Retailers using data to target customers often come up against some common hurdles. Retail Week looks at how to overcome them
Finding the time, resources and money to invest in data and personalisation can be hugely difficult for retailers. While solving the practical problems is a challenge it’s not impossible, and growing numbers of retailers are developing their use of data. Not just on their websites, but in their stores as well.
Here are answers to some of the questions asked by retailers about how to use data to personalise their offer.
1) How can I convince the board that improving our use of data is worth it?
Convincing other parts of the business that data projects are worth prioritising is not easy. However, it is possible to sell the idea of investing in data to the leadership by presenting the big picture of what things might look like in the future. In addition, it is wise to speak to each department about how the work will affect them, rather than saying what it will deliver because then you will be tied to delivering something specific. That way, it is possible to get buy-in from each team individually - even if it takes a couple of months.
Data analytics can also nearly always produce some quick wins fairly cheaply, so retailers can prove to senior management bit by bit that the return on investment is worth it.
2) What are the best ways of listening to customers and getting feedback?
Jenni Cumming, head of CRM at department store Liberty, says: “We’re good at structured data, but we want to differentiate ourselves in terms of service. But we don’t measure satisfaction – we should be listening to our customers and capturing that.”
This is often the second step retailers take on their data journey. Customer surveys can provide this sort of feedback, and a growing quality of data is coming from social networks. Some retailers are also using targeted customer review software, such as Feefo’s service, to gather feedback.
3) What is the best way of capturing data in-store?
This can be a challenge, both in terms of how much can be captured, and in terms of the quality of what is collected.
Store data is enormously valuable, and the process can be as simple as asking customers for their postcodes and linking them to the transaction. While the speed of this process depends on the point of sale system, at beauty retailer Space NK it adds only 45 seconds to the transaction, says Joe Pack, head of CRM and customer acquisition at the business.
E-receipts are another way of capturing store data – shoppers need to provide their email addresses for the receipt to be sent to, and other details are tied to each transaction as well.
4) How can we engage store staff and get them to help us?
Store staff can be resistant to the idea of data capture and ecommerce – even now, many are still suspicious of the website stealing store sales.
Pack says: “There’s still resistance to online among some store staff because some think it’s going to take all their sales.” He adds: “Store staff are all educated in terms of capturing email, but they don’t yet tie it to the transaction.” This is important because tying a
store transaction to a person’s email or postcode means their activity online and offline can be joined up and viewed as one.
The answer to getting store staff on board is to make data useful for them, by providing store managers with a report on why certain products or lines are performing well – or not – and on how their most loyal customers shop across channels.
Beril Becker, director of customer success at data firm AgilOne, says it’s best to make it useful for store managers. “Once data is captured you can tell more about who is buying a range. Managers will know if a range isn’t selling well, but they wont know that the range is also bringing in a lot of their high value customers, for instance.”
5) We have a huge volume of data – how do you know the right questions to ask of it, and what is the best way to manage the process of gaining insight from our data?
Once data has been collected into one place, or disparate systems joined up, algorithms can be run on the data and patterns will start to emerge. They can provide a good starting point for the type of work that should be done.
Alternatively, retailers should sit down with their analytics partner and work out exactly what questions they want answered. Paul Gibson, director of sales at AgilOne says: “There might be a group of people who buy once a year from you, and you want them to buy three times a year.”
Retailers need to identify their key goals for the coming 12 months and then find the metric that will help achieve them. Gibson says: “Marketers have very creative ideas but they don’t always have data that helps them understand what they should be working on now.”
The patterns coming out of the data – and a retailer’s priorities – may change over time, so retailers need to monitor and when necessary update what they’re asking of their databases.
6) How should we approach developing our use of data to include predictive analytics?
Predictive analytics don’t need anything planned or prepared other than at least six months to a year’s worth of data, which nearly every retailer has.
Gibson says: “People look at predictive analytics as something only
the Amazons and Tescos do, but it can be simple.”
At a basic level, predictive analytics can involve assessing how ‘likely to buy’ each online shopper is, based on his or her transaction and browsing history plus any other data that is held on them. The promotions shown to them online can be tweaked according to how likely to buy they are, or because they always buys a particular type of product at the same time each week, for instance.
7) Our data doesn’t tell us enough. How easy is it to pull in external information such as social or weather data?
It’s not difficult to add external data sets to your stable, but it’s wise to proceed with caution until your analytics is in a fairly robust state.
Retailers starting out on their data journey risk overwhelming themselves if using other data sets, so it’s important to get used to using transaction data, and to tying in stores where appropriate, before adding to it.
But once you are in a position to be able to experiment with third party data, the options are endless.
Weather is a popular one, and so is Facebook’s Custom Audiences data. Once a retailer has a picture of what their most loyal or highest value customer looks like, it can find other people on Facebook whose profile
is similar. This means your ads and sponsored news feed posts are likely to be seen by people who will
be useful to you – much better than a scattergun approach.
Data analytics is one of the areas with the highest level of innovation and entrepreneurial activity amongst start-ups and larger technology firms. Which means now has never been a better time for retailers to try it.