Kineo Australia

Behavioural targeting for the real world

Organisations are becoming more and more sophisticated with dynamic and personalised email programs.

What then is holding back the same targeting technology on websites?

Hit me

Take the delivery technology out of the equation, and one might argue comfortably that there is not a big difference between serving someone an email and serving someone a webpage. I will explain why a little later in the post.

Personalisation of websites has been around for a while and done with various degrees of success. By ‘personalisation’ I’m referring to customisation of certain images, text, and logic within pages created dynamically for an individual. Ideally information that drives the logic is based on data such as:

• First time or repeat visitor?
• Referring source
• Pages viewed historically
• Transaction history (for eCommerce)
• Search terms used
• Location (Country, city) by IP address

The logic follows that each user can be served a unique experience depending upon any number of variables. No two users might see the same page.

How much?

Degrees of sophistication range from automated personalising of advertising or recommended content (simple text/image changes) to entirely restructuring the site layout and process-logic on a user-by-user basis ie. a completely different website experience.

US Online retailers that offer personalisation - emarketer

Additionally some sites, such as the BBC, allow the end user to control the layout and granularity of the information/presentation they are wishing to access. In this vein it is entirely possible that no two users see the same BBC Homepage.

Research released on personalisation across major ecommerce sites reveals that almost 3/4 of them do not feature any sort of recommendations on products – be it upsell, cross-sell or alternative product merchandising strategies.

One might wonder if this lack of personalised targeting within websites is a result of limitations of technology, priority or imagination? Or maybe all three.

The Amazon Example.

Amazon has been personalizing its websites for quite some time using a combination of electronic logic guided by occasional human intervention. Amazon however is in the enviable position of having a vast pool of history of transactions across multiple properties from which to derive some pretty sophisticated and well informed assumptions.

It makes perfect sense for Amazon, with its huge turnover and volume to be making incremental sales gains for personalised targeting.

Amazon however is operating a enterprise scale ecommerce model, selling finite inventory, with the objective of maximising margins and sales.
So how can personalisation logic benefit other online models, with much smaller volumes and what then are the levels of sophistication one can (or should) aim for?

Email versus Website

I would argue that most organisations that have been sending personalised emails have been doing some sort of website personalisation already.
Sending a user an email is simply no different to sending someone a page from your website. The difference being that the customer has requested you to market to them at regular intervals of the advertisers choice.

Email marketing – “I like your brand. Market to me. Tell me what you think I need to know“.

This is one of the most attractive sales models around and explains why Email marketing delivers such a high ROI.

So if marketers have got smart enough to send users a single bespoke page (in the form of an email).. then why are we still letting users navigate multiple pages on our website without any form of targeting, relevancy or sophisticated personalisation?

3 things holding back website personalisation and behavioural targeting

Number one – Technology

The truth is that behavioral targeting technology is not a simple beast. Creating dynamic personalisation, in real or near-real time, on a page-by- page or session-by-session basis takes some fairly sophisticated programming, hardware and architecture to deliver suitable results.

Very few content management systems or ecommerce platforms are capable of this.

The sheer complexity of the integration of propensity based targeting algorhythms with other equally complex factors such as server-load-capacity and caching, is a specialised field. This is arguably outside the scope of skills and experience of most technology vendors.

Number two – Imagination

The concept of delivering personalised content and website experiences requires a lot of planning, thought, data, testing and informed assumptions from experienced professionals in order to be effective.

Very few technology vendors have the skills or experience “in-house”.

Number three – Resources/Low hanging fruit/Not my job

Personalisation and targeting can require an investment financially and from a resource/time perspective.

Many organisations cannot afford the time outlay in particular.

When was the last time you saw someone with the job title “Analytics and marketing, cross datasource, technical integration architect manager“.

The concept of “best practice optimisation” and sophisticated behavioural targeting is simply outside the scope of most individual’s job charter. In short it requires the drawing of expertise from multiple disciplines.

It is increasingly obvious that someone within the organisation, ideally a CMO/CCO/CEO needs to drive this sort of initiative, with the assistance of third party expertise. Kineo has driven initiatives like this for our clients before with significant results and a proven, measurable model. We have the experience and knowledge and we are in a position to add value to any brand using these techniques.

Summary

Brands must recognize that producing one-size-fits all websites is effectively a blunt instrument.

Web Technology, analytics and expertise has come a very long way and behavioural targeting as a concept, process and technology has developed to the point where it is viable.

The technology combined with the intellect supporting it backed by commercially proven results are now at a stage where the concept and benefits to brands cannot be ignored.

Brands must seek to implement behavioural targeting, personalization engines, even at a basic level in order to improve acquisition and conversion and ultimately to ensure a better overall user experience.

In doing so, brands will get much better ROI from their web channel spend.

Jon Bovard