Digital Transformation in Government Insight #58: Unauthenticated data personalisation

What is data personalisation?

There are a variety of personalisation experiences in action on most websites today. Data personalisation refers to building profiles of users from their data and then displaying content that’s personalised for them.

At the more basic end, personalisation can be based on data such as the user’s location, the browser they’re using (so content is optimised for that browser) or the fact they’ve come directly from a referrer’s website.

More complex personalisation will track a user’s current session and past sessions (plus anything else the site ‘knows’ about the user) to build a profile of that person and serve content that’s personalised for them. We also commonly see this information used to personalise advertisements (Google and Facebook are good examples of this).

Authenticated data personalisation

Authenticated data is from someone who’s logged in, so you know who the data is coming from. A good example of this type of data personalisation is the way Amazon displays recommended products for logged in users (based on the data it knows about you, which includes your past searches and purchase history).

Unauthenticated data personalisation

Unauthenticated data is from unknown users. It effectively tracks an unknown user’s behaviour on the website and builds a profile about them from that data alone. From this interaction, the data is used to display personalised content.

In terms of unauthenticated data personalisation, one of the key issues to consider is whether data is collected only from the current user session, or if it’s logged over past sessions for that user (based on their IP address), to provide cumulative data. Obviously the more information/data you have the better, so cumulative data is preferable.

Data personalisation and government

In the government setting, both authenticated and unauthenticated data are in play. Authenticated data comes from websites like myGov, when users login to access certain information or services. However, in many instances citizens don’t login when trying to find government information or services and so the data collected is unauthenticated (from an unknown user). In fact, unauthenticated data personalisation is definitely on the government’s radar, because the DTA recently posted a proof of concept opportunity, looking for a vendor who could work with the DTA to run a trial “to better understand the issues and opportunities”.

This is a new area in personalisation for government, and one we couldn’t find many examples of from here or around the world. For example, a recent Intermedium article (as part of a piece on the DTA’s proof of concept call-out) named some of the Australian government agencies currently using personalisation, but the focus was on personalised user journeys (personalising service delivery for a type of user/service), rather than data personalisation. The article mentioned Queensland’s Starting a Cafe trial portal and the Service NSW portal (which has its own user journey for opening or growing a cafe) as examples of government agencies personalising their user experience.

Overseas examples of governments using personalisation are similar, focusing on user portals and the online delivery of services — when citizens login.

Data personalisation in business

Probably the best place to find out more about data personalisation is from the corporate sector including ecommerce. In these areas, personalisation is much more advanced and widespread. While businesses often have authenticated data, sometimes they’re using data from unknown users.

For example, we came across a useful article called How to deliver personalized content to unknown users. This article gives five specific tips for personalising content to unknown users, although some of the tips don’t really apply in the government setting (e.g. based on ad clicks, or trying to match unauthenticated users to authenticated ones). However, there were two tips that could definitely be applied to government websites:

1. Behaviour — understand what content users have been viewing and use it to build personalisation rules.

2. Email click-throughs — personalising content when a person comes to the site from an email click-through.

Web analytics can provide a lot of detailed user data, that can be used to build a profile of a user as they’re on the site, and help serve personalised content. While email click-throughs will be less common in government, it is still applicable.

Personalisation tools

There are a variety of tools used to help personalise a user’s experience once they’re on a site. Sometimes these are linked to content management systems (CMSs), and sometimes they’re separate software tools.

A SmartInsights article on web personalisation and recommendation tools further separated web personalisation tools into four categories:

1. Personalisation integrated with web analytics (e.g. Adobe Target and Google Optimize 360).

2. Personalisation software as service (SaaS) for ecommerce (automated product placement and personal recommendations).

3. Personalisation within CMSs (e.g. Wordpress’s personalisation plug-ins and Acquia Lift for Drupal).

4. B2B, marketing automation, and publisher website personalisation tools (e.g. BrightInfo and Idio)

Obviously the personalisation integrated with web analytics and the personalisation within CMSs are the most relevant to government and both can be used for unauthenticated data personalisation.

Personalisation in the Australian public sector

Major Australian CMS initiatives such as the Department of Finance’s GovCMS program and Victoria’s Department of Premier and Cabinet’s Single Digital Presence are laying the foundation for future content personalisation. By consolidating government agencies onto a standardised platform, the Australian public sector will be in a better position to deliver more personalised content and services to citizens — whether that’s through user journeys, authenticated data personalisation or unauthenticated data personalisation.  

Salsa Digital’s take

The goal with unauthenticated data personalisation is to gather as much data about users as possible and then build user profiles/segments to personalise content. This can all be done in real time, so content is personalised based on the user behaviour. Yes, there are other ways to personalise the government web experience (e.g. portals and user journeys), but data personalisation is a tool that has the power to transform the way content is personalised on government websites.

Further reading

We came across these resource while researching this piece. Some were relevant and some...not so much. But they might make interesting further reading for you.

SmartInsights —Web personalisation software

Intermedium —DTA takes personalisation to the next level

R2i —How to deliver personalised content to unknown users

Masters thesis on egovernment portals

KPMG —Reimagine an age of more personalised services for citizens

ExchangeWire —Personalisation: An Untapped Opportunity for APAC Marketers

CMO —What AGL’s marketing chief is doing to embrace ambiguity

IT ProPortal —5 marketing predictions for a data-driven future

British Heart Foundation —Prime Minister challenges UK to transform care through AI and data science