While the rise of technology, the internet and e-commerce has generally been a good thing, within the fashion industry it has left a number of major casualties in its wake. Jane Norman, BHS, Austin Reed, Tie Rack, Banana Republic – these are just a select handful who have been lost from the UK high-street in recent years. So, where did they go wrong?
The problem is this – retail is notorious for being one of the slowest sectors to adapt and utilise new technological advances. Traditionally, many major brands used to play their cards close to their chest, working in secrecy and keeping vital information to themselves. They would base decisions related to the colours, pricing, style and fit of their garments off unstructured data, and wouldn’t think to watch competitors in much detail or monitor what their customers actually wanted to buy.
That all changed when Amazon came along. Using an abundance of new technology, such as machine learning and artificial intelligence, the retail behemoth began highlighting the importance of data science. By analysing big data, Amazon was able to stay ahead of the game, determining what was in trend and when. As a result, fashion brands had to sit up, pay attention and change their way of thinking.
Modelling a reliance on new data
The traditional closed-book method of analysing retail data meant that a number of fashion brands missed out on a lot of crucial information, such as data related to pricing, trends, insights and other must-have details. This may seem baffling to us now, given the competitive nature of the fashion industry and the importance of staying relevant, but it took a long time for brands to start using technology to their advantage.
In today’s market, that has all changed, and the fashion industry is now more reliant on data science than ever before. For example, specially trained data scientists can now predict whether a new collection is likely to be a success or not, simply by assessing previous sales data. This, in turn, helps companies ensure their money is being spent wisely.
Likewise, using concepts from predictive algorithms, visual search, natural language processes, and structured photographic data, brands can now identify trends before they’re in fashion. As a result, they can then develop bespoke designs they know will resonate with a target audience. In the past, companies would have relied on focus groups for this purpose – to predict whether a collection would be a hit or not. Nowadays, well-known brands like Ralph Lauren and True Religion use what’s known as ‘actionable product intelligence’ to determine how changes in product fabric, design details, colour and price affect a customer’s response. This all takes place before the item is even created, minimising the costs and the likelihood of it being a flop.
Insightful internet info
One of the biggest changes in recent years has come about through the rise of the internet, and the mass of data it now provides. Thanks to the advent of social media, there is now a wealth of information available for fashion data scientists to analyse and take advantage of. Whether through monitoring post engagement, watching out for Instagram trends, keeping an eye on Twitter hashtags, analysing the clothing styles of popular vloggers, or looking at the ‘likes’ and ‘reactions’ of popular celebrities, insights such as these are absolutely invaluable to clothes designers and fashion campaign managers.
For example, when releasing a new collection, many brands will post photos on social media to gather feedback and monitor the public consensus. They will then use these comments to make any changes required before its proper launch. This whole process is known as sentiment analysis, where publicly available information can be converted into structured, usable data for companies to utilise.
Using wi-fi to shape shops
Another area where the internet has significantly helped fashion retailers is in the layout of its high-street shops. While it may sound a little strange, through the store’s free wi-fi service, data scientists can track and use each customer’s connection to determine a number of things. They can tell, for example, how long customers spend in the store, how often they come back, and which sections they spend most of their time in. Using this information, they can then inform shop managers on how to shape the store layout, so that frequently purchased items can be placed in optimum positions.
Tracking big data also enables companies to determine the types of products to make, and provides information on the demographic of people who buy their clothes. If, for example, a store has a lot of male shoppers looking for new trainers, that brand will then know who to target and be able to create relevant designs.
In essence, the relationship between data science and fashion all comes down to keeping on top of the customer, using data to continuously track the who, what, when, where, how and why of purchasing decisions. Data scientists seek answers to questions like: what our customers are interested in? How do they make their choices? When do they decide to buy? Where do they shop? Which competitors do they buy from? Once they know the answers to these questions, the design and development of new collections becomes a breeze.
Looking to the future
Now that the fashion industry has started to take notice of the opportunities data science can provide, the future is looking bright. Further advances in machine learning, artificial intelligence and other sectors will only add to this, and shape the fashion of the world yet to come.
While the high-street may continue to struggle, brands who stay ahead of the game and prioritise data science will thrive. After all, the fashion industry is an incredibly Darwinian sector – only those willing to adapt will survive.