WRITTEN ON May 16th, 2008 BY Ruth Kennedy AND STORED IN Design: user-oriented, Foundation of Trust, Identity, Power of Information, What do we want?

We can’t say who said it or where, but the other night we had a meaty dinner conversation (apologies to veggies). Customer Relationship Management (CRM) hasn’t fulfilled its promise, and new ways of doing things are emerging. To what extent may this hold important lessons for government and public services?The big centralised databases of CRM have their limitations. There’s massive reduplication as every organisation seeking a relationship tries to maintain largely overlapping database records about the same person. Data is harvested with varying levels of consent and deception. Often as the data subject understands this process better it appears increasingly intrusive. Unease creeps in; even distinguished public servants treat CRM as a process which should be subverted eg with false middle initials or misleading information about buying power in the hope of making it work more in line with individual wishes.

At best CRM systems record a slice of history. Amazon’s state of the art CRM might know what you bought from Amazon, but not from Barnes & Noble, nor whether the books reflect your own interests or were a gift, nor how your interests have changed. CRM is not where buyers store their future intentions, and buyers don’t tell CRM systems when their needs change. CRM is expensive and the results can be disappointing. As one public servant with a huge customer base put it: “People don’t sit still long enough for us to devise a service mechanism to address them.”

Public services such as welfare, health, education and child protection are increasingly based on centralised databases. These must be universal, as the clients include everyone – there’s no picking and choosing of ‘customers’ in these parts of the public sector. They are often based on statutory requirements (for example to know quite a lot about a client to whom welfare payments are made). Over the last few years, the declared policy intention has been that such public services should also become increasingly personalised and customer-centric: the one-size-fits-all approach is no longer appropriate. Thus both the security-by-control and the personalised customer services agendas have depended on more data being gathered, aggregated and shared, setbacks such as recent data losses notwithstanding.

The news coming over the hill is that CRM is about to be turned on its head by a user-driven model of relationship data. In this “vendor- relationship management” (VRM) model the individual maintains the one comprehensive and up to date version of the ‘truth’, selectively sharing what they choose to with their preferred suppliers and partners. It avoids duplication, keeps the facts up to date, and can make marketing demand driven and therefore far more relevant and effective.

This is of course not how the world of marketing yet works, but contemporary web developments make it possible and a very strong case can be made that it is a huge and imminent trend. Straws in the wind would include search web sites and personalised health records such MS Healthvault and the Google/Cleveland clinic collaboration.

This change would be a fundamental inversion of how relationships are managed. But that’s not to say that it is universally applicable, that it is entirely “right” for public services or should be “sold” to government either in the political or commercial sense. But it does seem intuitively incontrovertible that the locus of personal data management should lie in many circumstances with the individual, and the availability of new tools to manage that data will be an important development for public services as well as the world of commercial marketing.

This raises deep questions about citizen and community empowerment, identity management, ownership of personal data and how we will successfully deliver personalised services.

A buyer-centric, citizen-centric or VRM approach to public services cannot be a universal answer for many reasons. There will remain a digital divide perhaps for generations. It’s essential that services are seen to be fair and equitable, challenging those who try to play the system, and that the benefits of new ways of doing things should not broaden yet further the gap between the haves and the have-nots. Such an approach is clearly inapplicable in certain areas: we’re hardly going to invite people to manage their own criminal records or history of traffic offences. But it is entirely relevant for any customer-like service relationship where the user chooses. There is a role for independently authenticated fields (eg a valid driving licence or NI number) when an essentially user-managed service is deployed, for example when hiring a car.

We caution against accepting a digital token or data field (such as a CRB check) as a substitute for common sense, whether it comes from a user-managed or a centrally managed system: no certificate of suitability removes the need for basic best practice procedures when working with children. Similarly we recognise that there are very real risks of aggregating data, whether it’s the comprehensive picture one individual needs to manage every aspect of their lives (based on perhaps 3500-4500 fields) or the aggregation of “slices” of data about millions of people’s health, child welfare or benefits.

If we look at the example of health records, some of this is professional confidential data. The data subject – patient – has to accept they should not be able to edit or alter their mental health history. But they may well wish to add subjective data, or share parts of their record with non-NHS care providers. This might give the patient a greater sense of ownership and responsibility over managing their own condition. Asthma sufferers might be willing to share their personal data with government (and other services) in return for helpful information about the weather, the geography of pollution and other knowledge which support self-management of the condition. The DoH might be open to interaction with personal health records such as the Google or Microsoft developments. Clearly there’s a risk of pursuing a centralised NHS approach to care records to the exclusion of what appears to be happening in the market. We could waste a huge amount of taxpayers’ money producing the wrong thing.

We’re grateful to the long-term care lobby for the insight that the risk of differential uptake (ie people moving forward at different speeds) should not stand in the way of progress. These are deep questions, and we saw that they challenged some of our deeply held assumptions about the journey to transform public services. That’s a good sign. And there is specific progress: the Government Gateway is piloting user- centric ID for local authority employees, and working on this with ContactPoint in children’s services.

How significant the potential for VRM in public services is perhaps comes down to how active or passive we see citizens as being. Are public services such as health, education or law & order only something we do to people? Or are these desirable community outcomes in which people and communities should and can play an active role?

It’s hard to avoid the conclusion that if government really wants to see personalised public services, choice and contestability, and increasing numbers of citizens actively engaged in creating and managing their own outcomes, it will need to be open to a more empowered user model. An openness to a user-controlled approach to information (even just in a few defined areas to start with) would also go some way towards restoring trust in government’s handling of personal data (or perhaps removing the requirement for such trust). The potential is great: let’s continue to explore how truly user- driven outcomes are likely to be based on user-driven data.