Personalization – The hard bit is the bet

Almost every day I ask myself why, after 15 years in this industry we are aiming toward the same goal with our clients that still seems unreachable – personalization.

No, we don’t call it personalization today, because that word is 15 years old and no-one wants to hear your 15 year old stories of ”it’s a bit like Amazon – only not the daft recommendations we all have a story about”. The cool kids talk about relevance, context and experience.

The idea is of course the same; you capture behaviour make some guesses on intent (implicit personalization), through integration you blend it with what you know about the visitor/customer/consumer/citizen (explicit personalization) and provide an experience that is seamless, trusted and intuitive. Voila! Uplift and customer happiness.

Why do I say 15 years? Well back in the day, when 15 years ago Vignette (you know kids, they used to be somebody) released Lifecycle Personalisation Server and Relationship Management Server, we talked the same talk, it could do just that.

Today, even though many organizations I’ve spoken to over the years are (by their own admission) stuck in the dark ages of solving the business pain of trying to just publish content – the truth is most CMS’s are idling way, with organizations using 20% of what they are capable of, probably badly.

So, the technology has been here – so why is it that my “personalized” experience of brands is limited to crappy ads in Facebook showing me what I just looked at or just bought? (more on that in this rant about re-marketing).

And lets face it, Amazon, the biggest remarketing sinner on my Facebook stream (oh for f**ks sake I just bought that), still has a rather crude recommendations experience, that sometimes hits the mark and other times offers stories that are not 15 years old, but just what happened to you yesterday (you know that recommendation for a little black dress as you bought something for your niece).

Today we also have unprecedented access to data, as people are now willing to share their wants and needs online to be harvested in exchange for a way to share cat photos.

So, we have technology, we have data, we also have bright folks in agencies and SI’s that can make this all work.

There are pockets of innovation here and there, campaigns, sub-brands and targeted banners – but, the most exciting a lot of digital marketers get is to play with multi-variant testing and let the crowd decide on which is the least unappealing piece of content.

In 15 years you’d think this would be mainstream, a core requirement of any web/digital project, table stakes while the true innovators crack the next nut of true multi-channel, artificial intelligence in content creation and curating or some other way of cutting through the noise feeding a message to my cerebral cortex.

So what happened to our dream?

The technology is easy; the hard bit is the bet.

With personalization you start making decisions about what content to push to this visitor, you are no longing hedging, sitting on the fence allowing the visitor to find (or not) the content, but making a bold play that THIS is the content for THIS visitor, possibly at the expense of a piece of broadly inoffensive content that tested well in the MVT.

With re-marketing you make lots of bets, you show a million people the thing that they just viewed on your website and 0.1% click through and buy it, while 99.9% think you are worthless and creepy.

That’s not an intelligent game, that’s like playing snap with a million cards and shouting snap every single time, regardless of whether the cards match – you win some games, but the other players are pretty pissed at hearing you shout “snap” for the 999,999th time.

Proper personalization isn’t snap with a million cards; it’s intelligently deciding what to play from your one hand (your website) to win this round.

In addition the bet requires you to think about who you don’t want to talk to. It stands to reason that the closer you get to be relevant for one visitor, the further you are getting from someone else’s interests.

You can please all of the people some of the time, or some of the people all of the time.

Scary stuff. In a world that has been traditionally measured by hits, popularity and not engagement.

With customer insight, analytics and social listening you can shift the odds into your favour. It doesn’t change the fact you need to make a bet – a decision about who to message and who you may exclude, but you can make data driven decisions about who these folks are.

The payoff is an engaged visitor who really gets you.

So place your bets please.