Hyper Personalisation is Boring

Peishan Tan examines the increasing use of data to create a more personalised experience in marketing. She posits that while hyper personalisation helps increase targeted conversions, consumers are starting to feel fatigued, and discusses how we can make it better.


Hyper personalised content and ads risk becoming too predictable.

In the past decade, I’ve welcomed the sophistication of companies like Google and news platforms to mine my preferences and serve me curated news articles they know I’d enjoy based on my browsing history. At the same time, instead of searching for that proverbial needle in a haystack for book recommendations, I could just scroll through the list of books Amazon and Goodreads suggested, based on similar books I’d enjoyed. It’s efficient, saves time, and feels familiar.

Recently however, I’ve begun to feel a little stifled.

Amazon and Goodreads keep recommending the same tired list of books that I’ve already browsed through and dismissed. I’m ready to move on to different types of books, but the suggestions I get from those sites haven’t changed. For that matter, if I’ve liked a particular style of photograph on Instagram — natural light landscapes for instance — that’s the only variation I see in the “Explore” tab. And if I’ve looked at Canon cameras on eBay, these cameras will follow me on my meandering path online.

I’m also not the only one that feels this feels this way. According to MarketingSherpa’s 2016 survey with US internet users, 17% of respondents unsubscribed from emails because they found the “content of the emails is boring, repetitive, and not interesting to me.”

I get it. Marketers don’t want to to be tone deaf, and hawk products and services that I clearly have no interest in. If the Instagram “Explore” tab shows me endless pictures of duck-lip selfies, for instance, I’d delete the app in a jiffy. But I wonder, have they — we (putting on my marketing hat now) — gone too far in data mining to hyper personalise to people’s preferences? Are we trying too hard to sell to consumers what they want / might have wanted? Have we done too good a job in pushing the confirmation bias that we sabotage our own goals as marketers to expand to new consumers, consumers who might not have even considered the products and services we were trying to sell? Have we hindered the power of discovery?

It’s a little ironic that back in 2004, Steve Jobs famously said,

“A lot of times, people don’t know what they want until you show it to them.”

Have we since swung to the other side of the pendulum? With cookies tracking, big data mining, we now know better than ever before, exactly what people want at a certain point in time, and we are now relentlessly reminding them of that fact, stalking them across the interwebs, egging them on to purchase.


How did we become so hyper focused on personalisation?
Personalisation of course, is not a new thing. Marketers have been doing this for years, if not centuries. The difference now, of course, is more people than ever before are living their lives online. They are “window shopping”, actually making purchases, chatting, dating, and even being intimate online. These activities are way easier to track digitally — and with pinpoint accuracy. And processing that data has gotten much faster and cheaper, allowing more companies than ever before to crunch all that disparate data into a ‘single customer view’, allowing their marketers to analyze seemingly minute details, calculate all sorts of probabilities, and react / adjust on the fly.

From having a hazy, impressionistic idea of who your customer is, marketers can now paint a sharp picture of who their customers are — their age, gender, average income, address, marital status, job title, their social connections, how much research they’d done before committing to a purchase… you get the idea. The thought of all that knowledge is sexy, sort of like Clark Kent using his super-bionic-eyes-to-look-through-elevator-doors-sexy. For marketers, having this trove of data at our finger tips is incredibly stimulating. We can model the propensity for prospects to convert, and boy, have the leaders in the field shown us what is achievable with data mining.

Netflix, for instance, hacked the world of television production with its original US series ‘House of Cards’, in part because they’d trawled through their massive subscriber database, and noted users’ penchant for the British version of House of Cards, the actor Kevin Spacey, the director David Fincher, and their viewers’ love for binge watching television series. Their winning bid of $100 million over entrenched channels like HBO and AMC for the rights to produce the series had seemed rich at first, but their data analysis paid off, big time. Within a quarter of releasing the series, Netflix reported an increase in 3 million subscribers.

While I haven’t personally bought into ‘House of Cards’ (not a big Kevin Spacey fan), I have recently been nudged into action to buy a bag because of a company’s savvy marketing. I’d been researching online for a new camera bag, and almost bit the bullet on a Peak Design bag, going as far as signing up for an account to check out. At the last minute, I decided to sleep on it; it wasn’t strictly speaking a necessity after all. A week later, however, I received an email from the company, entreating me to rethink my decision, sweetening the deal with a 5% offer. I was sold.

So clearly, personalised marketing works. When consumers are actively in the market, the right message at the right time can and will easily help close the deal. But we have to be careful. What people have expressed interest in at a particular point in time may change. Our role as marketers is to gently nudge, but not stalk. It’s tiresome to be receiving the same message over and over again. People still like surprises; want to be surprised, to discover new things.

So every now and again, let’s challenge them, and break them out of their comfort zone with something different.


While we’re all for using data to personalise content and experiences for our consumers, here are a few guiding principles for making those experiences more enriching, gratifying and satisfying.

Guiding principle 1: Past behaviour does not indicate future preferences
Personalisation, to date, has focused mostly on past behaviour. We have to remember that people’s needs and desires change and are constantly evolving. They get married, divorced, have kids, change jobs, move cities, countries

Say I had a friend who had been sharing a Netflix account with his girlfriend. They’ve enjoyed watching endless horror movies, and Netflix had been doing a bang-up job recommending similar genres. But they recently broke up. Needing a distraction, my friend turns to Netflix, but he just sees the list of suggested horror movies. He doesn’t want — doesn’t need — the constant reminders of what used to be. Why not give him an out? Give him the power to reset, to completely wipe clean the slate of past behaviour?

Spotify does a great job of using data beyond just listening history to recommend songs to users. Within its “Discover Weekly” product, they leverage data mining to make suggestions not based on the individual listener’s past playlists, but based on the playlists of their millions of subscribers. Consequently, the song selection feels not only personalised, but also constantly fresh and exciting. Their “Discover Weekly” has grown exponentially since it debuted in 2015, from 1.7 billion streams to more than 5 billion streams in mid-2016.

Guiding principle 2: Nobody likes a nag
I don’t want to belabour the point but, nobody likes a nag.

If you know that someone has browsed your products repeatedly over a period of time, but have not responded to a gentle sweetener, have the discipline to move on. They just weren’t that into you, or more specifically, that product they were looking at. If you want to keep the conversation going, maybe offer something different. You could add a footnote with a link that allows them to get back to what they had looked at before, but don’t talk about the product again in the main copy.

Guiding principle 3: Every once in a while, burst their bubble

Following the polarising 2016 US elections, it emerged that people were becoming trapped in news bubbles, in which they only read news articles by writers with the same political leanings. This was due in part to readers’ own unconscious seeking of confirmation bias, but also in large part due to news media and social media’s conscious, constant feeding of personalised content, articles they knew readers they would be interested in. This is problematic, and news and social media are starting to have serious conversations about the need to balance the creation of giant, digital echo chambers, exacerbated by hyper personalisation.

To counteract this, BuzzFeed developed an experimental feature that quite literally tries to burst their readers’ bubble. Termed “Outside Your Bubble”, BuzzFeed has added a module at the bottom of popular articles in which they pull in opposing perspectives from social media channels. Their goal is to at least attempt to expose readers to alternative points of views.


Wrapping up

At the end of the day, our objective as marketers is to help sell products and services. To demonstrate that we understand the audience we are marketing to, we should still try to personalise our messages and recommendations — but we should still strive to do it in a way that is more enriching and empathetic to our consumers, occasionally surprising and delighting them, and allowing them a greater sense of control and discovery.

Holler is a digital agency based in Sydney that transforms businesses through innovative uses of data, design & technology. If you want to chat, drop us a line at hello@holler.com.au.