MaxBLOG

In the youthful days of the internet, marketers were preoccupied with building online real-estate. Getting the right domain, acquiring traffic and ranking high in search engines were the key priorities. The rise in online advertising spend is well documented demonstration of this phase; we’ve all seen the graphs with their exponential upward curves.

Once marketers got good at acquisition, the brightest ones began to reach the limit of the amount of high quality traffic they could bring to a site in a cost-effective manner. At this point, the focus shifted to analysing what these visitors were doing once they arrived at the site. Marketers wrestled the controls of web analytics tools away from the IT department and began to analyse funnels and abandon rates.

Now we reach the present, our top of the class online marketer has taken steps to resolve the big issues that analytics highlighted and is looking for the next way to keep ahead of his (or her) competition.

In the UK and European market, a growing number of forward-thinking companies are harnessing content intelligence to drive the next stage in on-site conversion uplift. Techniques such as multivariate testing, visitor segmentation and behavioural targeting allow marketers to place the right content, creative and offers in front of the right visitors at the right time.

Multivariate testing combines well worn direct marketing techniques with some intelligent mathematics and automated methods of changing site content on the fly. What’s more, with self learning techniques, Maxymiser technology can learn what content drives uplift and display it to more future visitors.

Visitor segmentation and behavioural targeting represent the upper echelons of content intelligence. By discovering what content works well for visitors segmented by various criteria, a site can take account of the fact that not everyone is turned on by the same things.

Behavioural targeting takes this a stage further and discovers the content that works best on a highly granular visitor by visitor basis. We can even pull in data from client CRM databases in order to get demographic profiling by postal/zip code as well as looking at past interaction history.

Peter sets out his full vision of Content Intelligence in our whitepaper on the subject, you can download that here – Introducing Automated Content Intelligence.

NMA’s feature “Another Way”, published last week explores the role of sub-optimal journeys in online stores and the guys were kind enough to include a quote from me concerning the fact that around 60% of offline purchases satisfy “latent demand”, products that catch the eye of the shopper as they browse the shop.

The idea of serendipity, finding one product whilst looking for another is great and Maxymiser cross sell optimisation can help marketers to exploit this in the real world. We would suggest that these additional basket items shouldn’t be considered random, there must be a process to guide customers (without them sensing it) towards likely cross sell and up sell items to really take advantage of this.

Just because demand is latent doesn’t mean you can use random activity to get people to respond. If that was the case Tesco would use dumper trucks to deliver their goods to stores and tip them in through the open doors so customers could have a good rummage. Shoppers would belay themselves up to the tomatoes and their trolleys would be 4 wheel drive to cope with the terrain.

Capitalising on latent demand is about creating linkages between products and services in a way that customers like and find interesting. Linkages occur in product attributes, patterns of consumption and market dynamics. Customers find them so interesting that they decide they want some of something they had no initial intention of buying. The resulting increase in demand for a product can be 100’s of %.

An example is where a product is subject to a price change or something happens to bring it out of obscurity and into vogue. A flight to city X might appeal to user Y when price Z is offered and it’s reviewed in the Observer on Sunday, which user Y reads with his croissants every week. Prior to this user Y thought city X was type of vodka and they are amused to find out it isn’t. So Mrs Y says what they hell, we need a break - lets go!

Peter Ellen
MD (services)

Our technology has included the capability to segment traffic to a site based on "web environment data", that's criteria such as browser type, operating system and screen resolution since launch. It's relatively easy to do because, believe it or not, your browser supplies this information to every web server it connects to unless you have customised your privacy settings. You might be wondering why this is of any importance. When combined, these three criteria allow you to determine what type of device is being used to view a website. That means we have the capability to optimise your site separately for visitors using different devices. A consumer browsing from their Nokia smartphone while on the move will respond positively to different content than a teenager at home browsing from a games console.

We can, in many cases, even distinguish between devices and serve customised content. For example, the optimal page to persuade an Apple iPhone user to convert to a sale could be subtly different to that which persuades a Nokia N95 to convert. There are also significant usability elements. If your website can detect that a visitor is coming from a small-screen portable device, content can be served differently to speed up browsing and allow the visitor to get more from your site in a shorter time frame.

This is all juicy stuff but it is a common sense progression of multivariate testing. With visitor segmentation, we are performing a series of multivariate tests in order to optimise content for each target group of visitors. It is not, as a competitor termed it earlier this week a "quantum leap" in site testing and optimisation although it can add significant uplift in conversions.

If we were wheeling out our marketing hype writing skills, Behavioural Targeting would be the natural "quantum leap" candidate. With this technology, websites and offers can be personalised to a visitor by visitor level. Our behavioural targeting is capable of doing that today, we can input client CRM data about past interactions and other demographic information and serve the optimal content to ensure that the possibility of each individual visitor converting is maximised. Segmentation by device is a powerful tool but it's by no means the state of the art in website optimisation; behavioural targeting is and surprisingly, that competitor can't offer that technology yet.