Data usage & personalisation: when Big Data gets creepy

Marketers are walking a knife-edge when it comes to balancing which customer data to gather and how to use it appropriately. Use it too little and it can become a problem (questioning why you asked for it in the first place) and use it too much and it can easily become creepy and damaging to the brand. A consumer survey from Gigya found that 44% of people ignore all future communications from brands that do not target them correctly.

Consumers want more personalised marketing – let’s be clear on that. Customers will gladly hand over personal data in return for something of value, such as more relevant and personalised emails. A study by YouGov found that almost 7 in 10 consumers are willing to share such data for personalised deals and better customer service.

Big brands, particularly those that we interact with on a daily basis, such as Amazon, Google, Apple and Netflix, gather a lot of buying and behavioural data about us all. And we trust them to deliver a personalised user experience in return. However, does there come a point when using all this intelligence can become creepy? And, if so, how can marketers avoid it?

Data usage – how much is too much?

Cheetah Digital ran its first Data Masterclass for digital marketers in London recently, and a panel discussion examined this area, specifically where personalisation becomes too much. I was joined on a panel of experts by Justis Saayman, principal marketing consultant at Fresh Relevance and Steve Kemish, tutor at the IDM.

My view is that, ultimately, the issue is all about context and marketers using data appropriately – but make sure you DO use it. Popular questions from clients include ‘How much is too much?’ and ‘What becomes creepy?’ when it comes to data usage. I’ve seen the full gamut of answers – from simply putting a first name in the email subject line being seen as too personalised, all the way to content appearing too relevant and intrusive.

Not only is the ‘right’ level going to be different for every brand, but it’s going to be different for every customer for every brand. So each consumer is going to find a certain level of sharing and returning that is appropriate and that’s the thing that nobody has really struck upon – what precisely is the magic balance?

Think of brands like Google or Apple – we give them tremendous amounts of data and they give us good services in return. Until now, I don’t think consumers have ever been concerned with the security of that data, how that’s been used or even whether it’s been passed to anyone else (although we arguably should be, especially in the context of GDPR which is all about the right consent for data capture and how to use it).

Marketers should not limit their activities based on concerns that the audience might not respond well to it – that’s what testing is for. But I do believe it’s OK to send someone an email about their child’s birthday, for example, if they have shared that information with your brand and are a loyal customer. For example, if that was a really special thing I got for my child last year and the email is personalised in that way, it’s totally appropriate. Think of it from the consumer’s viewpoint – they gave you that data so why aren’t you using it? If you didn’t use it and you sent them something generic they might wonder why you didn’t personalise more because you asked for that information in the first place.

Here’s what my fellow panellists had to say on the subject:

Brands needs to focus on convenience

Justis Saayman.

Justis Saayman (@justissaayman), Fresh Relevance, added: “In today’s world, Jeremy Waite, an evangelist for IBM Watson, came up with the concept called ‘tinderisation’ which works on the majority of brands out there. Literally, I can swipe left or right to visit another brand – so, if I don’t buy a product from you, or the price isn’t acceptable to me or I don’t like the message that you’ve got, I simply swipe left and go elsewhere. Immediately, you’ll lose that purchase or lose that customer.

“Human attention spans are now around five seconds, or the time it takes you to read a few sentences, so the danger is that once the attention is lost, visitors move over to a competitor. Therefore, grabbing somebody’s attention has become far more important. So as a brand, the more you know about somebody, the more relevant content you can put in front of their faces – the more you can surprise and delight them. This actually works to the marketer’s advantage a lot more in today’s world than ever before. From baby boomers, to millennials, to generation Z, that attention span is short across the human race – and knowing things about what people want gives the element of convenience, thus increasing brand value as well as, ultimately, product value.”

So, just when do brands know too much?

Saayman added: “I think most won’t ever reach that level. However, the largest brands need to highlight the convenience of gathering all that data. Amazon, for example, is trusted because you can buy something and have it delivered the next day. Whereas, for some brands, knowing that your kid is going to be 11 years old next week might be considered creepy – but knowing that they frequently visit a certain area or that the best product to buy next would be X or Y – adds to the convenience. Personally, if I don’t have to go and look for it on a site, and I can get it immediately on my phone or PC, it’s a benefit.

“So, in short, don’t worry about it!”

Value exchange is vital for engagement

Steve Kemish.

Steve Kemish (@Skemmo), representing the IDM said: “I believe that attention spans are all relative. It all comes back to the relevance of what the brand does. Consider the brands you engage with, you’ll give more time to those that give you more back.

“If you’re getting bombarded by messages from brands and they’re just not relevant or useful to you, you will switch off and your attention span will be much shorter. But we commit more to organisations where we normally see value. Especially for all digital communications in a very time poor world, any of us can get better results by just being more helpful and adding more value to the touch points we use – whether that’s in their inbox, their newsfeed, or wherever it might be.

“So prioritise what you know about users – critically, you don’t want to be considered creepy. For example, knowing that Dan’s opened an email at 2.37am and messaging him with ‘Hello Dan, we noticed you opened your email at …’ becomes disturbing. However, some brands are particularly good at building lifecycle marketing and lifetime value, such as Mothercare/Pampers. They understand that if you’re pregnant and maybe you are in your first trimester of pregnancy, whether it’s your first child or your third child – it’s an exciting moment and you want to tell people. At that point, trusted brands can ask and get data from visitors easily, because you can give them back something of greater value – ideal if you’re lacking in data; simply look for the touchpoints around your business.

“Successful lifecycle messaging includes, ‘You’re pregnant, it’s your first time, here’s all the things you might need to think about’. Also ‘Here’s the things you’ll feel’ and ‘Here’s the stuff you can do’. Then, ‘By the way, here’s some of the products we sell, but don’t worry about those too much… when’s the big day?’ You then get a data point that says they are due on this day. Continue with ‘Upload your first photo of you and your baby and we’ll build you a virtual photo book’. All these things don’t cost much, but in the background are building value, in the perception of the consumer.”

Kemish summed up: “In terms of data (mis)use, typically I don’t think humans sit there and say: ‘He’s misused my data’. Unless you flagrantly misuse it – so long as you’re adding value people don’t really question it. Think about on and off-line – where are the points where you can ask somebody and you won’t get smacked in the face! They actually want to tell you and then, in return, you deliver them something back. I think attention spans increase greatly when customers feel like they’re getting help; you’re building a relationship – and, importantly, you’re adding value to their day.”

Testing and transparency will achieve personalisation goals

Undeniably, personalisation produces valuable return for marketers – a recent study conducted at Stanford Graduate School of Business found that adding a name in the email subject line increased the chance of the recipient opening it by 20%, increased sales leads by 31% and reduced unsubscribes by 17%.

Yet, it’s important not to overdo it. In short, it’s really down to striking that personalisation balance and sensitivity.

If you think it’s too much or you start getting complaints, then dial the personalisation back and make it more implicit. There is nothing wrong with asking your audience, ‘Would it be OK if we did this?’ In addition, when you capture the data initially you could say to them, ‘We will use it for this kind of thing, is that OK with you? Or would you prefer it not to be used in that way?’

That’s perfectly acceptable for setting clear expectations for what’s going to happen.

Have an opinion on this article? Please join in the discussion: the GMA is a community of data driven marketers and YOUR opinion counts.