A few examples of when personalisation oversteps the boundaries, making customers uncomfortable in the process - plus 3 actionable steps to personalising pleasingly.
‘Creating a personalised experience’ sits atop most brands’ customer experience to-do lists.
And it’s not difficult to understand why.
Rather than serving up generic content to your visitors, personalisation enables you to take a tailored approach. Using the knowledge you have about your visitor - whether they’re a loyal customer or a new visitor - you can make their interactions with your brand smoother, more relevant and altogether more enjoyable.
This improved customer experience, using the likes of personalised content and offers, engenders loyalty, increases conversions and boosts your revenue in the process.
But - despite the undoubtedly improved CX that personalisation brings your customers - there are cases when personalisation oversteps the line. On these occasions, brands use the data they have about their consumers in a way that makes them feel uncomfortable.
The practice of this kind of “creepy” personalisation makes it entirely understandable why some website browsers are reluctant to share their data with brands.
As we saw in the panel discussion at our Yieldify Customer Community, the tricky part for brands is that internet browsers have different thresholds regarding when personalisation actually gets creepy.
However, in some cases, personalisation is so far over the creepy/cool boundary that it’s clear the executors of the campaigns entirely forgot the ‘personal’ element of personalisation.
We’re going to look at a couple of these examples before offering some practical steps to follow - along with an example - on how you can mitigate any risk of your personalisation being deemed creepy and instead produce pleasingly personalised content for your customers.
When personalisation crosses the line
The creepy side of personalisation initially came to light with the growth of ad retargeting.
The improvement in retargeting accuracy meant some website users began to feel like they were being stalked as ads followed them around the internet. And with this sense of unease, online “creepiness” was born.
With the increase in on-site optimisation, as well as email remarketing, there are now more opportunities for personalisation than ever. Of course this means that - if you’re not doing personalisation right - there are now more ways to upset your customers!
Here’s a couple of examples of when brands or companies have managed to do this with aplomb.
Feeling vulnerable? You’d better buy this...
In a New York Times article commenting on retargeting back in 2010, one reference stood out.
The interviewee, Ms Maitlin, described how after viewing a dieting service she was still being followed around the web by their ads:
“They are still following me around, and it makes me feel fat”
Worryingly, there’s evidence to suggest that this may not have been a coincidence...
One marketing firm have recommended targeting women with beauty ads when they’re at their most vulnerable.
The agency suggested women in ‘prime vulnerability moments’, such as Monday morning or when they’re crying, are more likely to buy beauty products.
Image credit: Adweek
Not only did they conceive this manipulative strategy, they then told the world about it in a press release. Whether or not they had a PR department is up for debate, but the fact this predatory personalisation is some of the creepiest we’ve seen is in no doubt.
Despite the outcome, the team behind this campaign may not be a group of inherently evil marketers, sat stroking cats and working out how they can make women cry more so they buy more makeup.
You can imagine they were asked how to get optimal conversions for beauty ads - but somewhere along the line forgot they were dealing with real people.
Target pre-empt personal news
You’ve more than likely heard this one before, but we couldn’t miss out this infamous example of personalisation overstepping boundaries.
There’s certain intimate details and events in our lives that we’d either rather keep to ourselves - or at least wait for the right moment to share.
This was certainly the case for one teenage Target customer when the brand inadvertently outed her pregnancy to her father with some over-zealous personalisation.
Deciding that expecting mothers were an under-targeted group, Target used their reams of data to work out which of their customers were pregnant based on their purchases. (Purchase patterns including non-scented lotion and cotton pads were found to accurately preclude customers giving birth.)
They then targeted the appropriate customer segments with pregnancy related products and it was a set of these personalised coupons that outed one of their 14 year old customers to her parents.
Initially the father was indignant and complained to his local store - angry that Target would send his daughter what he perceived to be wholly inappropriate content. However, after a little investigation at home, to his dismay he found the pregnancy-related coupons would be more useful than he first believed.
Despite the unerring accuracy of Target’s algorithm - this is a conversation they shouldn’t have been privy to, never mind starting. Let’s face it, breaking family news through personalisation is decidedly creepy.
In the interim, Target have tried to tone down the creepy factor by adding other, unrelated, products to their predictive promotions - say a hairdryer next to the baby formula - making the personalisation look like a happy coincidence.
Spotify forget to spell out the benefits
This isn’t so much an example of creepy personalisation. More a lesson demonstrating your customers’ perceptions around personalisation and what happens if you don’t make it clear why you’re gathering certain data.
Spotify are generally in tune with their audience, even providing weekly personalised playlists based on the music you listen to the most.
However in 2015, the music streaming service nearly wiped out the goodwill they’d garnered from their customers with some cack-handed new terms and conditions.
The issue arose when Spotify’s new Ts & Cs noted they would be gaining access to pictures and contacts (as most apps do these days). What they didn’t do, was say why...
Their customer base didn’t like it one bit - and social media was soon awash with claims Spotify had turned evil. After all, why would a music app need your pictures and contacts?
Clearly something sinister was afoot...
Spotify actually just wanted to use the data to better personalise their service:
The photos for custom playlist covers and the contacts to make sharing music easier. (You also had to click a box consenting to sharing this data).
After explaining these benefits the whole hoo-ha died down. But there’s a clear lesson here - spell out the benefits users will get from giving up their data. If Spotify had made this clear in the first place they’d have saved themselves a lot of hassle.
This Spotify story brings us nicely onto some actionable tips on how you can personalise - without upsetting your customers.
In his recent piece for CMSWire our CTO Richard Sharp outlined 3 principles to steer clear of creepiness when personalising:
1) Use personal information to offer choices, not restrict them
2) Emphasize the benefits when requesting or using personal data
3) Don’t treat everyone the same
To finish up: non-creepy personalisation put into practice
Sure, the above rules are easy enough to write, but what do these key tenets of personalisation look like when put into practice?
Our CTO points to The Google US Mortgage Comparison interface as a good example:
Point 1 - use personal data to offer choices, not restrict them
The site presents personalization as a value-added choice - enabling users to enter personal information in order to have the system recommend specific mortgages for the circumstances if they want to.
Point 2 - emphasize the benefits when requesting or using personal data
Google highlight the benefits of providing personal information very clearly upfront (“Answer a few questions about yourself and we’ll show you which mortgages to consider and why.”)
Point 3 - don’t treat everyone the same
Users who want to take the personalized journey can do so from the landing page and receive specific mortgage recommendations; but users who would rather not provide personal information can go straight into browsing a rate table instead.
If you can stick to these 3 principles you won’t have any problems with customers deeming your personalisation inappropriate. Instead, your personalisation efforts will build positive relationships with your customers and increase brand loyalty.
Our onsite and email remarketing enables you to serve your visitors with personalised content, at the right time - improving their customer experience and boosting your conversions. Check out how our tech could work for you here.