Have you been a false-hope victim of e-commerce, buying something online and expectantly waiting to have it pronto - only to realise that receiving it soon isn’t soon enough?
The reasons for this are plenty, but two prominent issues immediately come to mind: e-com platforms not disclosing that what is advertised often still has to be sourced, and an age-old logistical problem that is part (and parcel) of home deliveries – the last mile.
The first is a betrayal of the what-you-see-is-what-you-get or Wysiwyg principle, and online platforms are increasingly urged to be forthright when it comes to inventory disclosure.
One such platform in South Africa, Takealot, last week professed to always indicate when something is in stock or not, a webinar declaration that has since been disputed (but more about that later this week).
As for the last-mile problem – market leader Amazon thinks it has a solution.
Part of the waiting problem, it explains, is that the current model works as follows: orders are honoured by “fulfilment centres”, after which they’re sent to “sortation centres”, followed by dispatch to “delivery locations”.
But what if “fulfilment” has an inventory problem and a new item facility has to be sourced?
In short, it turns out that the last-mile problem is often compounded by the Wysiwyg principle.
In other words, not only is the item you’re seeing on screen not in stock, but when it’s finally made available for delivery, the pressure on couriers is such that it involves rerouting, resulting in a ripple effect across the last-mile spectrum.
So what is Amazon’s solution?
Machine learning.
Through an intricate, artificially intelligent system of product coding and geolocation, it believes that all the little snags in the chain of supply can be eradicated by instantly mapping the shortest way between source and recipient.
Much still has to be figured out about this proposed solution.
But for the most part, it appears that sortation is going to be transformed in such a way that delays in the last mile are minimised.