactical business decisions such as optimal sales promotion, marketing investment, category and product management, pricing. It impacts in fact holistically on all market players, i.e., Shoppers, Vendors, and Merchants.

  • Shoppers use the widget to get more info about the searched products and consequently make more informed purchase decisions and gain trust.
  • Vendors gain access to unique almost real-time insights about the purchase behavior and demand predictions for a particular category or brand.
  • Merchants get a better understanding in the category and SKU related sales dynamics through tracking analytics and forecast based sales/product events and relevant forecasted weather.

Brand Performance Insights has been in use since January 2020 with 15k activities recorded in the web application by 10 external partners from different industries by the end of February 2020. In April and May 2020, we grew the number of users on the vendor or merchant side to more than 200 in four different countries. It became a tool that showed the direction in the particularly difficult times of Covid-19 market shifts and enable clients quick responses to fast-changing demand.  

Additionally, we use the application internally to support decisions related to partner management, consultative selling and building relationships with clients. We constantly improve the product and we also gained high-value evaluations through external user research in May 2020, which complimented the tool for the ease of work, design, and clarity. 

IMPACTED INDUSTRIES 

  • Retail
  • E-commerce
  • Brand Manufacturers / Distributors

RESEARCH & INNOVATION DOMAIN: Data integration, data aggregation, modeling, predictions, data analytics, web design

TOOLS INVOLVED: Qminer, MariaDb, MsSql

CONTACT PERSON: Darko Dujic darko@ceneje.si