Affiliate Marketing Tools

Marketers, You Are So Close to Doing Personalisation Properly. Don’t Screw It Up.

By Jonathan Rose

Marketers, You Are So Close to Doing Personalisation Properly. Don't Screw It Up. image Edited man with arrows

Earlier this month, we cast an eye on what 2014 held for content marketing and the shifts necessary to break it from the ideological stasis the industry has found itself in.

Today, we want to extend that to personalisation and examine the trends that are bringing content marketing closer to that nirvana of relevant and personalised delivery in the coming year.

If done right, 2014 will be a boon for brands that want to deliver personalised customer experiences that are relevant, useful and, ultimately, profitable. But to do this, marketers need to do several things:

Stop using useless ‘big data’ and start using content analytics.

With all the customer data websites are collecting, marketers should be on the cusp of being able to creating a personalised content experience across every customer touch point.

The problem is that up until now all that customer data – whether is be demographic, transactional, behavioural or social – has been inherently flawed for various reasons.

I am more than my demographic – a generalisation based on my age bracket, postcode and, perhaps, also my income as well. I am more than the product I’ve bought – indeed, my historic buys are rarely indicative of my future purchases. Social media has been a boon for marketers who have sought to better understand consumers at both an aggregate and individual level, however that too is fraught with problems: our social profiles are rarely reflective of our actual realities, instead we have a tendency to portray, a more flattering, unrealistic ‘curated self’.

However, my content consumption – what I read and engage with online – is highly reflective of who I am, what I’m interested in and what my needs are (both currently and in the future).

Content analytics is the process of structuring previously unstructured content, by extracting new information. It means computers can track an individual’s interaction with a piece of content and collect and draw trends about that individual’s tastes and interests. As content analytics practice matures – as it is sure to do as brands have to start proving the ROI of their content marketing – brands will be able to personalise a customer’s online experience by understanding what they’re likely to do or want next. For some example use cases, read our Guardian article here.

Accept that business rules are dead. Long live the algorithm.

If you’ve paid close attention to idio’s writings throughout this year, you’ll have seen us take issue with the popular opinion that “Content is King”.

Perhaps that was the case ten years ago when there were fewer online publishers, but now we are experiencing a content marketing deluge – content is no longer King, relevant content is (see our Econsultancy article on this).

One of the limiting factors of personalising a customer’s online experience used to be the need to customise the website content management system to be able to serve different versions of content to different audiences.

Web CMSs such as Adobe Experience Manager and Sitecore make it much simpler to define the complex rules needed for personalisation and to manage the content required to make personalised content experiences practical for many brands, however the fact remains that a lot of heavy-lifting is needed to creating the rules for these experiences.

Furthermore, the nature of business rules means that they are used to serve content to customers as part of a subset of a segment not as individuals.

The more savvy brands will begin to move away from using rules to power their content personalisation to using intelligent and adaptive learning algorithms that continually improve the content that is served to individual users based on their interactions with it.

Forget product recommendation engines, start doing content recommendation.

Recommending products via computer algorithms has historically been extremely difficult. Time was when Amazon had the monopoly on recommending products for consumers to buy, but now there are a whole heap of recommendation engines that can be bolted onto most websites and ecommerce stores that will provide Amazon-style recommendations on the fly.

These systems are getting better and better, with vendors such as RichRelevance or Barilliance now offering plug-in recommendation engines that automatically generate product suggestions based on user data, which are easily integrated into existing websites.

However, product recommendations are only useful if you’re in the market to make a purchase. Most of the time when you land on a brand website you don’t want to be ‘pushed’ an irrelevant or interruptive sales message of product. Indeed, it only serves to

We call this effect ‘interest abandonment’ (read more of our thoughts on this on Econsultancy) and the next generation of recommendation engines need to marry knowledge of a customer’s interests and needs (from the content they’ve been browsing) to make more intelligent content recommendations that engage – rather than rebuff – until they are ready to make a purchase.

As brands look to make their content marketing personalised for the end-user and profitable for themselves, marketers will need to lead the way in unifying the above trends.

Fortunately, the technology driving intelligent content personalisation getting smarter by the minute – let’s hope that in 2014, your content gets smarter too!


If you are interested in making your content marketing personalised and more contextually relevant to your customer, please get in contact – we’d be happy to have a chat!