Obtaining customer data for analysis: in a hassle-free manner

While Black Friday is known as retail's craziest day of the year, deeper analysis of customer data can make life easier. However, the data needs to be obtained without negatively impacting customers.

For by
Johannesburg, 24 Nov 2017
Partner content
Visit our company zone
David Cosgrave, Customer Intelligence Lead, SAS South Africa.
Read time 4min 20sec
David Cosgrave, Customer Intelligence Lead, SAS South Africa.

Black Friday serves as a wonderful springboard for online retailers to improve themselves and their services rapidly, thereby offering better, more targeted deals to customers in the future. After all, by their very nature, online retailers gain access to a wide variety of electronic data about customers, each time these clients visit their Web site.

In order to best utilise this information and improve their services, says David Cosgrave, Customer Intelligence Practice Lead: EMEA South at SAS, retailers need to understand that there are effectively two types of customers - those that are known quantities, and those that are unknown.

"Obviously, e-tailers want to collect online data in order to personalise the experience and give the customer what they want to see, in as few clicks as possible. While it might seem like they won't know anything about anonymous visitors, such as those customers who don't log in, this is not necessarily the case," he explains.

"Even anonymous customers leave behind interesting breadcrumbs of behavioural data from the moment they visit the site, such as the products they viewed, whether they added something to a cart, if they browsed reviews and even how they reached the site."

Cosgrave points out that this data can still be used in real-time to fine tune the experience. In addition, Data Management Platforms (DMPs) can provide online retailers with third party data about anonymous customers from their 'cookies', based on their browsing behaviour on many other sites. This can be used to instantly identify that a particular 'cookie' or device is most likely affiliated to - for example - a male between 30 and 40, in the high income segment, and with an interest in technology and gadgets. This is all very useful information to help personalise a site.

"At the same time, with the 'known customers' - namely those that have logged in previously, or have made purchases before - online retailers will generally have demographics such as age, gender, location and income level. They will also have access to previous purchase behaviour, recent patterns of behaviour that may indicate life stage events (such as a new job or child), as well as their browsing behaviour up to this point."

"Furthermore, social media data can be extracted if the retailer uses a 'log in with Facebook' option. Finally, retailers also have access to surveys and reviews that the customer has done, which can indicate additional insights such as their interests and their feelings about particular brands or about the Web site in general," he says.

This vast amount of data can thus be used by the predictive analytics and propensity models to inform personalised offers, as well as the types of adverts that are displayed on ad networks or even on the site itself.

Obviously, he continues, retailers should always be striving to extract as much data as possible, but it must always be done in a way that is effortless for the customer and does not excessively impinge on their privacy. "This is why the 'log in with Facebook or Twitter' and 'Like us on Facebook' options are so popular - they are a good way to obtain additional customer data in a hassle-free manner, which will ultimately ensure the retailer can provide a better customer experience," states Cosgrave.

"In addition, retailers can offer incentives, such as discounts or entry into a prize draw, to customers to either link their social media accounts, or to complete their profiles. Some also offer discounts if the customer gives their consent to be marketed to. Even non-digital retailers can get in on the act, thanks to loyalty programmes, which are extremely useful ways to track customer purchase behaviour and to drive repeat business."

Of course, companies need to ensure that customers are incentivised to use their loyalty cards for every purchase, and that the loyalty programme is sufficiently transparent and rewarding enough to make it worthwhile.

"Electronic membership cards are a great way to avoid the traditional problem of customers' losing their card or forgetting to bring it along. Furthermore, if a retailer owns several brands, the loyalty cards should be shared across brands. This will afford them the opportunity to mine even richer customer data and allow them to make more accurate predictions about customer behaviour and also identify customers may be getting ready to churn."

"The key to accessing this vast array of customer information, in order to analyse and then deliver on it, is to ensure customer privacy. Clients who feel their personal information is secure are always more willing to share additional data, particularly when they know they will get something - such as a more personalised experience, a discount or something similar - in return," he concludes.

Rodney Weidemann
ITWeb Contributor

Rodney Weidemann is ITWeb's contributor.

Have your say
a few seconds ago
Be the first to comment