Member Article

Fashioning better forecasts

When it comes to reporting on trading results, often the (not so) Great British weather remains the scapegoat of choice for retailers announcing poorer than expected performance.
Topps Tiles recently announced a 2.2% decline in like-for-like sales for its second quarter, with the blame pointed squarely at bad weather, while Kantar also suggested that the ‘Beast from the East’ – the unseasonably cold weather in the UK March 2018 – cost supermarkets £22 million that month.

In fashion, the likes of Debenhams and Primark are two examples of retailers that attributed lower than expected sales of their products to unexpected weather conditions this year. With a raft of new technology available to retailers, and with better data analysis, it doesn’t need to be this way.

In what is becoming an increasingly non-seasonal sector, fashion retailers should be looking to develop more accurate forecasting to drive greater full-price sales, and a premier service, while also maintaining their margins. More sophisticated use of data can help here – and it can also make retailers more adaptable when the unexpected happens, as it inevitably will.

Retailers can leverage the benefits of data science to optimise in-season performance in three key ways, as Robin Coles, Director of Inovretail, explains:

Flexible forecasting Online fashion house, Asos, is in the process of establishing a whole new merchandising system which will inform its decision making around markdowns, so that it sells its ranges at the optimum price throughout the course of the year. Data science and a burgeoning team of technologists, engineers and analysts are driving the change at the celebrated UK e-tailer, which will see the company make pricing decisions throughout the year based on various data points and real-time number crunching.

It’s just one example of the increasing reliance on data by some of the leading retail organisations. This scientific approach to retailing, which can be enhanced by working with an organisation like Inovretail, reduces the need to consult out-of-date spreadsheets, and allows retailers to take action based on real-time statistics and relevant business intelligence (BI).

Reacting to demand Aggregated data gathered from Inovretail’s clients using our data science solutions shows that, in most Mediterranean climates, it can be expected there will be up to a 30% increase in store visitors on a rainy day.

This type of intelligence is vital in fine-tuning store staff sizing and thus meeting real-time demand from customers.

But reacting to demand is not all weather-related. It’s about the general ability an organisation has to reforecast, down to a more granular level, during trading. It’s about having the ability to recognise, due to a local event, that different things are happening and a quick response is required.

And, crucially, it’s about making sure staff on the frontline are connected to the information held at head office about demand.

Empowered store staff Data science can be used to empower store staff in various ways. In fact, real-time data optimisation is only really achieved when shared with the wider retail team – there’s no use in BI being kept as the preserve of central office.

Imagine the power of filtering stock information or sales performance data straight to store staff via smart watches and other wearable devices, so they can use this intelligence ‘in the moment’ when serving customers or conducting store- or warehouse-based tasks?

This capability exists, and could have a transformational impact on fashion retail customer service, whether in detailing what stock is required or helping shoppers in the fitting room.

The truth is that more often than not in retail, despite best laid plans, the plan isn’t followed through. In the past, veering off course might not have been such a problem as it is today but businesses are now operating against a very different backdrop to the pre-digital age. Mistakes or unexpected outcomes in retail today tend to be punished more severely.

There are plenty of alternative, agile, and digitally-led businesses, such as the aforementioned Asos and its digital-led peers Boohoo and Missguided, which are available for customers if they are turned off by high street failings.

Dabbling in data science and putting trust in new forecasting systems that run on real-time information rather than historical sales data – and which can be used to empower staff at the coalface – should give retailers added flexibility, a better chance of adapting to unexpected events, and a greater opportunity to compete.

This was posted in Bdaily's Members' News section by Chelsea Reay .