Scarpamondo
Turning a 47-store retailer into an organisation that decides on data, reading omnichannel behaviour as a single purchasing journey.
-
Industry
Retail / Fashion
-
Services
LIFT Conversion, ACTIVATE Data, EARN Retention
-
Year
2022-2026
-
View project
scarpamondo.it
A retailer’s customer does not live separately across online and offline — they move between the store, window displays, online searches, apps and e-commerce, leaving signals at every step, and few retailers manage to read those signals as parts of the same purchasing behaviour.
Scarpamondo is one of Italy’s leading footwear and leather goods chains, with 47 physical stores and a national e-commerce platform. When we started working together in 2022, the goal was to build a system that would allow the team to interrogate data at every level, from daily operations to strategic choices.
Three work streams that feed each other
Il progetto si è articolato in tre direttrici sviluppate in parallelo, ognuna con logiche proprie ma collegate dallo stesso impianto infrastrutturale.
The project was structured across three parallel workstreams, each with its own logic but connected through the same infrastructure framework.
Omnichannel analysis
We reconstructed user purchasing behaviour between physical stores and e-commerce, cross-referencing CRM, transactional and navigation data to deliver an integrated view of the customer.
Customer journey optimisation
Through qualitative and quantitative analysis, we identified and prioritised improvement points in the online purchasing flows. The UX research and design team produced wireframes and optimised UIs, always measuring impact not only on the conversion rate but also on lifetime value — so as not to gain one at the expense of the other.
Purchase driver analysis
Machine learning techniques applied to CRM, web analytics and transactional data made it possible to label each customer within a price/brand matrix, distinguishing four profiles with different purchasing logic.
Moving from a channel view to a behavioural view
The difficulty was not technical in nature — it was representational. Measuring store and e-commerce performance separately is a relatively simple exercise, whereas what was truly needed was a common grammar for reading users as people and not as online sessions or in-store receipts.
The data work went in this direction, starting from the reconstruction of persistent user identities across online and offline and arriving at a geographic logic capable of directing advertising investments based on the real potential of each territorial cluster.
A system that guides decisions, not one that describes the past
Geographic analyses made it possible to diversify advertising investments based on the stage of the funnel (awareness, consideration, conversion) and to calibrate communication and CRM strategy on territorial clusters with the greatest growth potential.
The analysis of the transition from in-store to online purchasing made it possible to rethink free delivery thresholds — an intervention that alone generated a 13% increase in the conversion rate.
The price/brand matrix returned an actionable segmentation across four clusters (Price seeker, Brand follower, Price functional, Brand seeker), used to differentiate communication, promotions and assortment.
Data that enters daily decisions
-
E-commerce conversion rate growing
The revision of free delivery thresholds, informed by the analysis of omnichannel behaviour, alone delivered a 13% increase in the conversion rate.
-
More efficient advertising investments
The geographic reading of data made it possible to calibrate budget and campaign objectives based on the real potential of each area, reducing waste and increasing return per euro spent.
-
Actionable segmentation of the customer base
The price/brand model transformed the customer base into an asset used daily by the marketing and merchandising team for assortment, promotional and communication decisions.
The customer is one person, and data should tell their story as such
If the data from your e-commerce and your physical stores live in separate systems, you are making decisions based on half the story — because the customer who buys online and the one who walks into the store are very often the same person, and their behaviour reflects it.