A Slop University paper scores nine vending-machine contractors against a five-level restock-anticipation maturity model, and reports a landlord consortium's decision to tie premium site tenancy to the resulting score alone.
A five-level Restock Anticipation Maturity Model, built from route telemetry, has become the test a shopping-strip landlord consortium applies when renewing a vending contractor's access to its busiest sites.
A paper from the School of Emergent Priorities finds that how a vending-machine route driver times a restocking visit, not how often the round runs, predicts whether a machine stays stocked. Drawing on twelve months of route and sale telemetry across nine contractors’ fleets, Doctoral Candidate Tove Ilesanmi, Dr Lindiwe Achterberg and Professor Verity Marris built the Restock Anticipation Maturity Model, sorting each fleet into one of five named levels from Reactive to Pre-Emptive restocking. A regional landlord consortium has since adopted the resulting score as the basis for renewing a contractor’s access to a shopping strip’s busiest sites.
A restock sheet was never meant to answer a question this large. It just turned out to have the answer sitting in it the whole time.
— Tove Ilesanmi, School of Emergent Priorities
The paper draws on restock and stockout telemetry from 214 machines across 41 shopping-strip and supermarket-foyer sites, sorting each contractor’s fleet into a named level and tracing what changed once one of the fleets’ own landlords made the ladder a condition of keeping its best pitches. Contractors reaching the ladder’s upper levels recovered lost selling time at under half the rate of contractors held to its lowest levels, a difference the School regards as a case study in what a fleet already knows about itself before anyone thinks to ask.
“The consortium didn’t ask us to build an eligibility test,” said Professor Verity Marris, Director of the Trajectory Analytics Group. “It asked for a maturity model, and found the eligibility test sitting inside it.”
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.2h2j54.