John Giles Interview pt.2

Data-modeller John Giles’ consultancy work has stretched from working with his home state’s fire service and the Salvation Army to private medical firm BUPA.

Having admitted his primary interest has been in data modelling, in the first part of our interview with the author of ‘The Elephant in the Fridge,’ I was keen to learn more about his approach to helping businesses get the most from their data.

John said: “I’ve often been surrounded by people who understand ‘mainstream’ industries better than I do, and some of these people shied away from the unusual challenges.

“That meant I ended up doing things nobody else would go near, and I loved it.”

Leaning on that experience , the Australian data-modeller, has a few golden rules.

John tips a big wink to Len Silverston’s universal data model patterns. He said: “They give you productivity and let the client learn from the hard-won lessons of the experience gained by others.

“I have a kit-bag of commonalities based on Len’s data-model patterns. I also respectfully treat each client as having distinctions that are at the heart of their competitive advantage.

“When it came to the Salvation Army, they had people who carried out what you might call social care activities, while there was obviously the formal church and worshipping side of the organisation.

“I had to ask: ‘Are there two enterprises here or are they one?’ The answer was one, but with two distinct arms.

“Conversely, with the fire service, they said they had up to 17 organisations that came together to respond to any given emergency.

“In 1983, I thought my family had been killed in a fire so I have a personal passion for effective fire management.

“Needing to map all the different responses to everything from a car crash to a major bush fire – and whether the response came by road, water or by air – it was a very complex structure to unravel.

“Similarly with BUPA, grappling with the concept of a “customer” was interesting. It could be a private healthcare member, or maybe a dental patient not associated with the healthcare insurance side of their business.

“Nonetheless, all customers needed to be identified and defined correctly to make sense of the business.[/vc_column_text][vc_empty_space][vc_column_text]“So my second golden rule is to always embrace emerging technology, but never forget it’s all about people.”

Obviously, many consultants are faced with a problem when the client has a set outcome in mind, or a fixed idea on the solutions required to achieve their goals – which are not borne out by the hired-hand’s viewpoints or understanding based on their own experiences.

A common problem for data-mappers is when a client looks for a data warehouse solution that merely maps the sources – rather than creating a Data Vault that maps the business.

John explained: “Problems faced when making such an argument follows on from the golden rules. One massive mistake is devaluing automation tools – and thinking developing your own might be quick, easy and cheap.

“The counter mistake is thinking that just buying shiny new tools will solve real-world business problems, instead thinking that data warehousing – and data integration – can be run by smart IT people, who don’t need the business folk as fellow colleagues on the same journey.”

John revealed he has spoken to Data Vault founder Dan Linstedt about such problems.

John has summarised two of the problems by classifying Data Vault practitioners as “pushers” or “pullers”.

“You have Data Vault pushers – people who find source data, they use automation tools to push the data into the Data Vault, but they have got no idea how it is going to integrate.

“They say: ‘Look what I found! I am going to push it into the Data Vault and we’ll worry about how it will integrate later.’

“They might have a customer here, and a client there. They might be the same thing. But they don’t stop and ask that question because it will slow them down. They are the pushers.

“Push it into the raw vault, build up huge amounts of data but with very little value because they have not asked the right questions.

“The Data Vault pullers say ‘Give me one group of people in the business that want some data that they wish to analyse, and we will build a data mart to suit their requirements – and we will work out what Data Vault data we need to support that.

“These “pullers” then work out what source or sources can provide it,” he added. “In contrast to the pushers who deliver little value – but can say ‘look how fast I am’ – the pullers actually deliver value in the short-term.

“But then somebody else comes up with a second or third requirement, or even a fourth, and they are starting to slow down because the requirements overlap.

“That’s because nobody thought about the integration. Focussing on integrating data around business needs should not be revolutionary.

“Your technology solution should evolve from evolutionary business needs and views,” said John.

Talking of evolution, how did John find his way into the computing and IT industry in the first place?

John explained: “In 1969, I was 19 and my father had died. The economic reality was I needed a job. I started applying and found an engineering job.

“I had no idea what the job really was. They said I needed tertiary qualifications and I didn’t have any. ‘We can’t offer you that position,’ they said. ‘But would you be interested in working with computers?’

“At that point I had no idea what a computer was,” he admitted. After being set a four-hour aptitude test, he rushed through it in 90 minutes so that he could meet up with his girlfriend!

“I passed and got the job at a rope manufacturer,” John said. “Their NCR computer was physically massive – the total storage of the mainframe was 1.6k – yes 1.6k. We coded in machine language in raw bits and bytes.

“Each member of the team largely worked in isolation. You spoke to the business, you did the requirements analysis, you did the solution design, and you did the coding.

“Then you did the testing, which meant going to the mainframe, and if it wasn’t turned on, you started it up.

“Then you loaded you program, ran the program, you got the test results and you went back with them. Because you were responsible for everything through to documentation and end-user training, you did the lot.

“My approach has always been I want to enjoy my career – I look at something and I think I reckon I would enjoy doing that. Let’s give it a go but let’s have fun too!

‘So many things have not been planned or calculated – there was no strategic evolution of my career. It has very much been about having fun, but also keeping an eye on what’s happening.

John said: “As far as I can judge, the lack of formal education made absolutely no difference at all in my work.

“In fact, I was employed by a high-profile consultancy that didn’t know I had no tertiary qualifications and I didn’t know that a degree was a pre-requisite for employment… but it made no difference.”

In the third and final part of our interview with ‘Elephant in the Fridge’ author John Giles, we will discuss his thoughts on the current trends in Data Vault and data modelling.