flaconi is Germany’s largest online-only retailer for beauty products. The online shop’s extensive portfolio is made up of over 850 international brands and more than 55,000 products. From natural cosmetics to premium products, flaconi fills drugstore shelves with everything from perfumes, to make-up products, to hair-care products, to accessories.
Results
- High-quality real-time data
- Flexibility in structuring raw data exports
- Real-time data transfer for strategic real-time alerts
- React to user interactions quickly and in an extremely personalized way
- Provide an outstanding customer experience
The decision to go with Data Streams was a no brainer. Modern online business is highly data-driven and needs to be managed in real time. Our vision is to be able to offer our customers a personalized user experience based on analytics and machine learning and to reach an ideal level of automated marketing outreach by using data from real-time streams.
MARTIN NGUYEN
DIRECTOR STRATEGY, DATA & ANALYTICS
Challenges
NEEDED TO CHANNEL RAW DATA TO MAKE IT ACTIONABLE
The online market for perfumes and cosmetics is characterized by a high level of competition. The products from different retailers can be easily substituted for one another. Many end-users have a clear idea of which beauty products and brands they are looking for. That’s why, for flaconi, it’s all about finding the right sales trigger to stand out and following a principle that offers everyone products for a dedicated spot in their bathroom cabinet: Find Your Beauty. Everyday. Of course, it also helps to have competitive prices and the right product recommendations to set yourself apart from the competition.
flaconi also possesses an extensive amount of raw analytical data. Success hinges on identifying how users navigate the website, as well as distributing and processing this information internally. Providing a relevant and personalized experience for each individual user is based on successfully managing the transfer and processing of the raw data within the company. The biggest challenge for flaconi’s data analysts was, then, how to analyze the data even quicker. They needed to channel this raw analytical data and make it usable in real time.
Solutions
UNLOCKING TARGETED PERSONALIZATION WITH DATA STREAMS
flaconi took up this challenge together with Mapp, using Data Streams as the solution. Data Streams are raw analytical data that are event driven and exported to data warehouses and BI systems in real time. Apache Kafka provides flaconi with a dedicated infrastructure for further processing. However, the possibilities afforded by this infrastructure for the further processing of raw data have not yet been pushed to the limit. Data is processed using listeners based on Kafka or with Kafka Connect. This makes it possible for flaconi to transmit raw analytical data to their own data warehouse or to process it directly on the website server. The data is also passed on to Amazon S3 and saved there for other purposes.
High-Quality Data Guaranteed
Every single user interaction on flaconi’s website is collected and processed. Besides clicks and purchases, information such as a user calling up a specific product image, for example, is also taken into account. All this information is processed in real time. Catalogue pages with several dozens of product listings are taken into account too. The dynamic product sorting means that flaconi also faces the challenge of identifying, after the fact, what information the user will be shown and how they will react to the respective product information.
Existing Setup Can Be Used
The switch to Data Streams was implemented using the standard method. This means that flaconi benefited from the existing frontend that was already designed to structure raw data ahead of time. flaconi data analysts are able to choose what data they would like to use for Data Streams and in which form. This gives flaconi a large amount of flexibility in terms of structuring raw data exports for further processing.
Real-time Handling for the Most Intensive Demands
Besides high-quality data, real-time data transfer was extremely important to flaconi. That’s why real-time alerts were set up with Data Streams. This allowed flaconi to react to pre-defined events faster. They are now able to set specific events for which analysis reports are to be sent automatically. Decisions are being made on the basis of this information.
When there are significant traffic spikes on the website, for example, the data analysts cannot be certain whether it’s the result of an influencer or a comparable external campaign. Therefore, real-time alerts make it possible to temporarily adjust prices on the website or to activate a landing page, which intercepts influencer communication. This allows flaconi to significantly shorten the time it takes for their customers to make decisions.
With real-time alerts, there is a particular emphasis on relevance. Sending too many notifications can reduce recipient acceptance and should therefore be avoided. That’s why use cases are specified in order to trigger a real-time reaction. The alerts are based on pre-defined threshold values for when sendouts take place automatically.
Over time, flaconi has become better able to define bandwidths to fine tune its reaction. Normally, a responsible manager from first-level support receives the real-time alert. This manager then decides whether and how to react. Numerous reactions have been pre-defined in order to manage and plan common reaction scenarios. Real-time alerts mean flaconi is able to act in the best interest of their customers and provide an outstanding customer experience. For example, if it can be foreseen that a certain product will sell out very quickly and there are only a limited number available in the warehouse, the price can be adjusted at short notice. For an online shop for beauty products, this means that in the future the product on the website will be more likely to be shown to the user and they won’t be disappointed by a lack of availability.
Website Personalization with Live Data
Another central aspect is real-time personalization with Data Streams. flaconi uses live data to target each individual user on the website in a personalized manner. This makes it possible to easily differentiate between men and women in real time, for example. Mapp’s raw data exports ensure that flaconi is able to react to user behavior extremely quickly. Functionality like this is crucial for the business, because it prevents users from moving to a competitor and purchasing products from them.
What's next?
The bottom line
Thanks to Mapp’s Data Streams, flaconi can react to user interactions quickly and in an extremely personalized way. Just as before, the raw analytical data collected is used and processed. Data Streams has an impressive and strong professional design and can be easily structured and transferred thanks to the high-performance infrastructure. flaconi benefits from Mapp’s extensive experience and many years of product development. A tremendous amount of resources and manpower would be required for flaconi to build up and develop these competencies internally.
Platform
Channels
Services
Clients
Partners
Blog
Resources
Learning
About us
DE
