There has been an argument over the past few years about real-time vs right-time data amongst vendors and analysts, who approach the question from the feasibility of technology, talking about ‘federated databases’ real-time ETL. real-time purchase history. For techies these are valid questions and interesting arguments. But, if the data doesn’t reach the point of decision making in a form for someone – whether a customer service representative, a salesperson, a manager –  to make a decision RIGHT NOW then the value of the information in the data has been lost.  

If instead we approach the question from the point of view of the business process then we get a different result. By building rules into systems which build on models and historical data, this data can be turned into real-time insight. By telling people on the front line what decision to make as part of their interaction with a customer, rather than providing them with information that may be interpreted differently businesses can ensure that corporate strategies and tactics are implemented on the front line. These best practices are then shared as service across the business whenever a particular task is needed to ensure consistency. 

Examples of organizations who could use this form of real-time insight include corporate banks who use models based on Basel II rules to inform the risk profile of loans offered, telecoms companies who can offer appropriate upgrades to profitable and non-profitable customers, and media firms who can make their ad pricing more efficient by sharing the services which provide pricing guidance across the business.

Providing real-time insight requires good data quality and in particular the use of master data management services across operational and analytic systems. Predictive models can mine the historical data and identify the appropriate models and segments based on incoming customer information which a rules-based system can direct to the right prompt on the salesperson’s screen.  

How much need is there for ‘real-time data’ in this model? To make the most of real-time insight, businesses actually need to know as much as they can about a customer BEFORE the transaction actually happens – based on purchase history, website history, information provided online. After the fact is too late to actually impact the decision – meaning that traditional ‘real-time’ data approaches cannot be acted upon to change the direction of the transaction in your favour.  

Really useful real-time data means that the data is being used to make a difference to the business. This means real-time insight, over real-time transaction history. Real-time insight gives businesses the opportunity to align their operational decisions with strategic and tactical objectives, meaning that operational effectiveness is increased.  


I’ve been rereading Re-imagine and inspired by some of Tom Peters recent thoughts on implementation so thought I’d put some random thoughts on excellence, implementation down . So what is  excellence. And how does it relate to success. Does excellence on its own = success? This is a great question – because success is such a personal thing. For me, a little formula is success=excellence + passion + execution. Excellence being ‘best in the world’, stretching your internal beliefs about your ability, passion being love! hate! the real reason for doing anything, execution being ‘get it done’ – the art of inspiring and organizing  for results rather than politics. Try stuff, don’t fear failure as it’s a step on the road to success. Success in teams needs all these elements, in all the people.    More later.

This was in response to a post on linkedin  regarding the usage of business intelligence.

Traditional business intelligence on its own is nice. It lets you know what you have done as a business, and a little bit on where you could be doing in the future. I 100% agree with the other posters about MDM/governance and stewardship as essential enablers to efficent BI by the way. As other posters have intimated, this can be of significant value in making decisons.

However, how much impact these decisions have on the bottom line cannot often be measured, particularly within the typical ‘reporting tool’ environment that is normally seen in conjunction with business intelligence. These decisions are also typically tactical – which company to sell to – or strategic – which market to enter, which is only a small part of the scope of decisions made in the organisation.

 To make the most of the information and analytics you have to hand, operational intelligence and operational decision making needs to be integrated to either automate or suggest the decisions made by customer-facing people, and crucially record the effectiveness of this decision so that the model/parameters it is made based on can be adjusted. Without this closing of the loop and the positive feedback, the power of the information in the organisation (and outside) isn’t being used valuably.

Suggest you have a look at James Taylor on the website attached who is evangelistic about this sort of stuff. Links:

goals again


I just realised that I’d published that and i haven’t really put any thought into my own goals – where I am and where I want to be. A task for the upcoming weeks.

I work for a well known consulting firm. The problem with this is, that everyone knows we do consulting, but have no idea about what we actually do. To that end, we’re aiming to get out and use conferences, workshops and articles in trade magazines to build up the profile of the firm in data quality and the personal profile of our guys as experts. It should be fun. I presented at Business Objects’ European conference earlier this year which went down well and some other guys have done similar, including pieces in the press. we just need to get it a bit more organised over the coming year.

Ive been setting out my phases and timing for the next 6 months before IM France, I’m going for a reverse periodisation plan, with a month of prep given my rugby injury 8 weeks ago. Given I’m back in London for the next few months it means I can get back into training without worrying about lugging stuff to an hotel.

 I’m also planning on doing a mark allen style low HR high frequency runs and am calling this my ‘patience’ phase in conjunction with my power phase on the bike, then aiming to up the endurance over the last 3 months from March.

 Is it the same as reference data, or is it much bigger, or smaller? Dan Lindstedt in another post I’ve linked to says how to implement it, but as David Loshin discusses here there isn’t really a clear definition. At the moment I think that Master data becomes reference data at the point that it is consumed by the end system – Master data exists, even when it isn’t being used for reference; reference data only exists in the act of referring!

Dan Lindstedt has been thinking up new ways of organising data for years – remember the data vault? MD as a service could be a big opportunity. Master data management across the enterprise is becoming more visible, MD as a service could be the idea that makes MDM deliverable to its customers. 

Will have to have a bit more of a think about this.

The coming wave in operational intelligence – decision and rule management, using business rules to make the transparency of how decisions are made. This is the first in a series. Worth a read and think.