[tatsu_icon_group alignment = "center" ][tatsu_icon name = "icon-facebook2" size = "tiny" style = "plain" bg_color = "" hover_bg_color = "" color = "rgba(255,255,255,1)" hover_color = "" border_width = "" border_color = "" hover_border_color = "" alignment = "none" href = "" new_tab = "0" lightbox = "0" image = "" video_url = "" animate = "1" animation_type = "fadeInUp" animation_delay = "0" ][/tatsu_icon][tatsu_icon name = "icon-social_instagram" size = "tiny" style = "plain" bg_color = "" hover_bg_color = "" color = "rgba(255,255,255,1)" hover_color = "" border_width = "" border_color = "" hover_border_color = "" alignment = "none" href = "" new_tab = "0" lightbox = "0" image = "" video_url = "" animate = "1" animation_type = "fadeInUp" animation_delay = "200" ][/tatsu_icon][tatsu_icon name = "icon-social_pinterest" size = "tiny" style = "plain" bg_color = "" hover_bg_color = "" color = "rgba(255,255,255,1)" hover_color = "" border_width = "" border_color = "" hover_border_color = "" alignment = "none" href = "" new_tab = "0" lightbox = "0" image = "" video_url = "" animate = "1" animation_type = "fadeInUp" animation_delay = "400" ][/tatsu_icon][tatsu_icon name = "icon-behance" size = "tiny" style = "plain" bg_color = "" hover_bg_color = "" color = "rgba(255,255,255,1)" hover_color = "" border_width = "" border_color = "" hover_border_color = "" alignment = "none" href = "" new_tab = "0" lightbox = "0" image = "" video_url = "" animate = "1" animation_type = "fadeInUp" animation_delay = "400" ][/tatsu_icon][/tatsu_icon_group]

Data Science

Statistics are a form of wish fulfillment, just like dreams.
Jean Baudrillard

We live stunned by a tsunami of disparate data that threatens to leave us passive and confused in the face of a landscape so complex it seems indecipherable. Yet the numbers are singing! Statistics, Artificial Intelligence and predictive analytics, combined with outstanding data visualization allow us to be aware of what is happening and what is plausibly going to happen.

BI Second Drop

Drop has invested heavily in Data Science in recent years. We started in 2012 by collaborating with the Polytechnic University of Turin on a Big Data analytics project, which then became over the years the core of Drop.BI, a leading cloud-based platform for Business Intelligence for online retail.
Drop.BI is a platform that provides real time maximum control over all the KPIs of an e-commerce business through a series of tools and dashboards that bring order to the large amount and raw heterogeneous data. For more specific insights Drop performs dedicated analysis starting from data extraction through direct action on DBs to visual reprocessing of these through platforms such as MS Power BI.

Data Science and e-commerce

Data Science processes aggregate data to describe the past and give insights into the future: customer lifetime value, analysis of consumers and their behavior, churn detection, customer segmentation, cohort analysis, and consumer trend analysis. If we have massive amounts of data available, well: we will make this data sing! Twenty years ago we were catching up with the university heritage of Statistics, Numerical Computing, Automated Controls. As the numbers grew, in a natural way, we approached Big Data and the world of Data Science. Today we are moving, driven by Machine Learning and Deep Learning toward Artificial Intelligence applied to so many aspects of e-commerce:

  • recommendations
  • dynamic pricing
  • search
  • anomaly and fraud detection
  • chat bot
  • UI dynamic personalization

The direction is e-commerce management based on collaboration between human intelligence and AI (e.g., through Machine Learning Control).

Drop Method: Control Cycle

Good management is based on a repeating cycle, a cycle made up of tests and iterations. We like to recall on this theme the theory of automatic controls (the first example being the centrifugal controller of 1788) In automation science, the automatic control of a given dynamic system (e.g., an industrial plant or - today - an e-commerce) aims to change the behavior of the system to be controlled (e.g., turnover) through the manipulation of appropriate input quantities (e.g., product price or, not talking about quantities but conditions: type of checkout, UI of the product card, etc.). Today, in an online store the cycle sees as a determining figure the e-commerce manager who based on reading the data (i.e., doing BI) makes decisions and intervenes on the e-commerce e-cosystem. Just as Frank Zappa in some ways preferred electronic equipment to musicians, today we know that AI in the future will increasingly help the e-commerce manager and that the above cycle will be optimized and closed by Machine Learning Control. Changes in price, changes in UX, changes on amount or type of adv investment, all these and more cause performance changes in your e-commerce. The e-commerce manager must be able to quickly assess the effects of these interventions by performing A/B testing, sensitivity analysis. For this to happen in the best way possible, an analysis system, a business intelligence system, is needed. The improvement in our opinion lies on the ability that the team has to test continuously, in an "agile" manner proceeding by successive experiments and going quickly to the deployment of evolutionary interventions.

    Leave us your details, we will contact you back for a chat.

    * Fields marked with asterisks are required.

    I have read thePrivacy Policy and acknowledge that I understand and agree to its contents and terms.

    • Data Visualization

      There is a certain beauty in numbers. Think of fractals, the shapes given by reflection in a kaleidoscope. Nature and numbers are linked by geometry: a snowflake, a spider web, the cells of a beehive. Behind the incredible amount of numbers revolving around your business there may be a fascinating shape. Data visualization is the graphic representation of data. It involves producing images that communicate relationships among the represented data to viewers of the images. Check this out: We can help you with this. We can help you shape the numbers of your business, a beautiful shape, a beautiful color ... in such a way that you can visually, quickly perceive the result, the improvement or the risk to be averted.

    • Predictive Analysis and AI

      Moving beyond the study of historical data, the new fronts we are working on are Predictive Analysis and the application of Machine Learning and AI-based tools. Predictive analysis is a category of data analysis aimed at making predictions about future outcomes based on historical data and analysis techniques such as statistical modeling and machine learning. Predictive analytics can generate future insights with a very good degree of accuracy. On the machine learning front, the scope is already very broad:

      • Customer Segmentation, Personalization of Services, and Targeted Campaigning.
      • Optimized Pricing
      • Fraud Protection
      • Optimized Search Results
      • Product Recommendations
      • Customer Support (ex: chatbot)
      • Stock management
    • Data Sources and Tools

      It all starts with the numbers. Where do we get the raw data from? We acquire streams for even realtime analysis typically from systems such as: E-commerce platform:

      • Google Analytics
      • Google Ads
      • ERP
      • WMS
      • WHO

      And any other source of data useful for further processing. The data are aggregated and made available for final processing. Depending on the result you want to achieve we use different processing tools. In case you are looking for a strongly verticalized default dashboard in online retail Drop uses its proprietary Drop.BI platform. This contains almost everything a digital retail manager wants (click here for more info). For more specific insights we use MS Power BI which allows us to quickly build dashboards that also update realtime in a very sophisticated data visualization environment. In the case of web-based dashboard design we can develop from scratch high-level dashboards usable from any device based on frameworks such as Bootstrap or VUE.js (e.g.,

    ROI calculation on UX or Digital Marketing operations.

    Beyond everything, being e-commerce managers, we know very well that any operation and investment must be motivated by a positive return. We have done this so many times that we have become good at understanding and showing through numbers and graphs what brought good and what did not. The topic is obviously vast, but the numbers at the end of the day are always the same: 1,2,3,4, ... And it is to these numbers that it is necessary to return in order to simplify. In the end, the clear answer that must emerge is simple: yes, do it again. No, it did not bring good. Ni: try again, but differently.

    Design of decision-making dashboards

    We can help you design the perfect dashboard for your needs. We can work on our Drop.BI platform or develop from scratch a web-based interface that updates on data streams from your infrastructure. The dashboards will be technologically state-of-the-art and show data with state-of-the-art data visualization.

    Analysis on product assortment to serve the buyer

    The buyer is often faced with deciding what to buy in a short time and under pressure. A retrospective analysis of how products, sizes, colors, markets have performed in the past is necessary at this stage. With an awareness of what has happened in the past, correct choices can be made for the future.

    Implementation of AI-based tools to improve UX

    Today it is no longer effective to handle cross selling or up selling by hand, it is simply anachronistic and ineffective. There are many Machine Learning and AI-based components on the market that we can integrate into your website in such a way that the content on the website adapts to the visitor automatically. AI logic analyzes everything that happens on the site and the visitor's own behavior. Based on all this, the right content (product) is proposed, at the right time, to the right person. All this will optimize the CR of your ecommerce.

    Product insights

    The retail manager needs to dynamically monitor sales performance by category, color, market. It is necessary to predict and highlight possible stock breaks, especially on bestseller products. Drop can provide a clear view of this.

    Extracting financial reports

    We can go dig up data in every nook and cranny of your digital ecosystem to re-aggregate it and provide it in "beautiful copy" to your CFO.

    Customer clustering

    Obviously, we are moving toward such granular clustering that we will no longer speak to groups of people, but to the individual person. This is the new front. However, today it still remains important to simplify and imagine models for breaking down your customer base into groups. We can help you with this by starting with specific drivers that we are able to identify through a good reading of the numbers.

    Software selection

    There is an infinite number of tools for any need in the "data science" sphere. We can help you choose your BI platform from MS Power BI, Christal Report, Qlik, Tableau and others. We can help you identify the right AI based widget to integrate into your e-commerce front-end. The operation for us curious, passionate people is almost fun, which is why we do it well.