As you have probably noticed, if you’re a regular reader of this column, we’ve been focusing quite heavily on the use of CRM by the public sector in recent editorials (see How to develop a strategy for implementing CRM in public sector, Emerging trends in customer – supplier relationships, The impact of CRM & e-government on democracy in US and UK, and CRM Opportunities in e-Government). In the next two weeks I want to continue that focus by going down a level and looking at how some CRM techniques can be applied to the public sector. Readers should be aware that this is from a CRM practitioner who has worked in the private sector (primarily financial services) on CRM for the past 15 years or so. This may lead to some freshness of approach with regards to the public sector, but that freshness may stimulate new thinking.
Now the potential for using some CRM techniques including contact centers and Internet-based self-service sites in the public sector should be fairly obvious, so I’ve decided to focus in these editorials on some techniques where their applicability to the public sector is not so apparent. I have selected four: multi-channel event-driven campaign management and data mining (this week); customer profitability and aspects of multi-channel contact management (next week). Covering four topics in two editorials will necessarily limit the level of detail, but if you’re interested in exploring any of these in more detail, don’t hesitate to contact me (see email address at the foot of this editorial).
- Multi-channel event-driven campaign management.
As the architect of one of the earliest (perhaps the very first) multi-channel campaign management systems (Mind), a key component of the CRM environment, perhaps I’ll be forgiven for starting with that. Such systems enable organizations to manage the delivery of relevant communications across a variety of channels to a large customer base. As such it is associated in many people’s minds with classic Direct Marketing campaigns (mail, phone, and email). Now I’m sure we all (or most) of us will agree that the last thing we want the public sector to do is to add to the huge volume of mail shots, phone calls and emails we all receive already from the private sector. The communication channels are becoming clogged up with irrelevant communications from organizations trying to attract our attention, particularly with the growing move towards customer managed relationships (CMR) or buyer-centric markets (see Emerging trends in customer – supplier relationships).
However, note the use of the word ‘relevant’ in the description above. Campaign management systems can deliver highly relevant communications based on the use of a number of techniques to improve targeting, notably: segmenting the customer/citizen base into similar groups; measuring the level of need of each customer for particular products and services (propensity modeling); and perhaps most powerfully, the use of events or trigger to deliver the message at the right moment. The use of all these techniques can lift response rates from a classic direct marketing response rate measured in single-digits (1% or 2% frequently in financial services) to 30% - 50% when used effectively. The presentation, Creating Value through CRM Techniques in Retail Financial Servicesa, outlines how these techniques have been used in five financial services sector companies, including First Direct.
We may still feel some lack of sympathy with the idea of local authorities ‘touting for business’, even if highly targeted. However, a number of the concepts of campaign management appear to be relevant:
Life events: Financial needs, and public sector service needs are frequently triggered by a life event. Change of job, marital status, accident or illness trigger requirements for services. The public sector is doing good work (e.g. see www.leap.gov.uk) in identifying life events relevant to that sector.
Transaction events: However, although life events are significant, it is worth noting that in financial services, transaction events (e.g. deposit of a significant sum; changes in spending / saving ratios; return / non-return of application forms) have been as powerful in identifying customer needs. There may well be room within the public sector for using such transaction events in one service to trigger communications in other services (e.g. attendance at A&E department of hospital, triggers communications on rights to social security payments). This could be provide a useful means of coordinating activity across public services to provide a ‘joined-up’ service. Of course the data protection / privacy issues need to be addressed, but the benefits could be significant.
Fulfillment communications: One of the less glamourous, but highly rewarding aspects of campaign management systems is their ability to use events to manage contact strategies. A set of standard communications can be triggered by customer events such as the receipt of an application form; a particular telephone query, etc), automating many of the routine communications which can take up so much administrative time, as well as providing a means for managing these communications processes.
- Data Mining
Event-driven campaign management systems depend on segmentation and propensity models, as well as events to target communications effectively, and data mining is used (amongst other things) to build those segmentation and propensity models.
One of the more interesting omissions from the [email protected] consultation paper is the whole area of Data Mining as a technique for use in the public sector. There seems to be little understanding of its potential value to the public sector and I’d like to try and flag some of the potential benefits here. Of course, there may be concern about the data protection / privacy aspects but note that most of the potential uses are at the aggregate customer level, rather than the individual customer, so it is rather similar to the use of census data, but without the limitations of the census (data once every 10 years). So where are the potential benefit areas? I shall cover two: strategic decision-making support and segmentation and propensity modeling:
- Strategic decision-making support:
In the private sector, as the data available in data warehouses and customer databases grows ever larger, covering more of the business, that data becomes more and more valuable for as a resource for supporting strategic decision-making. Of course the data itself is not useful – there’s too much of it. However OLAP, and other M.I. tools are providing managers with the tools to analyse that data to provide the information required to support strategic decision-making (see Practical Data Mining for beginners...in a typhoon) for a slightly flippant example of this process). The ability to explore changes in customer behaviour over time; the take-up of particular products over time; and the aggregate activity in branches provide managers in retail financial services companies with the information required to make strategic decisions about which products to focus on; changes to make to the branch network; as well, of course, changes in the management of customer segments).
It is inconceivable that such tools would not be of value in the public sector at both local and central government levels, as well as in the public services. Indeed, once every 10 years, the government goes to immense lengths through the census to gather information to allow such analyses to be undertaken. However, trend and other analyses on transactional data gathered by the public sector has to complement census analyses, being far more timely and far less costly.
Why does there seem to be so little interest from the public sector? Well the most likely scenario seems to be the fears surrounding data protection and privacy again. However, it should be stressed that the value of this information is at the aggregate level, and not the individual citizen level, though of course we need to make sure that data privacy / protection concerns are addressed. But they should be addressed and not avoided by failing to make use of such powerful tools.
- Segmentation and propensity modeling:
Segmentation and propensity modeling are used in the public sector to breakdown their customer bases into different groups (or segments) with different needs and requirements with two main objectives in mind: firstly, to get a better understanding of the needs and requirements of the customer base; and secondly to deliver relevant communications to that customer base.
Now it seems obvious that there has to be real opportunity for a local authority or public service to improve the services it offers if it has a better understanding of the needs of the citizens and/or customers. The demographic profile of different local authorities ‘citizen bases’ is going to be significantly different, and hence the profile of services required is likely to be different. Understanding these differences will help local authorities to tailor the public services to the need of the local population (as well as, perhaps, allowing them to measure their performance against other local authorities of a similar profile).
There is perhaps a less obvious need for outbound communications promoting services provided by local authorities to specific customers, particularly if local authorities adopt a buyer-driven (or citizen-driven) model of CRM (see Emerging trends in customer – supplier relationships), which at first sight seems most appropriate
Nevertheless, the support to strategic decision-making, and the better understanding of the local population profile seems to make Data Mining a necessary part of the public sector’s CRM armoury. I wonder if both local and central government understand the significance of data mining to the public sector?
This week we’ve tried to look at some of the potential applications of CRM techniques coming from event-driven Campaign management systems and data mining for the public sector. Next week we move on to consider the potential use of customer profitability and multi-channel contact management in the public sector. As, always, we’re interested in your views on this subject, so please make them below or email me at mailto:[email protected]