As is often said, knowledge is power. For retailers, knowledge is money. The more you know about your customers the better you can tailor your offering and the more sales you will make.

It’s a fairly simple formula that belies the complexity of understanding what people actually want.

After all, many of us, when confronted with questions about our preferences don’t immediately know the answer. This is shown time and again through customers’ “declared data”, which is so frequently flawed or downright false. So if we don’t know (or can’t express) what we want half the time, what chance do retailers have?

The answer is a much better chance now than even a few years ago. Thanks to the sheer volume of interactions the average person now has with organisations it is much easier to understand what people do and like. However, it is also a double-edged sword. With so much data flying around, it’s getting hard to detect the signal from the noise. Thankfully, new data science techniques are helping organisations to get value out of the data they collect.

In an ‘omni-channel’ retail world – an ugly buzz word, but a useful catch all for the different ways retailers interact with consumers – there’s plenty of information for retailers to collect on consumers. For those with concerns over privacy, I should point out at this juncture that for all intents and purposes consumers are essentially anonymous and the data collected is publicly available.

Customers happily offer up data to retailers all the time

Amongst the data collected in store, online, via mobile, telephone sales and market research, is what a consumer states they want – their declared preferences. For example, someone might state to a clothing retailer, via a survey, that they aren’t interested in celebrity fashion labels or prefer to buy items on the spur of the moment rather than well in advance.

However, by analysing social media preferences it might actually become apparent that the customer is very interested in a particular celebrity – especially the clothes they wear. These clothes may actually be from the celebrity’s own fashion label.

It may also be clear from how they browse online or use social media that the customer heavily researches clothes before they make a purchase.

Perhaps they usually ask friends on Facebook for advice, or spend a long time on a clothing website comparing different versions of an item. These are very simple examples, but what it shows is that behaviour can reveal the ‘implied preferences’ of consumers.

By using sophisticated data science techniques, every thread of information can be woven into a complete picture of a consumer. Marketing, pricing, stock, logistics and customer service can then be tailored accordingly.

This scenario is obviously very attractive to retailers. However, creating the conditions to make it possible requires a lot of work.

The first step is to put in place adequate data collection and storage procedures. ‘Adequate’ means a lot more than just having a means to gather data. It covers collecting the right information, in the right format and in such a way that it is easy to access and use.

Hand-in-hand with the storage of information is determining how it is accessed, who can access it and how insights are created and shared. It may sound obvious that if data on customers is available, those working within the company should know broadly what it says to provide the best service.

However, a surprising number of businesses continue to ‘silo’ data. Breaking down these barriers and viewing data as valuable to all functions within a retailer goes a long way to improving how a business functions.

Combining your data insights makes them more powerful

When you have a rich set of information on your target consumer base – including data on social media preferences, online and offline retail habits, demographics and declared preferences – the next step is to analyse it. The best people to do this analysis are data scientists.

Data scientists will marry computer science with statistics. They use techniques such as machine learning (creating an algorithm that adapts as it receives new information), clustering (creating groups based on seemingly disparate data sets), or hidden decision trees (a technique to quickly mine a huge amount of data – for example, transactional data to detect fraud).

Employing a data science team, or hiring a consultancy to do the work, is worth the time and money. It can reveal information that can give retailers a profound understanding of their customers, and crucially, an edge over competitors.

Gaining insights from a data science team is only half the battle. Business owners need to decide what to do with what they have learned.

It can be disconcerting when data science reveals that long held assumptions about a customer base were wrong and consequently, the company should head in an entirely new direction. Repeated testing of the results obtained from data science can go a long way to ensuring that the insights are correct and building confidence.

However, as with any new business technique, it requires bravery to actually act on the results. Retailers who use data science will find that it can inform every area of their business and tell them things that they never thought were possible about their customers. Those who don’t risk getting left behind by their data-savvy rivals.

Mike Weston is CEO of data science consultancy Profusion