Case Studies



Task
  • Assess the profitability of books range and promotional spaces
  • Across all store groups, book categories & promo space types
  • Hypothesis: large no. of unprofitable SKUs and promo spaces across particular store groups
Challenge
  • Task was not well defined at the start
    of engagement

  • No prior profitability analysis performed
  • Very large volume of data across disparate sources: 50K SKUs, 600 stores, 32 weeks, ~ 30 GB
  • Complex calculation required – unmanageable in traditional tools, i.e. Excel and Access
Approach
  • Identified key elements of value chain, in collaboration with client teams
  • Agreed assumptions, iterated & regularly tested analysis with client on small-scale data
  • Developed approach to promotional
    space-level modelling

  • Performed analysis on full data set, distilled as simple outputs

Solution
  • Disparate & messy data collated into single

    cloud database

  • Defined calculations and outputs on small scale in Excel, testing continually with client
  • Profitability calculated by combining cost-centre PPUs with sales, stock & throughput data
  • Analysis on full dataset executed in Java
    and Python languages

Outcome
  • £2m potential annual saving from removing unprofitable range SKUs
  • Visibility of profit associated with any book or promo space in any store
  • Identified promotional activity as overall profitable
  • Highlighted potential for significant improvement in promo space allocation
Timeline