The Smart Cube’s advanced data analytics deliver insights to support growth for US e‑commerce business, UrbanStems

Key highlights
  • UrbanStems wanted to improve its demand forecasting, through a greater understanding of volatile purchase patterns
  • The Smart Cube’s analytics specialists leveraged business data to undertake Purchase Pattern Analysis, and Demand Sensing and Management
  • The client gained new contextual intelligence and actionable insights to manage volatility and improve sales
Download a copy

Business challenge

UrbanStems is a nationwide modern floral and gifting business. Established in 2014, the e-commerce outfit has expanded fast: its three-year growth rate of 1,192% saw it positioned at #385 on the 2019 Inc. 5000 (a ranking of fast-growing private US companies).

Alongside strong growth, UrbanStems was experiencing very high demand peaks. To develop a sustainable business strategy and maintain customer service and satisfaction levels, better demand forecasting was required, underpinned by an understanding of volatile purchase patterns.

UrbanStems engaged The Smart Cube as an external partner, to leverage the company’s specialist data analytics skills and expertise, and retail knowledge.

The Smart Cube solution

Following consultation and project scoping with the client, The Smart Cube recommended an analytics proof of concept, based on its Revenue Growth Management solution. The solution was customised to address UrbanStems’ specific challenges, focusing on Purchase Pattern Analysis (to understand customer profiles and buying trends) and Demand Sensing and Management (to drive evidence-based planning).

The Smart Cube’s specialists gathered internal information from UrbanStems – two and a half years of transaction data, including customer demography, products and distribution points – which first needed collating and cleansing, to create a data dictionary for trend analysis.

Using custom-built algorithms, The Smart Cube leveraged the combined dataset to identify time-based purchase patterns and forecast overall demand, through three stages:

  • Analysis of purchase patterns, for demand over different times of the day, days of the week, seasons, etc.
  • Filters created to view purchase patterns by specific factors including gender, region and categories
  • Forecasting model developed for time-based purchases using time series models


The Smart Cube provided UrbanStems with intelligence on purchase patterns across times, seasons and buyer groups, for SKUs and product categories. Insights derived include:

  • Overall order volumes (identifying highest and lowest volumes across days and times of the week)
  • Event-based orders (the order trends follow similar patterns across the years – including Valentine’s Day and Mother’s Day seeing peak orders)
  • Delivery mechanism (including the preferred delivery providers and speed of various services)
  • State level (six states showed high level of localisation, with around 50% of orders being sent and received within state)
  • Customer level (including repeat buyers – the majority of orders are now placed by repeat customers despite fewer promotions or discounts, indicating an overall increase in customer loyalty – and time to repurchase)
  • Category level (identifying trends related to purchase of plants compared to flowers, and add-on products)

Using the data for demand sensing and management, analysts were able to forecast numbers using time series models for the year ahead (with highest accuracy for the next three months). UrbanStems has also been able to leverage The Smart Cube’s intelligence to enhance in-house forecasting models, leading to greater levels of accuracy and improved forward planning.

Value delivered

The partnership with The Smart Cube has provided UrbanStems with:

  • Specialist data analytics partnership: As a relative start-up, UrbanStems recognised that partnering would provide access to specialist skills and capabilities, with The Smart Cube’s analysts working as a seamless extension to its team
  • Actionable insights through timebound methodology: The analysis was conducted over a period of four weeks, leading to immediate and relevant insights which could be used to manage volatility, improve forward planning and drive sales performance
  • Strategic opportunities to leverage indepth insights: Working in collaboration, the partners have identified areas where The Smart Cube could provide insights at greater depth, adding more value to  UrbanStems’ business planning and growth strategy


Vivek Kumar, Chief Revenue Officer, UrbanStems said: “I particularly found the purchase pattern analysis very insightful and interesting. It shows that The Smart Cube team was able to quickly pick up the nuances of our operations, and deliver insights based on the data and how we think about the business. Very impressive!