Montage Blog

The CIO who built a career on big data

Tue, 03 Nov 2015 10:30:00 +1300

A recent article in CIO tells the story of our very own Chip Felton's career path from social worker to CIO in New York to principal consultant at Montage.

"Chip Felton is cognisant his career trajectory to CIO was not typical.

Felton, who is originally from the United States, started as a social worker, doing direct counselling work with clients, while finishing his master’s degree in social work.

He rose to senior deputy commissioner and CIO at the New York State Office of Mental Health.

Seven years ago, he moved to New Zealand and is now a principal consultant at Montage, providing business intelligence consulting services to organisations across New Zealand.

Now based in Nelson, Felton relates how this multifaceted background is a great backdrop for his current role."

Read the full article here

Get ready for the next major release of Tableau - version 9.0, coming soon.

Tue, 17 Feb 2015 11:30:00 +1300

Tableau version 9 is scheduled for release in March - right around the corner! I've been beta testing the software and there are lots of productivity enhancements as well as new features in both Tableau Desktop and Server: drag and drop analytics, smarter maps, easier and expanded calculated fields, and much faster performance - the list goes on and on. Perhaps the most powerful new feature is something called Level of Detail (LOD) expressions - these allow you to define the level of detail (aggregation) at which a calculation should be performed. See this link for a video example of how LODs work: https://www.youtube.com/watch?v=BnSF6i4MW0E

Business Intelligence vendor Yellowfin receives 2014 Australian Excellence Award

Tue, 16 Sep 2014 09:30:00 +1200

Global Business Intelligence (BI) and analytics software vendor, Yellowfin, has won the US Institute for Advancement of Trade & Commerce’s (USIATC) 2014 Australian Excellence Award for the BI industry.

Each year the USIATC conducts business surveys and industry research to identify companies that have achieved demonstrable success in their local business environment and industry category.

“This award signifies Yellowfin’s growing marketplace presence and penetration, not just within Australia, but the US and other major global markets too,” said Yellowfin CEO, Glen Rabie. “Such recognition is testament to Yellowfin’s ability to meet growing market demand for BI solutions that balance the self-service needs of end-users with the governance requirements of enterprise IT.” Click to read the full article.

2014 Gartner Magic Quadrant for Business Intelligence & Analytics Platforms

Thu, 24 Jul 2014 09:30:00 +1200

Each year, Gartner analyses every vendor in the Business Intelligence and Analytics market. Their research has particular significance because it often identifies the innovations that drive the market. As part of that report, Gartner also releases the Magic Quadrant, which shows the relative positions of the market's competitors. Tableau once again is recognised as a Leader in the Magic Quadrant for 2014.

The trends and innovations that Gartner has identified this year are familiar ones. Business Intelligence customers have continued to reward BI platforms that are easy-to-use, accessible and readily integrated with existing systems.

Montage is a silver partner for Tableau Software in New Zealand, the first in New Zealand and one of only 15 Premium (Gold and Silver) partners globally.

To download the full Gartner report, click here.

National Health Informatics NZ Conference (HINZ)

Tue, 03 Dec 2013 12:34:21 +1300

Montage recently attended the National Health Informatics NZ Conference in Rotorua. The event is a great opportunity for a large range of organisations and individuals to gather together each year to share progress and experiences whilst taking a look at the future direction of IT within Health.

Key messages taken from the conference:

Information being shared across Health sectors

- A number of regions are working on solutions to share patient information between health care providers. In Canterbury a shared care record view allows health professionals from GP’s, community pharmacy, DHB, Community care to access up to date patient information.

- At Montage we are assisting a Primary health Organisation to obtain a consolidated view of patient information across all regional general practices. Some of this information is being shared with the local DHB emergency department.

Great strides have been made over recent years and there is a real sense of optimism regarding what is possible.

- Our years ago, there was a great deal of discussion and acceptance about what needed to be done. At this year’s HINZ there were a number of solutions displaying what has been done and just starting to touch on what is possible.

An openness and willingness to collaborate

- A number of DHB’s were presenting solutions and actively looking to share learnings and specifications.

- At this year’s conference NMDHB presented a Pharmacy and an Inpatient dashboard developed in conjunction with Montage. To demonstrate the benefits of collaboration these two solutions have since been ‘re-plumbed’ by CDHB at a fraction of the effort and cost.

Best Practices for Designing Efficient Tableau Workbooks

Thu, 14 Nov 2013 12:34:21 +1300

A must read for ALL Tableau Software ethusiasts, Alan Eldridge, introduces some best practice ideas for fine tuning and improving your Tableau Software workbooks. Read More

Success Through Partnerships - Nelson Marlborough District Health Board

Tue, 12 Nov 2013 15:31:24 +1300

Decision Making at the Speed of Thought:

Table and chart reporting is a mainstay of hospital clinical information, however clinical users need data visualisations to better understand disease trends and care patterns. Well-designed dashboards and presentations are more intuitive, and can provide the ability for users to answer their own questions

How Measurement Drives Behaviour - Four Part Series

Tue, 16 Jul 2013 12:34:21 +1200

In a series of papers, author and industry expert Frank Buytendijk examines the people side of business performance and business intelligence. How does measurement affect human behaviour? How does organizational culture contribute to business intelligence? What does a rational decision making process look like? Which number games do we recognize in business? What is the best methodology to manage business performance?

HRIS The Road To Success

Thu, 11 Jul 2013 16:20:15 +1200

When evaluating your current and future HRIS needs, it’s important to align people, process and technology for the best outcome.

Montage's Simon Lind outlines the approach to take to move from being a cost centre to a value centre. Read More

Information Visualisation

Tue, 23 Apr 2013 10:53:37 +1200

Charts and other graphics are a powerful means for sharing information derived from data. But how do you decide what chart or other graphical representation to use to best convey that information? Are such decisions just a matter of personal preference, taste or aesthetics? Or is there a firmer, more scientific basis for choosing the best visual representation?

Montage's Chip Felton introduces us to the science underlying today's Information Visualisation best practices.

Success Through Partnership - Christchurch Casino

Tue, 26 Mar 2013 09:08:42 +1300

A good client story which focuses on the value of investing in a review process (HRIS Discovery) to save time and money in the long run...

The second annual Getting IT Right Survey by IDC

Tue, 12 Mar 2013 09:29:18 +1300

The second Getting IT Right Survey finds greater misalignment of IT with business strategy, a key cause of project failures.

Data Governance Whitepaper - Part 1

Mon, 17 Dec 2012 15:31:24 +1300

With so much talk and focus on Data Governance do you find yourselves asking these questions -What is it? What should I do about it? How does it apply? How
will it affect my world at work, my team, my unit, my organisation, all the processes, systems and technology within it? What is the benefit? Will I see it?
And on goes the list...

With this 2 part white paper I would like to introduce you first to what Data Governance is all about and why is it essential and, secondly the framework and implementation methodology.

Demand-Driven Business Intelligence

Mon, 17 Dec 2012 15:31:24 +1300

Demand-Driven Business Intelligence

Few areas of business are immune to the forces of supply and demand. Business Intelligence (BI) is no exception. BI Teams supply a product, information, which aims to meet the demand coming from information consumers. Of course there are differences between BI and widget-making: information is less tangible, and BI’s end consumers are often stakeholders internal to the organisation. But the same principles apply, and we would be wise to heed the lessons of supply and demand from more traditional industries.

Being Demand-Driven

One important lesson from industry is that product design and manufacture should be driven by demand. Even in Agriculture, one of the most traditional of industries, there is a strong drive towards demand-driven production. It’s no longer a case of just rearing as many sheep as you possibly can, herding them into a truck and hoping for the best. These days every part of the process, from conception to final packing, has a deliberate focus on the requirements of the end consumer - busy shoppers in London or Paris.

In the same manner, BI should be demand-driven. “If you build it, they will come” may work in the movies, but in BI it is merely a recipe for disappointment, even disaster. Sadly, lacklustre project outcomes are not uncommon in BI – in fact industry analysts Gartner suggest that around 50% of BI projects fail to deliver, and the most common reason for this is the failure to understand and deliver to the requirements of the organisation.

How can BI be more Demand-Driven?

Here are two simple approaches that every BI initiative should include:

1. Use a Translator

If you liken the BI team to a bunch of Kiwi farmers, then the CEO is an uncompromising Parisian chef. In the case of farmers, to be demand-driven they need to understand their consumers and what they are demanding. To get the best result from an exchange with a Parisian chef you would be wise to involve a translator – someone who can not only speak the language of both groups, but who also understands the realities of both the paddock and the kitchen. The same is true of BI – the demands of the CEO (or other information consumer) need to be captured and understood, but the BI developer is not always the best person to achieve this result. The ‘translator’ in the BI case is usually a Business Analyst (BA) but the concept is the same: the BA’s job is to translate organisational requirements into language that makes sense in the world of the BI developer. To do this effectively, the BA needs to not only understand the practicalities of BI, but also have a good awareness of the overall organisational context as well.

2. Nail Down the Functional Requirements

If you’re developing a new smartphone to take to market, you need to have a clear idea of what the consumer wants the product to be able to do: it needs to make calls; it must have GPS, and so on. The statement of functional requirements of the phone (i.e. what it can do) is the principal point of contact between the demands of the phone consumer and the engineering that actually puts it together. A BI project is no different – the statement of functional requirements should be the core document that spells out what is “demanded” of the BI project, and be the primary point of contact between the information consumers and the BI team. Since BI is an information-based product, the functional requirements will usually involve statements concerning the information content that will be provided, or perhaps even specific questions or analysis that the BI deliverables will address.

There are many benefits to a carefully constructed and detailed statement of functional requirements, but some of the key ones that influence project success are:

Being top-down and demand-driven, it ensures that the BI project is focused on tangible results and agreed benefits from day one.
It effectively defines the scope of the project. Any activity or deliverable included in the project is only useful in so far as it supports the functional requirements. If it doesn’t support them, why are you doing it?
It provides a touchstone that every part of the BI project can be referred back to.
- Q: Is this a good solution design – A: How well does it support the requirements
- Q: Which source applications do we need to integrate – A: What data is necessary to meet the stated information requirements
- Q: What testing should we do – A: Test that each of the detailed requirements are met
- At the end of the project it is very easy to demonstrate success. A clear and demonstrable connection between requested functionality and delivered product builds up the credibility of the BI team.

In Summary

Business Intelligence is as much about supply and demand as any other area of business. As Business Intelligence professionals we can improve our ‘product’, and the success of our projects, by becoming increasingly demand-driven in our approach to Business Intelligence.

Peter George is a Senior Consultant with Montage - Business Intelligence

Top Ten 2013 BI Tends

Fri, 14 Dec 2012 15:31:24 +1300

Tableau Software release their thoughts on the BI trends for 2013.

Gartner Research: Customers Rate Their BI Vendors on Costs

Wed, 12 Dec 2012 15:31:24 +1300

Gartner Research: Customers Rate Their BI Vendors on Costs

Gartner, one of the world’s most influential Information Technology analysts, is out with new reports about the costs of business intelligence platforms. Read the full report and see how each of 25 BI vendors performs. Gartner compared vendors on cost elements such as average implementation cost, license costs and other product attributes.

Gartner Research: Implementation Costs

Gartner Research: License and Hardware Costs

ETL in Depth with Ralph Kimball and Bob Becker

Wed, 21 Nov 2012 12:30:24 +1300

A thoughtful observer’s point of view: ETL in Depth with Ralph Kimball and Bob Becker

Spending four days on a course with the father of dimensional modelling and a thought leader in Business Intelligence is, as you can imagine, any BI consultant’s dream!

The course started with Ralph’s overview of his career back to his involvement with PARC in the days when the icon and mouse were designed, through to Red Brick and then the Kimball Group. His fascination with how people and computers interact has driven the passion and has always been at the heart of his career.

He summed up the personality requirements for a BI professional as:

Someone who is interested in the business world, how businesses work and discovering business processes.
Someone who is pretty good with technology. Not necessarily a computer scientist or a programmer; but more a thoughtful observer.
Someone who is interested in people. Why people interact with computers, what they need from it and how they prefer to interact with it.

Thinking about his summary of the ideal candidate, it reminded me of one of Montage’s principal methodologies around the BI process and what we aim to achieve with our clients: bringing together people, process and technology to deliver successful BI solutions.

This being an ‘in depth ETL course’, Ralph continued delving into the depths of ETL highlighting the ultimate goal of ETL as “speed and simplicity to the end user”. This was a short and sweet one liner, but says so much about the process. What we want to achieve from a business perspective is ease of use, and fast query retrieval. Thus if the actual process we as ETL professionals have to go through to achieve this is hard, that’s ok. It’s almost “supposed to be that way”. Else there would be no need for the professionals and all users could spend their days and night in Excel to answer their questions. But alas, we soldier on.

I’ve experienced a few scenarios where someone (client and/or project manager) throws a requirement at me as to exactly what they would like in a fact table, and 9 times out of 10 that is due to their reporting requirements. During the course it was great to get the reconfirmation that that ‘fact tables have nothing to do with how a client expects to see the output of data”. Fact tables serve one purpose only, ‘to represent a single measureable event and therefore belong in the back room’.

Having said that Ralph introduced a 5^th type of fact table which filled in a gap where I’ve always kind of experienced one: a Consolidated Fact table which aggregates the underlying Fact and joins Dimensions as required to simplify reporting. Thus you’re able to keep your underlying dimensional model clean (i.e. only properly structured Transaction, Periodic Snapshot and Accumulating Snapshot Fact tables in your star schemas) and stick to the ultimate goal of ETL: “speed and simplicity to the end user” without making the report writers sweat too much. If you’re wondering about the 4^th type of Fact table, the answer is “Factless Fact”.

Another new idea around dimensional design was the concept of a Fact Surrogate Key. Up to this point I’ve always included Surrogate Keys in the Dimension tables only and used a composite key in the Fact table. Ralph introduced the idea of adding in an (auto-increment) Fact Surrogate Key into Fact tables. Some advantages which helped convince me of the usefulness:

Easy identification of a single fact record. Use a single column value instead of a combination of a number of Dimension Surrogate Keys to identify a single fact record. Mostly for back room and testing purposes.
Allows you to easily resume or back out of a failed job.
Using a Fact Surrogate Key you can also now insert and update records in a Fact table without any primary key violations
Use the Fact Surrogate Key to serve as the parent in a parent-child relationship scheme.

After 2 days with Ralph, Bob Becker took over and stepped through Dimension and Fact tables in even more detail. This was the really in depth part and we got into the technicalities without getting too technical, a perfect mix. The course is very detailed without focussing on any specific software vendor. Bob included some examples of different software vendors which was great to see. I mainly use the Microsoft BI stack and WhereScape for ETL, so seeing other vendors’ products was great exposure for a consultant.

One of the most important things I took from the days spent on dimension architecture is the importance of conforming. You can only really have success in your design if you conform on all levels. Thus not only conforming Dimensions, but also Hierarchies, Attributes and Relationships. One version of the truth can only be achieved if that’s what the underlying data portrays.

At the end of the 4 day course I felt inspired, my ETL passion re-ignited and ready to take on the BI world and transform all Data Warehouses into fast and efficient data cleansing machines.

It’s refreshing to realise the Kimball Group comprises only 6 people in the US, yet their influence spans across the globe into so many companies, data models and warehouses. Meeting these people was a truly inspirational experience.

Course details:

ETL Architecture in Depth

Melbourne

17 – 20 September 2012

http://www.kimballgroup.com/data-warehouse-and-business-intelligence-courses/data-warehouse-and-business-intelligence-training/

What Chart or Graph is Right for You?

Wed, 07 Nov 2012 15:31:24 +1300

Tell high-impact stories with data.

Transforming data into an effective visualisation (any kind of chart or graph) is the first step towards making your data work for you. In this paper you’ll find best practice recommendations for when to create these types of visualisations.

Download pdf

Agile Business Intelligence - Aberdeen Group Report

Tue, 23 Oct 2012 15:31:24 +1300

Aberdeen Group compare the performance of organisations based on the type of BI technologies and approach they choose (Traditional BI - VIsual / Interactive BI).

Download pdf

Success Through Partnerships - ANZCO

Tue, 02 Oct 2012 15:31:24 +1300

The challenges faced by ANZCO are typical: they had invested in new core business systems and applications, but their Business Intelligence was still driven by spread sheets. This brought a number of issues around scalability, transparency, data quality, and timeliness - and meant that they lacked a ‘single version of the truth’.