Can you achieve personalisation at scale? That was the question facing delegates at Marketing Week’s roundtable event in partnership with data management platform Relay42.
Personalisation is always a hot potato but never more so than now as marketers sit on the cusp of the EU’s General Data Protection Regulations coming into effect in May 2018.
While GDPR may well cut down the volumes of information marketers have access to, they are still faced with a huge ‘data lake’ from which to plot customer journeys and draw insights.
Steve Forde, director of online product and marketing at ITV asked: “How much data is enough? There is so much more now and it’s always been available. We have a pure vision of how much data we can see but just because you can use it, does it mean that you should.”
Knowing what data will be relevant is about understanding the relationship customers want to have with you. Ian Morgan, managing director, digital channels at Barclays UK explained that how the organisation manages customer relationships has a direct impact on the sort of data it uses.
“One of the biggest changes has been that banks were product-focused in how they used data. That ignores how relevant those products are to the customer. Now, how do we use data to add context and relevance? That is what makes us more customer-centric,” he said.
Can volume ever be personal?
Mattress manufacturer Tempur Sealy’s head of customer experience Barbara Callicoot and Gary Booker, CEO of consultancy Selinon and former Dixons Carphone CMO, both got to the heart of the personalisation challenge.
“We still want to be customer-centric but expectations change. There is never enough. The ideal is one-to-one but it’s hard at scale,” Callicoot warned.
“The challenges are cost and IT infrastructure among others. If you can, go for it. But how much should you spend and how much incremental gain will you get from it?” Booker asked.
We still want to be customer-centric but expectations change. There is never enough. The ideal is one-to-one but it’s hard at scale.
Barbara Callicoot, Tempur Sealy
Rupert Bedell, CMO at employment benefits company Unum explained how he put the brakes on spiralling personalisation costs and took another tack while in his previous role at RBS.
“We had too much data and went so far down the personalisation track it was getting expensive and complicated. We stopped and instead put it in the hands of relationship managers. Sometimes human relationships are better than building huge engines. We could see the net promoter score (NPS) going up as a result.”
What does intelligence mean?
When talking about doing personalisation at scale, it’s no longer about changing the name at the top of a promotional email blast. The industry is moving towards dynamic content and adjusting it to the insights marketers have gleaned about individual consumers. But are marketers getting ahead of themselves?
Premal Patel, commercial director of Catalina UK insisted: “We’ve got to have consistency of approach. Right now customers are happy to give data but they’re doing it for a reason. That reason being value. The problem is, they’re not always getting it.”
Booker emphasised the point by citing retargeting, exclaiming that a retailer keeps trying to sell him the same kettle.
Callicoot added: “Often once a person has bought something, the company forgets about the post purchase and customer lifetime value piece.”
Approach with caution
Strategies for approaching personalisation varied immensely around the table. This could be down to the relative stages of data management of the companies involved and their appetite for risk.
Personalisation itself is nothing new and the industry is at the point where it has the luxury of learning from its mistakes. As a result, Npower’s head of marketing, Edward Madden said: “We approach all of this with caution. Stories about getting it wrong resonate with us. We are famous for implementing new software in 2014 that sent out a number of wrong bills to customers. We are taking baby steps so that when we progress to advanced personalisation, we get it right. GDPR is going to be a real catalyst.”
It is tempting to launch headlong into a large-scale complex personalisation project. But there is still a lot of conversion ground to be made simply by getting the basics right.
“Amazon’s board initially wouldn’t sign off on the recommendation engine that’s now so famous,” explained ITV’s Forde. “So engineers limited it to two different laptops and hard-coded a recommendation for a laptop bag. Everyone wants a laptop bag when they buy a laptop. And it worked.”
The question of where and how to test personalisation to get a true view of its potential is complicated by the number of channels involved, as well as the data being gained.
The challenge is still felt in banking despite most choosing to interact via mobile, Morgan noted. “The typical customer goes into a branch at least four times a year. But they’re doing something transactional. In traditional retail settings customers are open to other potential purchases but the challenge in banking is to find the opportune moment to start a conversation about other products.”
More complex or big-ticket purchases in-store do allow marketers to reach out using data but it’s still a challenge joining interactions up. Dixons Carphone’s marketing manager Claire Playle explained: “Store colleagues can gather data – email, name, telephone – and if you leave without purchasing you’ll get messaging later. But we would still like to be able to marry that data with what customers have also looked at online.”
Where’s the cost, what’s the value?
In the end, attendees agreed that a lot of what was possible came down to cost and measuring the relative returns of various actions.
It’s not that marketers are unwilling to spend, it’s just there is an ongoing lack of certainty as to how to measure that spend and whether or not they are using reliable metrics.
Callicoot pointed out: “If we haven’t connected the dots [on the customer journey] how do we measure the ultimate impact of a strategy? What exactly has happened? What influenced their decision?”
Relay42’s solutions architect David Wendt suggested it should not be about ROI in the channel but ROI in the use case. “If you set KPIs, customer experience for example, that’s enough to take to the CMO and show the benefits of your strategy.”
We are taking baby steps so that when we progress to advanced personalisation, we get it right. GDPR is going to be a real catalyst.
Edward Madden, Npower
He he cited travel comparison site Skyscanner as an example. When shoppers abandoned searches to go to other websites, the travel site’s data management platform acted to deliver personalised messaging, pop-ups and timed emails to funnel the customer back to purchase.
In the end, attendees agreed that customers are keen to be contacted by brands as long as they have granted permission, the content is correct and appropriate, and they perceive some value in the exchange.
It was noted that while millennials are certainly more free with their information it would be foolish to presume they are also less hard to impress when it comes to communications.
Npower’s Madden said: “Finding the right balance is really hard. You have a lot more licence to get it wrong if you are a favourite brand. Unknowns have much less ability to get it wrong.”
Morgan added that navigating data and personalisation was still something of a mountain to climb but not a challenge that brands should shy away from, stating simply: “If we don’t do it, someone else will.”