What if public services could be more like Amazon’s?

May 14, 2015

For all the advances we have seen during the course of the last parliament in the development of better online transactions, they are not ‘user-centric’ but ‘users-centric’ (plural). They are designed so that anyone, anywhere will find them simple, quick and intuitive. This is huge progress, and a step change from how government operated in the past. Yet it remains a world away from our most positive experiences of transacting online: with internet retailers and tech firms, where user-centricity is at its most pioneering.

There is no government equivalent of Amazon’s suggestions for additional products or services we might like based on what similar users have bought. No part of interacting with government resembles Tesco ClubCard’s personalised offers, or Google’s services based on our location or past behaviour. Some consumers decry these as unwelcome intrusions on privacy by corporate giants. Our online behaviour suggests the opposite: more than 90% of UK internet users search with Google; at least 24 million Britons log on to Facebook each day.

All this should lead us to wonder: What would personalisation of public services look like? What would it take to do it? And would it be desirable if we could?

(Those interested in this area are strongly recommended to read WPP’s report: Me.gov: The Next Generation of Digital Government by Laura Citron).


Personalised services would:

Address the needs of the individual: The digital services offered by central and local government are still largely designed around the information needs of government and its siloed departments. Applying the online retailer model would flip this around to what a specific person wants or needs. The information and services I see should not be the same as those used by my parents or grandparents.

Spot gaps: If government had a joined up view of an individual’s interactions with different public services, it could spot needs that were unmet. It would ensure that vulnerable people who depend on support from multiple organisations do not fall through the net. When more than 30 separate organisations work together to support a single troubled family, shared information would focus on the families’ needs first. Pioneering work in Chicago social services has demonstrated how this can be done securely.

Enable digital nudge: Personalised government would suggest other services based on a person’s age, demographic or location. This kind of technique would offer considerable scope to expand the use of nudge factors: “The last X many people who completed this transaction also found Y helpful…”. For example, why not be prompted to sign up for organ donation or giving blood when renewing a driving licence online?


Many are put off the idea of personalised services because of the implications for personal data privacy. The reaction to Care.Data and ID cards confirmed for many policymakers that the public are simply not prepared for government to combine and analyse their personal data. Two things could help address this:

Make it about the citizen benefit first. In the 2015 Budget, George Osborne announced the ‘death of the self-assessment tax return’. Instead of an individual having to enter details such as their past year’s salary via PAYE; the amount of interest they earned; and their student loan repayments, the government will pre-populate the form (as it already collects that information). Citizens will then simply check that the details are complete and correct.

The move received a universal welcome by the self-same media outlets that decried Care.Data. Yet it will work by using analytics to intelligently join up an individual’s records across multiple different IT systems in different departments and agencies (HMRC, DWP, Student Loans Company, etc.). The difference in the reaction to Care.Data is that the initiative provides an immediate, personalised and tangible benefit for citizens. That citizen-benefit-first model should be used in future initiatives.

Personal Data Stores. As I argued in Small Pieces Loosely Joined, if citizens are truly to feel confident about more ambitious uses of their personal data, the long term solution may be Personal Data Stores. Starting with our medical records, we should have access to and control of our data so that we are in charge of which organisations share and use our information to deliver more personalised services.


In a word: yes. Three benefits stand out.

  1. Savings: At a time when the public sector is aiming to save £billions, let no one pretend that business as usual is an option. One-size-fits-all digital services will not be enough to change that. Personalisation could help pre-empt need and address problems before they become serious and expensive to resolve, moving to a more effective model of prevention rather than cure.
  2. Making services better: Why shouldn’t citizens be delighted and surprised by government services? Given that an increasing amount of our experience of government is of digital government, the impression we receive from our online transactions will become ever more important in how well served we feel as citizens. As government seeks to help busy, hard working people, services that better meet their exact needs, and are well coordinated, will surely be valued.
  3. Supporting the vulnerable: The privacy lobby likes to cry foul whenever there is any suggestion that personal data will be joined up across government. Yet given that 80% of public services are directed to the poorest (and often most vulnerable) 25%, the right to privacy can all too easily end up being a right to suffer privately. Citizens have every reason to demand that government’s right arm knows what its left is doing. Those working in the Westminster bubble must remember that most people do not think about DWP, HMRC, DCLG and BIS, etc. They just see: government. We would not tolerate such disjointed action from other providers.


Of course, it would be naïve to deny that there are significant challenges. Some big questions need to be answered:

  • Is it acceptable if some people choose to opt-in to personalised services, and everyone else gets the standard version?
  • Would this lead to two-tier public services? (And would it matter if it did?)
  • Should private sector data be included in public sector transactions? If so, how?
  • What are the ethics in delivering services based on predicting people’s needs? Could this lead to ‘predictive determinism’ where people’s situations become self-fulfilling?

It is incredibly hard for government – any government – to lead on this agenda. They would far rather respond to a demand for personalised services from the public. While that is understandable, the old business model of running public services needs to be overhauled. Personalisation, done carefully and prudently, is one possible innovation that is surely worthy of further investigation.

Join our mailing list