Imagine a world where price, promotions and stock can be optimised with minimum human intervention

Ali Athar | 04.10.2017

Can a software robot replace merchandisers?

Imagine a software robot, which could track and forecast sales for each product (or product type) at each location (online and in-store) every week, or even daily.

A robot which would know instantly whether you were likely to run out of stock, and would automatically raise a new allocation or replenishment order ensuring no lost sales.

Imagine that it could ‘forecast’ that your end of season sell-through was not going to be on target, and could recommend promotions that would bring it back on target.

What if it could predict promotional uplifts, and automatically change the replenishment parameters to ensure that the right amount of stock was available to support the promotion, and would know to allow stock to run down as the promotion came to an end?

Imagine it being able to achieve all of those things without any human intervention.

If when the ‘Sales’ were about to start, this robot could know which products to clear and by how much.

Imagine if all of that were true – what would we ask merchandisers to do?

Well, you no longer have to imagine. We have such a software trading robot.

We have not yet gone as far as turning off human merchandisers. With every alert or recommendation Profimetrics makes today, the decision is ultimately still left to humans.

Our goal is to dramatically improve the merchandiser’s productivity by doing the heavy lifting for them.

If you believe in the 80:20 principle, you will recognise that 80% of the work can be automated, leaving the merchandisers to concentrate on the tricky 20% where some judgement is required.

If you understand statistical and neural net techniques, you will recognise that, in most cases, systems can do statistical analysis and correlations faster and better than most humans. Our brains are not wired for intuitive statistical analysis.

In fashion or continuity retailing, the problems are different as we go from short season forecasting to life cycle modelling, but the statistical principles remain the same.

Stock optimisation and price/promotions optimisation are two sides of the same coin. If a product is selling too fast you need to balance stock distribution, if it’s selling too slowly you need to simulate sales.

The benefits are not only improvements in merchandiser productivity, but also improved margins (2-4%), increased sales (3%) and reduced stock.

>> Read more about Profimetrics