In the technology news space, and definitely amongst the PredictiveIntent team, there has been much discussion over Eli Parisers book “The Filter Bubble”. You can read our previous post to understand the ideas put forward.
In essence, Eli argues that personalisation technology is creating, with only good intentions in mind, a world in which users are surrounded by news, topics and information that personalisation systems think they will be interested in. He argues that personalisation is placifying the average human by creating information bubbles, protecting us from news and topics that may upset our equilibruim. However, with more consumer reliance on data-connected sources to keep engaged with the world around us, personalisation is inadvertantly creating an air of dissonance with the world we live in – as we read news stories through rose-tinted glasses, we are more inclined to internalise these messages and therefore feel at conflict with reality when expectations are not met.
The Filter Bubble has been a must-read book for everyone in the office, and it’s created some thought-provoking discussions, which has lead to me (James, marketing manager) writing this blog post.
We deal with both e-commerce and content personalisation, recommendation and discover technology, and it’s suprising how few differences there are between different industries. Retailers, content aggregators, mobile operators, all have a mixture of the same goal – to increase revenue, whether by increasing the conversion rate, by presenting more relevant adverts in hope of a paid click-through or to increase time spent on site to maximise cost per thousand advertising rates. Putting relevant content that the right visitor is interested in at the right time can help do all this – and more.
But… herein lies the problem. The more relevant content put in front of the visitor, the more they will engage with it. The more they engage with it, more targeted, relevant content will be put in front of the visitor. This is the Locked Loop, and this is how it can kill your revenue.
The Locked Loop is when visitors are given the bare minimum of ways to escape from the Filter Bubble. In e-commerce, visitors can be highly profiled on their behaviour, but when they return to purchase a gift, the system doesn’t recognise the visitor intent and still attempts to adapt the purchase experience around favourite colours or brands, fighting a losing battle until the visitor clicks off site. For content sites, visitors attempting to declare a new interest, or read a story from another point of view can be warned off by opposing views and content.
It’s not just personalisation and recommendations that can keep your visitors in the Locked Loop. If your navigation and search isn’t up to scratch, visitors will simply give up and visit a competitor – if it takes visitors only 7 seconds to gauge their attitude to your site, imagine the damage to your brand that could be done if they are unable to find what they want in a reasonable amount of time.
As visitors try to escape the Locked Loop they lose the battle, give up and try somewhere else – the damage this can create to your brand and opinions of the experience you offer, and more importantly, the damage to your revenue, conversion rates and engagement will be difficult to repair.
However, it doesn’t have to be like this. You don’t have to lose visitors for good because your search and discovery mechanisms aren’t good enough; you don’t have to say sayonara to surfers stuck in the Locked Loop; there is no longer any need to say auf wiedersehen to weary visitors just trying to buy a gift for their grandfather.
Release your visitors from a regime of repetitiveness
It’s not difficult. The web is built around the notion of the long tail – providing everything in the knowledge that someone, somewhere will be interested. Sites have to allow discovery of even the smallest niche vertical product.
Too many recommendation engine systems, particularly those provided by academics, believe in the total control of artificial intelligence. That is, “black box” computer programs with little or no human input, deciding the best content to suggest based on what they know about a visitor and their behaviour. Whilst it’s a continual academic goal to give computers common sense, unfortunately it’s not quite there yet.
This is why our IntentPredictionServer uses a powerful combination of mathematical algorithms and predictive modelling with flexible attribute and business rule support to inject common sense into personalisation, recommendation and discovery decisions. It’s why we have some unique, patent-pending Suggestion Choreography® technology that looks to understand visitor contexts, such as whether they are a known or anonymous visitor, whether there is any profiled information available, if we can predict the goal of their visit through entrance analysis, and dynamically adapt the algorithms and filters used. It’s why we work with customers and agencies to hold discovery sessions, to find out how your users use your site and how best to accommodate and facilitate discovery and personalisation.
Giving users the chance to jump in and out of their filter bubble, to escape from or sink back into the Locked Loop and to find the products and content they want to see, using all of the valuable behavioural data that can be gleaned from your visitors, through all of the different discovery methods available will enable your business to dramatically increase engagement, conversions, revenue and a wide range of vanity and important metrics. But most of all, it optimises your visitor experience to encourage loyalty and repeat transactions, to enable visitors to connect with your brand through multi-channel touchpoints with experiences relevant to them.
By adding intangible qualities to your brand, a well thought out personalisation, recommendation and discovery strategy will only positively affect your chances of commercial survival.
What is the Locked Loop?…
There’s a pretty good low down on the Locked Loop here – http://predictiveintent.com/2011/08/the-locked-loop-why-your-recommendation-system-is-reducing-your-revenue-and-damaging-your-chances-of-commercial-survival/...
[...] to allow users to declare new interests, to ensure they don’t get stuck in what we call the Locked Loop, a continual and relentless regime of repetitiveness where users can’t escape from the Filter [...]