Put simply, personalisation is the ability to provide content and/or services that target an individual user. This can be as straightforward as displaying a welcome message with the identified user’s name to a more sophisticated mechanism to provide tailored content (and links to content) that is tailored to an individual’s interests or preferences. It is essential that personalised content is as relevant as possible to the end user as this is a powerful way of building customer loyalty and provides a more direct, one-to-one relationship between a website and a user. When done right, it also reduces the number of clicks a user has to make to get to content that is of interest to them and provides a mechanism where content can be served to the user that they may have not found by themselves. The result of this is to provide much wider access to the content on your website where a navigation structure or content is deep. In addition, a good personalisation solution provides rich and meaningful data about customer habits and interests through the use of analytics.

The Ixxus approach is all about identifying ‘why’ personalisation is important to you, ‘what’ the requirements are and given these objectives ‘how’ we can deliver a solution to best meet your needs.

Why?

The first step in the Ixxus personalisation process is to understand the goals that will make personalisation a success for your website; what is it you want to achieve by implementing personalisation techniques? For example, to: build a better long term relationships with customers and improve loyalty; provide a better user experience whereby content is easier to find and suggest possible topics of interest that the user may not have thought about and allow them to explore more areas of your website. This can not only increase loyalty but also visitor duration.

What?

The next step is to get a deep understanding of the all requirements that are relevant. This gives us a detailed view on the types of personalisation offerings required. This may include:

  • Product and service recommendations - this may be based on a set of user defined preferences (explicit personalisation) or by inferring areas of interest on behaviour (implicit personalisation) by tracking what the user has clicked, bookmarked or purchased. If a user is a member of a group, providing recommendations based on what other users of that group have clicked, bookmarked or purchased.
  • Personalised feeds - feeds such as RSS/Atom that are tailored to an individual’s interests, mailing lists/emails on relevant topics, customised forum/blog views that are filtered to a user’s requirements.
  • Personalised content - relevant content or links to content based on explicit or implicit rules. Content can include text, images, video, etc. as well as events, exhibitions, talks, etc.
  • Personalised messaging - this can include welcome messages or tailored message banners that can be targeted at specific users or groups of users.
  • Retail personalisation - customised products (e.g. build a CD with tracks based on a user’s interests), prices or promotions.

Once we have this information we can also determine the impact on other aspects of a website, such as:

  • The visual design – are there any changes required to the design of the website? If there are any targeted messages where do they need to be displayed? There may also be a need to add new sections to pages to provide specific options to the user, such as ratings or ‘bookmarking’ of pages.
  • Navigation – is there any additional navigation or modification of existing navigation required for options such as login/logout (and possibly registration, forgot password, change password, etc.).
  • What placeholders or slots on each page are required to display personalised content – is it only on the homepage or does it require changes across the entire site? How do these slots relate to the content on the page and to each other?

How?

We then need to understand the interaction required from the user. Does the user need to explicitly provide information about their interests? Should there be a rating system on the site whereby a user can score an article or piece of content based on how relevant or interesting it is to them? Should we track the behaviour using analytics and use this to infer other content of interest to the user? Is there a need to provide content depending on the location of the user? For example, this can be used to provide content such as events that may be happening soon to users who live nearby.

There are many other techniques that can also be used to provide information and metrics which can be used to provide a robust and detailed personalisation solution. We need to decide which technique or combination of techniques to use and weigh up the pros and cons of each to best serve the types of people who will actually be using your website.

The key to good personalisation is to build detailed knowledge about your users and to understand the interests of the targeted audience. This can be achieved in many ways including making use of any existing user behaviour information from web analytics. For example, this can be used provide a detailed list of topics that a user can subscribe to for explicit personalisation. Next we need to build accurate and useful segmentations based on aspects such as location, types of content of interest (such as events, talks, video, photography), role or user type (teacher, pupil, scientist, archivist/librarian), topics, etc. This defines information we need to know about the user. Once we have a good understanding of who will be visiting the website we need to ensure we have detailed and useful metadata associated with the content. Ixxus always strive to make best use of any information that has already been defined, such as keywords or website structure. During this stage of the analysis we would determine if there is any additional metadata required to enrich the content to provide the best personalisation experience to the end user. This allows matchmaking rules to be executed to deliver the best fit personalised content by a personalisation engine, such as the Percussion personalisation solution.

It is important that once you have a personalisation solution in place that you can easily track its effectiveness and to be able to make tweaks and modifications on an ongoing basis. Again, this can be achieved using technologies such as web analytic software such as WebTrends or Google Analytics. Other techniques can include questionnaires or market research. This provides vital feedback to ensure the quality of interactions and personalisation if kept at a maximum and does not end up providing the effect of de-personalisation caused by misinterpreting user behaviour. An example of this can be seen on some websites where you have once purchased an item as a gift for someone and then receive continuous recommendations based on that item which are of no interest to you.

By Ixxus going through this process with a client it helps to ensure a successful personalisation project that is maintainable and can be owned and understood by those who manage the system on a day-to-day basis.