In this guest post, Bonnie Roche (pictured below), senior manager at Metrix Consulting, argues that when it comes to loyalty programs, we need to stop talking about love and start talking about logic – particularly in the age of data.
Unsurprisingly, loyalty programs claim to be about increasing loyalty or love for brands and retailers, but with consumers typically having 3.8 loyalty cards in their wallet, can we expect these cards to be an indication of, or vessel for fostering emotional connections with a brand? We say no!
Now, before you assemble the troops, we are not saying that loyalty programs are void of emotional attachment… simply that this emotion is only a small component of what a successful loyalty program is and can be, so hear us out.
Let’s stop talking about loyalty programs as a one-dimensional measure and indicator for consumers’ emotional connections with brands and start thinking about them as a rational tool. One that consumers can use to get a better experience and one that businesses can use to better understand, anticipate the needs of, and communicate in a personalised way with their customers – all of which can be achieved through the understanding and strategic use of data.
Why data? Because data is the tool that helps us improve the effectiveness of our marketing spend (ROI). By leveraging this data to communicate with our customers in a more relevant way, we improve our chances of them visiting us or buying us more often and spending more with us over time. Whilst reports indicate 57 per cent of Australians are more likely to return to a retailer if they have a loyalty program, in reality, consumers are going back to the retailers who are able to target them effectively, about things they’re interested in or see value in, and in a way that they find engaging.
Businesses today have access to more data than ever before, but without a clear strategy regarding its collection and use, it has little value. You need foresight to identify what needs to be captured, how you will capture it, and finally how you will transform it to create real value.
Let’s focus though on the quest for better customer data. From my perspective, there are two key questions you need to ask yourself:
- What can I get?
- What can I do with it?
What can I get?
The data that you get from your customers can be classified in two ways:
- Inherent data: this is the information that is collected upon sign-up (sign-up data) or from the customers’ day to day interaction with the company (passive data), such as number of visits, time and day of visit.
- Volunteered data: this is the type of data that can be asked of your customers without the need to pay them for the information. This information is beyond the day-to-day interaction (things like feedback and preferences).
The amount and type of data you can get within each of these categories is driven by the quality of your relationship with your customers. Specifically, how ‘committed’ consumers are and your share of their wallet. Whilst share of wallet is an easy enough metric to calculate, ‘commitment’ is a multi-faceted measure which depending on the category, can be comprised of:
- Category involvement (personal interest in the category).
- Existing brand relationship (emotional attachment or loyalty to the brand).
- Frequency of purchase in the category.
- Perceived longevity of relationship (which encompasses time to reward or benefit).
- Perceived loss of non-commitment (loss of rewards or benefits).
This commitment element is where we start to see emotional attachment play a role, but as can be noted from the list above, emotion is only one element of a multifaceted measure.
Why do we need this definition of commitment? Because it not only shapes the types of rewards that your customers are going to expect from you, but also the data that customers are willing to give you.
For us, the easiest way to conceptualise the ‘quality’ of a relationship you might have with your customers, is to think about it in human terms:
To bring this to life, let’s apply the quadrant to two different categories.
Imagine an ‘everyday’ worker with no strong preference regarding where they buy their coffee. Let’s say they visit a coffee shop that has a 10-for-one coffee loyalty program and they ‘sign up’. This customer has low category involvement, no relationship with the coffee shop and risks nothing if they don’t go back frequently to the coffee shop. This is a low commitment, small share relationship – a one night stand. In this relationship, there is no data collected on the customer and no communications, but given the nature of their relationship, this isn’t required or expected by the customer.
Compare this to someone who is a member of the Qantas Frequent Flyer program. Customer commitment to the program is stronger given the longer-term reward structure and the risk of ‘loss’ if they exit the program. In this relationship, there can be a lot of data collected (particularly if the customer has Qantas credit cards or travel cards) and regular communication between the airline and the customer. Given the nature of this relationship, this expected and permissible.
Once you have determined the type of customer relationship that you have, the other lever that impacts the amount and type of data your customers will give you is the incentive you provide them for doing so.
Incentives and quantity of data can be thought somewhat like a linear relationship (of course, there will be exceptions) but simply, the more incentive you provide, the more data you can get.
Continuing with the coffee example above, you’re unlikely to give up your name, email and address for a low value, delayed reward such as a free coffee after you purchase 10.
But if a coffee shop (like one we know in Byron Bay) was offering a free coffee right off the bat, if you sign up to an app that provides them with your name, age, address, payment details and access to your location and order history, you might be willing to give up your details. This immediate and tangible incentive gives customers enough of a reason to provide information that would otherwise not be captured.
But let’s be honest, there is absolutely no point in even collecting data if you don’t know what to do with it. Which brings us to question two…
What can I do with it?
So, now that we understand the type of data that customers might be able and willing to give us, it’s time to think about what can be done with it from a reward, targeting and communication perspective.
Rewards is the simplest of the three to explain. What you can do with your data to reward and incentivise your consumers must abide by the boundaries of the relationship you have with your customer. Is it appropriate for a one night stand (coffee shop) to wish you happy birthday? Never, creepy!
Do you expect Qantas to change the rewarded points based on the value of your flight? Of course.
When we consider the rewards that are expected and deserved by customers, the status of the current relationship and the projection of what it will look like in the future needs to be considered (moving from one box to another).
Perhaps one of the best examples of a business correctly identifying the boundaries and status of a relationship, and using their data to create real value for the business and their customers (targeting and communicating), is Woolworths via their Everyday Rewards (EDR) Program.
Each time I shop with Woolworths I swipe my EDR card and in turn, receive a point value equal to my commitment (i.e. per dollar spent). This incentive is adequate for me to swipe my card every time I go into the store… a committed relationship.
The result is that Woolworths knows what my ‘typical’ basket looks like. For example, that every Sunday I buy a two-litre Great Ocean Road Milk, fresh bread and nut bars. They also know that given my typical inter-purchase intervals, I am due to buy toothpaste this week (a purchase I make every three weeks) and that I usually buy Sensodyne regardless of what is on sale (inherent data).
But imagine this Sunday – no transaction registers against my card. Monday rolls around and this is picked up by the Woolworths data system and so on Monday, I receive a tailored (one-to-one) email with a unique set of promotions to match my needs:
- Great Ocean Road milk is on special
- Edwards Sourdough bread is on special
In addition to these known behaviours, the algorithm also gives me specials based on what other people with similar behaviour have bought (think about the famous case of Target in the USA predicting a girls’ pregnancy:
- Butter and dental floss earn you double points this week (data analysis).
But it doesn’t stop at me receiving a tailored email, because a good loyalty program should continually learn from the data it collects.
The Woolworths data system runs 70 million queries each week to develop one-to-one communications, with all 50,000 products ranked daily to determine which offers would be most relevant for the customer.
You’ll notice that a Sensodyne offer didn’t make it into my promotional email. This is because the system has learnt that I buy this same brand no matter what the price.
On Tuesday when I finally make it to Woolworths, I buy the Sensodyne toothpaste, the Great Ocean Road Milk, the dental floss and the butter, but don’t purchase Edwards Sourdough bread – I opt for my regular fresh bread instead. It’s unlikely I’ll see another offer for packaged bread until my behaviour indicates I’m interested in it… a continual test and learn process.
And what about volunteered data? Well, whilst Woolworths doesn’t currently incentivise customers to gain additional data from them, in the relationship we have detailed, if I was offered dollar rewards for my feedback on a store layout or customer experience, it would be within the bounds of our relationship.
The Woolworths data processing and one-to-one communication approach is five times more effective than sending the standard ‘weekly special’ email and has improved email open rates by 16 per cent. Talk about impact! Woolworths is a great example of a company knowing the limits of the data they can ask for, providing the right incentives for customers to share their behavioural data and continuously learning from the data they collect.
Summed up as a diagram, the whole process looks like this:
So let’s reframe the way that we think about loyalty programs – let’s make it less about love and more about logic.
Rather than just asking how we can use these programs to encourage customer loyalty, let’s start asking ‘how can I obtain and leverage the right data from my customers, to better target my customers and achieve the desired commercial outcomes for my business?’