6月 052015

Have you heard of Meskimen’s Law? It states the following: “There’s never time to do it right, but there’s always time to do it over.” If you work in software development you’ve probably come across colleagues who seem too ready to apply this law in the realm of software quality.

Meskimen’s Law is of course meant to be tongue-in-cheek, but sacrificing software quality for more functionality or faster development is no laughing matter when you care about the bottom line. Back in 2002 a study commissioned by the National Institute of Standards and Technology (NIST) in the US found that between 50 and 75 percent of development funds are consumed by software developers identifying and correcting defects, and that software defects cost the U.S. economy almost $60 billion annually.

The upshot is that by focusing on quality a software company can significantly reduce costs and boost profitability. Here are three other reasons why attention to quality makes business sense:

Focusing on software quality promotes business growth

When software is made properly from the outset free from defects, it’s much easier to scale and adapt as business needs change. If a software project is successful and delivers good value to the organisation, there is a better chance the project will require additional features and functionality and be deployed in different areas.

Consider what Executive General Manager for Group Security and Chief security officer John Geurts had to say about Commonwealth Bank’s requirements when investing in software for fraud detection:

“We were […] looking to achieve an economy of scale, reducing data storage costs, enabling reuse across the group. In addition, we needed the flexibility to add new products, services and channels to the platform at a far lower incremental cost than installing another customized fraud detection system.”

Read the full story of how the Commonwealth Bank saw a 95% increase in check fraud detection efficiency.

Software makers who produce quality software thus have greater opportunities to cross sell and to sell into new markets. But if adapting or re-purposing software becomes too expensive or time consuming due to poor original implementation, these opportunities for additional sales are less likely to materialise.

Focusing on software quality promotes customer loyalty.

When a company consistently delivers quality products to its customers, those customers tend to keep returning for more. They also give positive referrals to others. Thanks to the proliferation of social media sites such as Twitter and Facebook, the impact of these positive referrals can be far-reaching. So too can the impact of negative referrals.

Companies that use software for high-stakes activities such as fraud detection or credit risk modelling can potentially incur millions of dollars in damages due to glitches resulting from poor quality software. You can be sure that when something like that happens it doesn’t take long for the rest of the marketplace to find out about it. By focusing on quality software, companies can ensure their customers remain happy and tell their colleagues about it.

Focusing on software quality increases brand equity

The value of a company’s brand is derived in large part from customer experience of its products and services. Brand equity is difficult to measure but it can impact the ability to raise capital, hire top quality employees, and charge a premium. Software quality problems can significantly affect the experiences customers have with a brand and the damage builds up over time. Is your idea of a good time spending the morning on the phone with Tech Support talking about error logs? (No offense to Tech Support teams!) A good solid brand with a reputation for high quality is a powerful driver of business growth.

For more on the Quality Imperative and SAS' commitment to product and service quality and customer satisfaction, download this free technical paper (no registration required).

tags: quality, software

The post 3 reasons why focusing on software quality makes business sense appeared first on Left of the Date Line.

5月 262014

I recently attended an event where a speaker from LinkedIn presented. He mentioned an interesting trend – the demand for analytical professionals has increased in the last five the years. We all know analytics is the hottest field around. Better than mining even - as of last year, the median analyst in Australia was earning over twice the median Australian salary. Although, some evidence suggests that salary growth may be cooling somewhat overseas.

Big data’s hot, then it’s not. Machine learning’s cool, then it’s old school. LASR, Spark, in-memory, and good old data mining; what’s a person to focus on?! Given the rate that everything’s changing out there, have you ever wondered how the skills you choose affect your salary?

Don’t guess, find out. The Institute of Analytics Professionals of Australia is running their annual skills and salary survey for the second time. As with last year, it covers:
• The industries that people work in
• The tools people have used and are using
• The challenges people are facing

New to this year, it also covers:
• The degree to which analytics is being centralised or federated in organisations
• Whether people are interested in formal or informal education on business analytics

With the recent launch of a Masters of Data Science at the University of South Australia, the Masters of Business Analytics at Deakin University, the Masters of Analytics at RMIT and the Masters of Analytics by Research at UTS, there’s clearly a lot of pent-up demand for learning.

Respondents have the option of receiving the report once it’s completed. And, given what was in last year’s report, it’s sure to be interesting reading.

Regardless of what you use or where your focus lies, as long as you’re an analyst and you’re working in Australia make sure to complete it before it closes. Tell your friends, tell your enemies, tell anyone and everyone you know who works in the field. Tweet it, blog it, talk about it over the water cooler - as the most comprehensive survey of its kind in Australia, it represents a unique opportunity to really understand what’s hot, what’s not, what’s trending.

Complete the survey

SAS is an active sponsor of IAPA and supports the broad development of skills across the industry.

tags: analyst
5月 292013

At SAS Global Forum 2013, one of the key announcements was that a new platform SAS release, 9.4 will be available from June 2013.  While 9.4 underpins some of the headlines around high performance analytics and visual analytics and the cloud initiative,  it will be of great interest for SAS users in its own right.

While SAS 9.3 has been our current release for the past two years, 9.4 offers a huge number of enhancements, 
adding lots of new goodies that people will find useful.  SAS 9.4 can be viewed both as an incremental release, as well a platform for exciting and totally new features.

 SAS 9.4 will offer  a straightforward migration path, using the same approaches proven in the migration to 9.2 and 9.3.

Vincent covered the main highlights but I'm going to concentrate on the new SAS foundation features that will benefit SAS analysts and programmers:

  • In the area of analytics, we are seeing some of the new algorithms pioneered in our High Performance Analytics products implemented in an SMP (single server multiple cpu) environment in the “standard” editions of SAS/Stat and other analytic products.
  • New languages such as DS2 to allow code to be submitted from Base SAS sessions to run in-database to perform advanced data manipulation without moving the data out of the database taking advantage of parallelisation.  Furthermore extending implementations of existing languages such as the ANSI 1999 compliant FEDSQL.
  • Support for the latest operating system versions and other third party product versions, such as Microsoft Office 2013.
  • New trigonometric functions such as COT, CSC and SEC.
  • One of the areas that will be really exciting for many people will be the long awaited production availability of  ODS Layout and ODS Report Writing Interface.  For old timers who remember the DATA _NULL_ report writing with PUT statements, these new features allow unlimited flexibility in creating highly customized PDF and HTML documents.  Read this SGF paper for more information.
  • ODS Graphics is now in its third generation and has many new features that give you more flexibility and control.   This combines with the LAYOUT to create publication ready content.
  • The ODS EPUB destination  creates output optimised for eReaders and tablets.
  • In SAS/Access a new pipeline implementation can improve performance especially when data is being streamed from a database into a complex SAS data step.
  • One useful little goodie is the ZIP Filename engine; this allows reading and writing of files inside a zipped archive.

So when 9.4 is released, investigate the “What’s New in 9.4” documentation on the SAS Support site to find out details of these and many other new features.  I’m sure you will find some that will give you real value – and for each person they may be different.  I'll be very interested to hear from you as to what you found valuable so we can share them!

Stay tuned to this blog where I will discuss new features for SAS Administrators in a later post.

Want to find out more about the exciting announcements, you can attend SAS Forum Sydney on August 8th.

 Click here to register

tags: analytics, big data, cloud, data visualisation, high performance analytics, SAS 9.4
3月 252013

I have been on a plane non-stop for four weeks travelling around Asia working with key customers looking to expand their analytics footprint.  It's clear that analytics is no longer taking a back seat to business intelligence (BI).  Customers I am talking to have squeezed as much value out of BI as possible, and while it is still a necessary capability and key building block, analytics is now the frontier companies are leveraging to drive competitive advantage.  A recent article from McKinsey refers to the tangible benefits companies are already seeing “when companies inject data and analytics deep into their operations, they can deliver productivity and profit gains that are 5 to 6 percent higher than those of the competition”.  Question begs though; is this you or your competitor?

Typically the benefits of leveraging analytics have been reserved for companies employing rocket scientists – abstract and nerdy.  However recent innovations in data visualisation, hardware, analytics and mobility are providing this same insight to anyone, anywhere and anytime.  With the introduction of SAS Visual Analytics, SAS is saying that there must be a shift in the mindset around analytics and precisely who it should be meant for. The key premise underlying this change is the belief that “Analytics is for Everyone”. Michelle Holmes provides two great examples on how to practically leverage these innovations from both the everyday manager and a data scientist perspective.

Leading organisations have adopted analytics in the board room and at all levels of the decision making process. By making analytics approachable, companies can drive its adoption across previously unreachable segments of users, hence instilling an analytics culture within the organisation. Analytics has been the backroom engine of some of the best performing global companies, providing insight to not just their key management and analysts, but as a corporate behavior that can positively impact all key business users in the way they gain insights and make better decisions. The results from analytics processes don’t always have to be meant for the few, and be boring and hard to consume.

To make analytics approachable, data visualisation is key. To convey the analysis results effectively, both aesthetic form and functionality need to go hand in hand, providing insights into a rather sparse and complex data set by communicating its key aspects in a more intuitive way. SAS has managed to combine the best in advanced analytics and data visualisation, and coupled these with innovations around in-memory, advanced analytic server design and Hadoop. This combination of high-performance analytics and an easy-to-use exploration interface is important; big data brings a unique set of challenges for creating visualisations. If you are working with large data, one challenge is how to display results of data exploration and analysis in a way that is not overwhelming. You may need a new way to look at the data that collapses and condenses the results in an intuitive fashion but still displays graphs and charts that decision makers are accustomed to seeing. You may also need to make the results available quickly via mobile devices, and provide users with the ability to easily explore data on their own in real time. Here are a few examples of how companies and departments are leveraging the marriage of advanced analytics and data visualisation:

  • SM Marketing Convergence Inc. (SM-MCI), an affiliate of SM Retail Group delivers customer insight with pure speed. The ability to analyse in seconds, billions of rows of shopping cart and loyalty data from multiple channels is key to making timely decisions on campaigns and products. Using data visualisation and advanced analytics SM-MCI now have a more in-depth understanding of customer buying patterns, behaviour and trends, and can use this insight to offer more relevant promotions, acquire new members and identify profitable up-sell opportunities.
  • Australian Institute of Health Welfare (AIHW) and leading government agencies of today are evolving to focus on outcome-based-budgeting and evidence-based-policy making, improving the transparency and effectiveness of decision making.  AIHW have moved away from a traditional focus on historical reporting to supporting agencies with the ability to explore and understand the data helping them to apply the right analyses. Leveraging SAS Visual Analytics AIHW shares insights with other agencies ensuring they have the data they need, when they need it, and helping them draw insights from that data with visualisations.  Analysts now get the information they need in real-time, rather than make them wait 18 months for a report.  The marriage of advanced analytics and data visualisation are key to implementing and succeeding with these programs.
  • Cosmos Bank provides information and insight in real-time to key executives on-the-go.  Executives and decision makers have easy and immediate access to explore dashboards and reports from their mobile devices in the board room and on the road, anytime and anywhere. Cosmos Bank executive management has access to up-to-the-minute risk intelligence and customer intelligence for decisions that keep the bank ahead of its competitors.
  • Hong Kong Efficiency Unit (HKEU) obtains smarter insights from analysing data from 2.65 million calls, 98,000 e-mails and 300,000 complaints from citizens about infrastructure, health, defence and public services.  By applying advanced analytics, text analytics and data visualisation, the department can quickly identify key patterns and relationships in the data. HKEU make better decisions and has developed smarter strategies that improve delivery of services to the public. Click here to read more

If a picture is worth a thousand words, stop writing and check out some of the ways data visualisation can help all your employees make faster smarter decisions.

Learn how to tell your own story - try it for yourself for free right here.

tags: Asia, big data, business intelligence, data mining, data visualisation, ROI, skills, visual analytics, visualisation
7月 232012

Bringing analytical insight to the point of decision
Organisations are moving their focus from being product-centric to customer-centric. I recently spoke to the CIO of a leading Asian insurance company who discussed their CEO’s new mantra … the desire to focus on doing business at the speed of life.

He went on to elaborate:

“Consumers today demand instant satisfaction; regardless of the occasion, whether it’s online purchases, money transfers, package tracking, social communication, and even over-the-counter purchases. Organisations have it tough trying to convey a relevant message in a timely fashion, especially considering the move towards acquiring knowledge using the web and social media, as well as the proliferation of marketing messages everywhere you look, flip and click. Consumers now expect a personalised, relevant and contextual customer experience. Organisations face a decreasing window of time to detect threats and take advantage of opportunity. The challenge facing us is how to combine ten years of ERP data, 20 years of customer history and combining that with the current customer interaction data. Add to this that some of that data is unstructured, and it quickly becomes apparent that our information technology environment and our enterprise data warehouse simply was not designed for this real-time contextual focus.”

This reality is where many organisations find themselves today. Industry leaders see big data as an opportunity and high performance analytics (HPA) as the enabler to gaining competitive advantage. The focus is how to bring analytical insight to the point of decision, whether that be the sales team, customer service representatives, ATMs, websites or other operational applications.

The recipe to great customer experience
Organisations that are successfully delivering great customer experience through real-time insights are consistently:

  • Accurate. Poor insight, sped up, means angry customers faster. Increasing accuracy of insight requires analysing existing structured data (source of truth) and unstructured data together. It means looking at all of the data not just sampling the data.
  • Event-driven. Doing business at the speed of life means we need to understand, predict and listen for the life stage triggers. Knowledge of when and why a customer is interacting with you is key to improving your cut through. The opportunity is not to make an offer, it’s to make the right offer in real-time. We have all received a credit card offer in the mail. Somewhere there was an event that was captured and the offer was relevant at that point in time. Three weeks later it’s spam.
  • Empowering the channel. Operationalisation and delivery of insight to the interaction points with your customers is vital to success. Some have labelled this the consumerisation or democratisation of analytics. It could be as simple as a pop-up window on your website, a dynamic script for your call centre representatives or a salesman looking at location based reports on a tablet. Placing insight in the hands of those who make decisions about your business every day is the real-time analytics pay off.

HSBC improving customer experience through analytical innovation
Traditionally credit card fraud is detected by sampling some transactions post the event and looking for fraudulent patterns. Then kicking off a process to investigate, notify and fix the fraud. HSBC moved to real-time fraud prevention, differentiating their credit card product by improving security, lowering fraud and improving the customer experience. Think about the customer experience if they get it wrong. The customer is paying for shopping or expensing a corporate dinner. The stakes can get no higher, a negative experience here would likely end in a customer churn. HSBC analyse hundreds of millions of credit card swipes every day ensuring their customers have a higher confidence of security, all delivered using all of the data all of the time without being obtrusive.

HPA delivers insight at the speed of life
HPA enables organisations to drive real-time competitive advantage in three key ways

  1. Ability to look at all the data – Improved accuracy requires the ability to aggregate structured data from your enterprise data warehouse and unstructured data from inside the firewall (email and voice) and outside sources (social). HPA embraces technologies like Hadoop to bring unified insight.
  2. Operationalisation of insight – Delivering insight faster is crucial for organisations to take advantage of that ever shrinking window of opportunity. Shortening the decision cycle requires the ability to expose the insight as decision services that can be embedded in applications like your CRM system, online store and website. HPA manages the end-to-end data-to-value process allowing the velocity to increase exponentially. This is achieved by capturing, transforming, and storing data, and performing advanced analytics over in-memory storage to provide pure speed while managing larger datasets.
  3. Mobility – It is paramount that the insights can be placed into the hands of those interacting with customers. As such the ability to support mobile delivery by web and native applications means the right insight gets to the right person at the right time and in the right location.

Real-time analytics is the enabler to a customer-centric strategy.
So where do you start? Look for areas where you can bring structure to a key decision making process, that positively impacts the customer. The challenge is how to do this profitably, and that’s where analytics comes in. Here are some examples to kick start your thinking:

  • Offering real-time personlised marketing offers to a segment of one at the point of sale. Include detail on location and type of transaction to bring more context and relevance to the offer.
  • Predictive maintenance of assets affecting customer experience. Analysing machine data looking for the beginning of historical fault patterns E.g. electricity grids, telecommunication towers, ATMs, production lines
  • Individual risk based pricing in insurance. Highly personalised policies can help differentiate you in the market while looking at greater profitability in the long term.
  • Disruption management in transport and logistics. The ability to understand the ripple effect of an error and re-coordinate the different assets and resources to achieve the best possible outcome for the customer and your bottom line. Oh and doing it in real–time.
tags: analytics, Asia, Australia, big data, business analytics, business intelligence, challenges, cio, customer intelligence, data integration, high performance analytics, New Zealand, operationalisation, scalability, Singapore
6月 252012

Four ideas for customer delight by elevating customer intelligenceFor the past four days I have been knee deep in advanced business analytics; in particular its application to customer intelligence, at the Business Analytics course at Macquarie University Graduate School of Management.  Camille blogged about the need to embrace customer intelligence in order to address the increasing demands of the intelligent customer.  Discussions with many regional organisations have made it apparent that only the leading organisations are leveraging customer intelligence to meet customer expectation.  With this in mind I found the last day of the course very helpful in providing practical ways to kick off the business analytics journey in an attempt to address this expectation, and to do it profitably. I scanned the market analysts and trusted media to understand the four key trends and have provided some helpful hints to jump-start your projects and improve your customers' delight.

  1. Big customer data.
    Every minute you are not using Big Data means a lost buying opportunity…  Big Data is not going away, it's just getting bigger.” Joe Cordo, Chief Marketing Officer Extraprise.
    Wade through your big customer data and look to identify key influencers. Instead of applying churn and propensity models to view a customer as an isolated individual, look at how an individual can influence friends and family.   Start by capturing more data about customers and building network views by analysing transaction histories to identify consumer networks.  If you’re a telecommunications company look at call data records networking them using numbers; banks can look at transfer records networking them using bank accounts; and retailers can look at basket analysis networking by address details from loyalty data.
  2. Real-time personalised customer experience.
    "From a business perspective, it's about understanding a customer so that communications are relevant and timely," Karen Ganschow Customer Relationship Marketing and Digital Chief, Westpac.
    Empower your call centre representatives to make trigger-based offers.  Start by looking at the top three next best products after analysing your segmentation, product acceptance propensity and use the current event to determine the best offer.
  3. Leveraging social.
    The fact that 75% of brands and agencies have executed social marketing programs over the last year shows marketers are taking social very seriously~  We are seeing tremendous progress of social marketing efforts in the region…" Ken Mandel, Managing Director of Buddy Media Asia Pacific.
    Look at understanding brand sentiment to improve the way you go to market.  Start by analysing unstructured data, determine the sentiment of content, categorising it according to your products and services, and then allocating work to be actioned.  Listen to how Nedbank are using social monitoring of Facebook, online and Twitter to improve sales and take business from competitors.
  4. Right information, at the right time, in the right location ... and at your convenience.
    “33 percent of business intelligence functionality will be consumed via handheld devices by 2013” Gartner 2011.
    Let’s not forget your internal customers. The need to make informed decisions in the moment means that managers and leaders need information on the go.  Look to implement and disseminate key reports on product and regional sales performance reports.  Key things to remember are to; keep the format easily readable on smart phones and tablets, easy to navigate interface and focus only on the key metrics relating to sales performance, like actual v target using bright colors to indicate problem areas.  You can download the free SAS Business Intelligence app here.

All journeys start with a single step.  The hype around some trends can make them feel out of reach.  It's clear that leading companies are not twiddling their thumbs waiting, they are taking a measured approach early.  The advancements in analytics make taking the leap less risky and easily accountable.  Start your journey today.  Stay tuned on our blog to learn how organisations in our region are leading the way and the best practice tips for success.  If you are after a more formal education check out the Business Analytics course at Macquarie University Graduate School of Management.

tags: analytics, big data, business analytics, business intelligence, customer intelligence, Gartner, HPA, mobile, Mobile BI, sas, social media, twitter
5月 242012

In Left of the Dateline we have been discussing topics ranging from the evolving role of the CIO, to the skills and competencies that organisations need to be successful in utilizing business analytics to compete with their rivals.

I was recently at SAS Global Forum in Orlando FL, listening to Thornton May. He mentioned something that really hit the mark when it came to decision making. He mentioned something called an OODA Loop. I had never heard this term before, so off I went and did some research. The OODA loop is a concept that was developed by Colonel John Boyd of the US Air Force and is really a cycle of events that surround the decision making process in military dog fighting

The OODA Loop
The OODA Loop

Colonel Boyd suggested that those who could get inside the rival's decision cycle can gain the advantage. I for one agree with the good Colonel, so here's my take.

O - Observe

Decisions are made on observations. Of course today observations are many and varied, they come from internal and external sources and need to be managed consistently and constantly. Boyd suggests the observation includes unfolding circumstances, outside information and the interaction with the environment - sounds the same in business data. Big data plays heavily into this part, with plenty written on this, so let me leave this challenge. Can YOU manage the data you need to in order to OBSERVE?

O - Orient

When you read what has been said about Orientation you see that it's the part of the decision process where organisations apply their lense to what is observed. What does your corporate lense look like? Is it clean? Scratched or damaged? Is the culture of the organisation or defined norms bending the truth on what you observe. Are you ready to let the data you have available to you orient you for success? It's time to let analytics shape your view and today you can with High Performance Analytics.

D - Decide

When you have to make any business decision, it could be as simple as asking "what’s the chance my customer will leave me". How would you know this? How do you measure the propensity to churn? I would advocate using analytics of a predictive nature to do this, otherwise you simply won't be deciding, you will be guessing and you will be looking backwards rather than forwards. The decision process needs to deliver an action (we will get to that in a moment). At every stage of the OODA Loop you will see that we are looking to feed back into the observation stage, we look to re-orient and then make additional decisions. Deciding more than once on anything is an iterative process. A process that is supported by analytics and governed by a robust information management and model management framework.

One simple piece of  advice I would offer is to decide to decide. I have seen so many people keep waiting for the right time, the perfect data or the perfect model. Let me assure you it never comes along, decide to start and start to decide before your rivals make you irrelevant.

A - Act

Now you have a decision, ACT on it. As Boyd says this will produce an unfolding interaction with the environment, take that and add it to your feedback loop, add it to your "observations" and cycle on through.

The OODA Loop
The OODA Loop

When military thinking is applied to the decision making process, especially when combined with highly capable people and robust technology, you get sustainable competitive advantage.

So what's holding your organisation back from better decision making?


tags: analyst, analytics, Asia, Australia, big data, business analytics, cio, high performance analytics, sas, Thornton May
5月 222012

img source

Let's look at some examples of how analytics is helping enterprises in India become high performance organisations:

Power and utilities: India relies heavily on power and and the distribution of power is highly regulated. Regulators specify that the utilities are required to plan, estimate and report their daily demand to the load dispatch centres, which in turn schedule the available power to them for the next day's consumption. Any shortfall or excess in estimating demand incurs either penalties and losses due to wastage. Today, analytics for load forecasting is helping leading power and utilities companies forecast short-term and long-term power demand with accurate results, thereby minimising losses and penalties.

Stock exchanges and regulatory bodies: The magnitude of daily transactions on exchanges brings with it surveillance challenges. On one hand there is the challenge of analysing the behaviour of investors and scrips over a period of time looking to uncover fraud, including insider trading and circular trading. On the other hand the key role of regulatory bodies is to protect the interests of the investors and citizens. New regulations, increased scrutiny and scandals have all increased the need for sophisticated analysis and monitoring for malpractices. Analytics is helping exchanges and regulators in India to identify relationships between investors, fradulent behaviour, manipulation patterns and new unseen patterns for investigators to identify and detect fraud earlier.

Automotive manufacturing: India has a large base of local auto makers as well as foreign players. Add to that a huge customer base with varied preferences and a mindset largely focused on pricing, and you gain an understanding of the challenge to gain profitable market share.  Analytics is helping leading players in India to determine accurate demand forecasting of various vehicle models, warranty analysis, campaign management and social media marketing. The high performance organisation has a single customer view, enabling them to run targeted marketing campaigns, to optimise spend and improve sales.

Government and Public Sector Undertakings (PSUs): For a country of over a billion customers (citizens), the Indian government can be viewed as an organisation that needs to cater to each of them in various ways. This duty carries the many complexities of multiple regions, central policies, state policies, a large number of ministries and departments; and governance in such a scenario becomes an almost herculean task.  Analytics is playing a pivotal role in the Indian government, enabling the policy makers to implement citizen welfare policies in areas of healthcare, inland security, defence and income tax to excise, customs and census.

Finance Industry: The banking industry in India has a long history, from the traditional banking practices to the reforms period, nationalisation to privatisation of banks, scheduled and cooperatives to foreign multi national banks. The opportunities and challenges run hand in hand. The quest for acquiring more customers through providing various services to the issues such as risk, fraud and economic uncertainties. Analytics is helping Indian banks and insurance companies in the areas of integrated risk management, identification of customer characteristics that lead to deliquencies, cross-selling to existing customers in areas like rate making, claims fraud and campaign management.

Retail: The Indian retail sector is amongst the largest in the world and is experiencing rapid growth to address swelling market demand in an ever-changing customer driven world. Growth in the sector has created a highly competitive environment with foreign players opening shops. In such a market place, retailers are under constant competitive pressure for customer wallet share, improving customer experience and loyalty, meeting growing customer aspirations, increasing its breadth of merchandise, and expanding store operations in to new markets.  All the while needing to maintain profitability.  Analytics is helping leading retailers in India with insightful analysis of merchandise, assortment and inventory management, loyalty, campaign management, shelf-space optimisation, including real time knowledge about sales, and store performance.

Telecommunications: The telecommunications industry in India has a big market potential and is a fast growing sector. India has the world's second largest number of mobile subscribers, with over 900 million as of beginning this year. Many challenges are facing this sector, like ongoing price wars and high government taxes.  Hyper-competition is also bringing in other challenges like eroding revenue and decreasing profitability. Communication providers are now focusing on data and content services to increase revenue and profitability. Analytics has been key in helping these organisations to effectively address issues from churn, cross-sell/up-sell, price plan optimisation and network optimisation,  to areas like identifying potential high value users, campaign management and nurturing profitable relationships.

Data is key to decision making. Big data has placed pressure on Indian organisations' ability to manage the conversion of data into insight to make better decisions.  Welcome to the era of competing on analytics.

QUESTION: What could analytics do for your industry?

tags: analytics, Asia, big data, high performance analytics, HPA, India
5月 082012

After debunking the myth that big data is just for the big end of town I set out on the road to listen to what is happening locally.  For the past two weeks I have met with over 50 Chief Information Officers (CIOs) around Australia and Asia discussing their 2012/13 priorities.  I thought it would be useful to hear about the specific goals and challenges facing them as they move from a mindset of keeping the lights on to that of a strategic seat at the boardroom table. It was very refreshing to see that the CIO was working hard to align IT capabilities to business goals.  A quote from Gartner has certainly sparked some action in CIOs.  The quote discusses the moving trend in Chief Marketing Officers to spend more on IT - predicted to surpass CIO spending by 2017.

Their line of business peers are asking for an increase in the trustworthiness of insight, increased accuracy of data, insight delivered in near real-time, and critically, delivered when the customer is interacting with company.  The big data hype is applying even more pressure to costs but also asking questions as to whether CIOs have the new capabilities to derive value in a consumer-educated world.

Some examples of what the specific lines of business were trying to achieve are:

  • Marketing arm of an insurance company looking to make offers based upon understanding of static history and then using context of current situation and interaction to make a more relevant offer.
  • Risk officer of a bank looking to prevent fraud in real-time to reduce costs in detection and investigation.
  • Chief financial officer of a gaming company asking how to deliver a more personalised experience to punters based on history and current playing habits.
  • Head of marketing for a telecommunications company looking to obtain a better understanding of consumer needs by analysing social and online data and then combining that with their existing CRM and transactional data.
  • Chief operating officer of a transport and logistics company looking to improve the way it reschedules resources, freight and customer expectation based on unforeseen events like the tsunami in Japan.

Here are the top five challenges facing CIOs in trying to deliver to these business goals:

  • Top of the list is data governance.  Specifically the need to automate the way data is martialled and transformed from data entry through to insight and action.  As one CIO put it, “Data is cheap.  Mining the data is expensive and timely.  We need to optimise the data supply chain”
  • Secondly, data quality is back on the table.  A government CIO remarked, “a move towards evidence based policy means data must be trusted and reliable, thus bringing into scrutiny the quality of the data”.  With all the different applications and citizen or customer touch points how do we ensure quality?
  • Close third is an inability to meet performance requirements of the business with the existing platform approach.  Interestingly the problem was not just in shortening a one-off time-to-delivery but in making sure insight could be delivered regularly in shorter intervals.
  • Fourth is an old chestnut - single customer view, asset, product, vendor or employee.  The difficulty is that the customer is strewn across different lines of business with differing details.  A retail bank CIO gave an example of why it is important. “We have a customer; Maryanne Smith for a credit card, Mary Smith for personal loan, and Joe and Mary Smith for home insurance.  Currently the bank are marketing to approximately 20 million individuals when it’s clear we only have around 5 million unique customers. There is a lot of needless cost and effort spent on irrelevant marketing offers with low response rates. Haven’t we all experienced that? So what does that mean to customer experience and churn?”
  • Fifth is the inability to manage and harness value from unstructured data.  While some had experimented with Hadoop none of them had successfully implemented value.

If this sounds like you, then take comfort in knowing there are options out there.  It was clear that making better decisions relied upon increasing the ability to deliver more timely, reliable, trusted and accurate data.  While we have been recently discussing the power of high performance analytics it is clear that data governance, data quality, master data management and data integration are seen as the key to unlocking sustainable value from business analytics.

We've got more to come as we go explore examples of how local companies are addressing these issues to drive value from big data.  In the meantime let us know how your how your data governance initiatives have delivered value.

tags: Asia, Australia, big data, cio, high performance analytics, New Zealand
3月 232012 that the home renovations have slowed somewhat, I've had a bit more time to think about things that don't involve moving bricks, arguing with tilers, or dealing with the seemingly endless rain we've had this summer. One of those things is high performance analytics, partly because it's interesting but also because I'm still working out how much it's going to change how we view analytics. That sounds somewhat hyperbolic, but I also fundamentally believe in the truth of it - ready or not, I think we're at the start of an inflection point.

I don't say that lightly. I'm painfully aware that information technology thrives on the latest and greatest; arguably more than any other, the industry leaps from buzzword to buzzword, rebranding the old as new. Analysts love it, experienced curmudgeons hate it. Cloud computing draws a lot of parallel concepts from the old and heady days of mainframe virtualisation. Web services were the latest iteration of loosely coupled messaging frameworks. And, let's be honest - even with all the recent attention to analytics, it's not like Defence hasn't been applying analytics to aid with logistics optimisation since the 50's.

Of course, there's a catch. It's dangerous to take things at face value and assume that simply because there's a new name, we've already been there and that the latest iteration is just more of the same. Look at analytics versus business analytics - the difference is just one word. That solitary word, however, cuts across so many financial implications and cultural considerations that it's like night and day. It's why some are successful and some firmly believe it's just another point on the hype cycle. Words are important, and we ignore them at our own peril.

High performance analytics is like that.

Supercomputing and massively parallel systems aren't new to me -  while I was working at General Motors, we had our own personal supercomputer on-tap. I used to enjoy heading into the server room just to check out the blades; for a guy brought up on an Apple II and a PC-compatible 286, the whole concept of a room stuffed to the gills with racks and racks of parallelised servers was just ... cool.

It was also frustrating. We mainly used the platform to help model computational fluid dynamics, simulating airflow patterns over wireframe meshes to work out the aerodynamic efficiency of vehicle prototypes. Very cool, but not what I personally did - my focus was on modelling consumer purchasing behavioural characteristics and long-term migratory patterns to identify and quantify emerging markets. And, for that, the platform was useless.

None of the software I had would actually work! It was so specialised that despite having massive potential, it might as well have been my old 286. Imagine this for a second: in less than a minute's walk from my desk, I could wonder at a massively parallel computing platform. And yet, I had to do all my modelling on a (admittedly still impressive for the time) quad-core server with a basic RAID setup. I had the most powerful computing resources I'd ever seen sitting there right in front of me and yet I couldn't do a thing with it.

Ironic, right?

The problem was that while it was spectacularly powerful, it was also pretty much single-use. If I wanted to do something on it, I had to roll my own code in C or Java; the platform didn't actually have any real abstraction. It made it kind of hard to justify a discovery project where the ramp-up just to get some usable software was a minimum of six months worth of software development ...

Which, in a round-about way, brings us back to our announcement about high performance analytics. The thing that fascinates me is the general-purpose potential of it - I'm still coming to terms with the types of problems that can be solved now. It's a very real case of breaking down preconceptions about what's possible. Stuff like interactive market-basket analysis with no limitations on sales history, transactional granularity, or geographic divisions. The probable (but not guaranteed) eventual death of OLAP as general-purpose interactive in-memory capabilities become increasingly available. The freedom to not worry about memory allocations, lack of computing resources, or algorithmic limitations. And there's more, lots more.

High performance analytics is going to change the way we work - take the time to have a look around and re-examine what you take for granted. It's not about the technology; it's about delivering better outcomes. Whether it's being able to do things faster, solve harder problems, or make decisions in real-time, high performance analytics is going to affect everyone, not just those with big data.

It's an exciting time. If you want to read more, check out Alison's series on high performance analytics or have a look on Twitter using the hashtags #SASHPA and #SASANZ. Or have a look at our new microsite. It's worth it.

tags: business analytics, high performance analytics, scalability, supercomputer