The three words no customer wants to hear

If you were one of the many robbed of a bargain during this year’s Click Frenzy online sale, which saw inventory and technical issues leading thousands of would-be customers into a social media outrage, you’d be more inclined to think the latter.

In today’s digital age of instant gratification and the plethora of shopping channels available, consumers are more likely to act with bemusement bordering on horror when an online retailer cannot immediately supply the items they desire.

Consumers don’t see it as their fault for being beaten by the rush – but rather the retailer’s for failing to predict and plan for the unexpected demand.

Cut-throat? Perhaps, but that’s exactly as it should be. Every retailer should aim to have the right products available at all times.

With so much competition in the online retail market, businesses that fail to give customers what they want will quickly lose both reputation and market share.

This means retailers should have a clear grip and visibility on inventory, globally. Operational efficiencies must be enhanced in order to run an effective supply chain.

This theory will be put to the test again this holiday shopping season. Aussie consumers are expected to spend over $50 billion during the upcoming Christmas trading period (2.8 per cent more than last year), according to the annual pre-Christmas sales predictions from The Australian Retailers Association (ARA) and Roy Morgan Research.

It is imperative during this time that online retailers run a tight ship and avoid having excess or insufficient stock.

A complex machine

With the increasing complexity of retail operations, every cog in the machine, from manufacturing and supply chain to stock management and dynamic pricing, depends on the ability to forecast demand accurately.

And while it’s relatively easy to pore over historic data and identify the reasons why certain products were so popular at various times (for example, because of a heat wave or a new sales promotion), this insight is not enough on its own.

Retailers must move incredibly quickly to implement these findings across every area of their operations, from the warehouse to the point of sale, if they are to react effectively to changes in demand.

This requires retail businesses to crunch enormous volumes of historic and up-to-the-minute data, while also automating processes throughout its operations to ensure that they can quickly implement the right changes to fulfil orders and keep the shelves stocked.

Software and silos

The biggest barrier to implementing an effective demand forecasting system is the traditional silo-based approach to data, which leaves this valuable information trapped in a multitude of separate applications.

The goal for retailers must be to break down the walls of these silos so information is shared quickly and efficiently across the business.

This depends on effectively managing the many different sources from which data is gathered, transforming data processing and storage so that it is available to every relevant line-of-business application, and introducing agile, need-based processing that requires no intervention from IT.

In simple terms, an effective data governance strategy coupled with the technology platform is critical for effective operations.

Modelling the future

Once the retailer has a handle on their data, they can then begin analysing the many variables that affect demand.

This requires mapping the relationship between sales and factors including (but by no means limited to) seasons, holidays, promotions, times of day, and previous customer behaviour.

Using exploratory data analysis, the business can break down massive data sets, construct and test a number of predictive models.

Through regular analysis, retailers will soon establish patterns and relationships that will enable them to identify and forecast trends that have an impact on stock-keeping unit availability, and then build algorithms that automatically make the necessary changes within the relevant areas of operations, from supply chain to pricing.

In the words of Arther C. Clarke, any sufficiently advanced technology is indistinguishable from magic, and there is something magical about predicting the future of retail demand.

Harsh Gulati is the senior director and regional head of Infosys.