In my last blog, I explored the balance between the new digital marketing analytics and the more conventional CRM analytics. This blog was based on my wider work on how large corporations use insight. Interestingly, I recently received a notification from the Journal of Research in Interactive Marketing that my article on the implications of the explosion of digital and CRM data for the organisation of business intelligence, written with Neil Woodcock, Chairman of The Customer Framework,where I am Research Director, had achieved an unusually large number of downloads. The article was entitled “Interactive, direct and digital marketing – a future that depends on better use of business intelligence” is available.
Amongst other topics, the article explored the organisation of insight. In many corporations, users of BI are organised into departments like “Consumer Insight” or “Customer Intelligence”, replacing the older “Market Intelligence” or “Marketing Information”, reflecting the move from less targeted forms of marketing towards precision marketing and creating personalised customer experiences. However, the BI community is not always good at managing customer insight (CI) users. CI has emerged as a powerful BI user, but is still learning what to do with so much data, whether owned or non-owned (such as data arising from social media), structured or unstructured (e.g. voice, text, video). For some corporations, markets – customer needs, channels used competition, devices – are changing fast. Capabilities are changing too – witness the move towards real-time or near real-time marketing.
The CI community is right to demand better self-service, but an unrestrained and unplanned drive to self-service may lead to wrong conclusions due to poor understanding of how data arises or to use of the wrong analyses or tools. Improved planning and governance is needed to cope with a powerful, self-servicing CI community. We believe that self-service should be rolled out in a careful, prioritised and targeted way, using criteria of need and competence. BI’s experts, who know the software and are in command of metadata development and data definitions, should work across both communities. Their knowledge of what works well and what does not is crucial to the interface. Governance is needed at all levels of the organisation.
We noted that where BI and CI communities move in different directions or have different priorities, CI users do not get the support they need, while the BI community may take the route of technological optimisation rather than meeting users’ needs. If this is so, we argued that the company should either slow down deployment of new BI technologies, and/or ensure better alignment in plans and operational delivery for the two communities. Applying maturity modelling to the relationship between them helps, whether to look backwards to see how the relationship developed or forwards to see how to improve it, towards the nirvana of CI and BI marching in step, along a road where each step’s value is understood by the business and supported throughout it.
As excellent BI self-service tools become common and as users such as CI become more competent at using them, the question, “Who guards the guardians?” must be asked. In many corporations, the BI community’s role is to support users with data and tools for ‘self-serve’ data access and analysis. But what if users get it wrong? What happens if they are looking in the wrong direction when a tsunami comes, as many financial services companies were? Or should the BI community abdicate its role as custodian of ensuring that users use data correctly, in the face of a well organised CI community which professes expertise in using BI in marketing, sales and service, keeping the BI community busy with demands for new self-service tools to model, analyse, explore, view, measure, collaborate, virtualize, and then write back results, citing the advantages of more agile decision-making, reduced delays and frustration for users, and so on, and reports that show a strong correlation between self-service and good business results? These questions can only be resolved by partnership between IT and user communities. However, partnering processes often evolve over years, without careful consideration and analysis of how well they work and meet the demands of the user community or of the governance needed to supervise the partnerships. This must change.
Meanwhile, wherever the question “how far should self-service users be allowed to go” is asked, in circumstances where users are highly motivated to use self-service tools and where central BI teams are limited in resource (which seems to be the norm!), then a good strategy is for the central teams to use modern tool-sets to provide users with many different ways of selecting, analysing and forecasting using given data sets, but for the central BI team to maintain some control over the content of the dataset, so that the corporation as a whole can be sure of the validity of the results.
Of course, if you work in a digital marketing agency, happily using one or other analytics package to track the effectiveness of your activity for clients, you might not recognise this problem. But believe us, if you are managing insight in a large corporation, such as a bank, insurer or telco, you certainly will.
You can visit Merlin Stone’s blog here.
