What is Cluster Analysis? And Why Use It?

2015-09-14 15_01_17-K-Means Clustering - Store. MicroStrategy 9

Why would you want to use cluster analysis on your retail sales data? Well, cluster analysis helps you identify non-independence in your data. Here is an example to help illustrate the point. Lets say we want to ask loads of teachers from many different schools what they think of their principal. If you ask two different teachers from two different schools, you will get two completely different answers that will be independent. But, if you ask two teachers from the same school, the answers will not be completely independent and could be very similar – but not EXACTLY the same.  Now if your job was to take the raw data and try to predict which school each teacher came from based on their answer – then you have an application of clustering.

2015-09-15 09_53_10-K-Means Clustering - Store - Original. MicroStrategy 9

The same thing can be applied to Walmart store performance for a supplier. You have some data points for a store like how long that store has been open, how many competitors it has located in its vicinity, what was your products sales performance for that store, some demographics for that area like unemployment and population, possibly even some historic weather data. Now you use a clustering algorithm to group your stores that are most closely related. This could be the first step in identifying under performing stores and why. It could give you a viable store list for a product test based on more than sales performance. It might help you further identify your product identity and who your actual customers are using enough demographic data. You might not find anything you didn’t already know. The important thing is that you are diving into your data to truly understand it on a level you never have before, and uncovering one of these nuggets could be millions of dollars difference to your company.

Once you’ve built your base analysis, and in our case we built our report that you see above, turned it into an in-memory cube, and then built a MicroStrategy dashboard on top of it – we can then explore slicing and dicing our data along the different data points to help identify if any of the metrics in our analysis are a key contributor to a cluster alignment. This way we can determine what factor affects sales the most. Could it be store age? or store square footage? or unemployment? Ethnic breakdown? What of these are driving markdowns?

The great thing about using this analysis as a MicroStrategy dashboard is that it is pretty easy to tweak to look for your top performing stores, and refreshing the data source is very easy. In fact, this report could be automated each week and emailed to you. There might even be an application to look for cluster changes and have something like that generate an alert so you only need to be bothered if anything changes.

Contact us today to discover how Vortisieze analytics can help you explore your own data science.

post

New #BusinessIntelligence skills in, BI developers out

Consultant Lee Feinberg says organizations need to build up their data storytelling and visualization capabilities — and details a competition for finding people with the right business intelligence skills.

You’re investing a lot of money in business intelligence tools and applications to keep up with your organization’s changing business needs — perhaps hundreds of thousands of dollars, if not more. But are you investing enough in the people who you expect to use those tools — or in the right kind of people to begin with? Probably not.
Read more . . .

Contact us today for your complimentary BI consultation.

Source: New business intelligence skills in, BI developers out

#BusinessIntelligence
#CategoryManagers
#CPG
#CPGMarketing

post

Vortisieze Reduces Stress Related Deaths at #CPG Companies: #CategoryManagers Take Note

Well – ok – that’s a “wee bit of a stretch,” as my Irish grandfather used to say – but what is reported about long hours and early death is not a stretch – and it’s serious.

Two Yahoo articles published today show a direct coorelation between working long hours on the job and increased risk of stress- related early deaths from stroke, heart attacks and suicide.

In Japan, death by over work, or karoshi, is a legally recognized cause of death.

While the demands of CPG category managers and sales managers grow, there is pressure to keep staff levels at predetermined levels, sometimes without regard to the amount of work to be done.

This presents real headaches (and worse) to CPG vendors.  However, one possible solution is outsourcing some of the routine and mundane aspects of gaining insights – building reports and dashboards.

As a category or sales manager for a CPG company you are paid to gain insights from what is happening in your retailer environment.  But do you really need to know the nitty-gritty of building dashboards or reports?

Probably not.  At Vortisieze our founders have over 25 years combined experience in building analytics in the CPG category management arena.  Even if you aren’t ready for big data we can take some of the grunt work off of your desk.  We understand BI – included traditional DSR data warehouses – and especially the analytics engines.  MicroStrategy is our primary expertise but we know other tools as well.

Contact us today to discover how we can lesson your workload so you can focus on what is important – growing your brand.

 

Sources:

The 100 hour work week in Japan

Working longer hours increases stroke risk by up to 33%: study

 

#CPG

#CategoryManagers

#BusinessIntelligence

#CPGMarketing

 

post

ZuumSocial Releases Facebook Leaderboard – Most Engaged #CPG Brands on FB

Last month Zuum announced it will begin monthly ranking of the 25 most engaging brands on Facebook.  The rankings are limited to the top brands operating in the U.S.  This is how they summarized their criteria for selection:

The goal of this is to capture the Facebook content and community traits of the top brands operating in the US. So we’ve put several constraints on which brands are selected. Brands must be operating in the US, we’ll use the US page unless there’s only a global page, brands must be consumer products vs B2B, and we’re excluding technology and entertainment brands, as our goal is to better understand how typical consumer products work on social media, and both of those categories tend to skew towards extremely high engagement due to the product type.

For July and August Monster Energy Drinks captured the top.  This month Zuum has released the rankings for a specific category CPG Dips and Dressings.  Here is the lead-in to their rankings:

CPG is a very broad business category, with sub-categories like soft drinks having brands with some of the larger social media fan counts anywhere. In our July ranking of the 25 most engaging brands on Facebook, the three top brands are CPG.

Of course, the CPG umbrella includes many smaller brands, with more niche or regional appeal. One subcategory in particular is dips and dressings. Below is the Facebook Leaderboard for some of the top dips and spreads brands in the US for the month of July.

You can see that the fan counts, while not what you’ll see from Pepsi or Coke, are still substantial. Posting volume is a little light in this category, with even the top brand, Sabra, only posting about 1 time per day.

Once the novelty of these new rankings are gone what value does this have to a CPG category manager or sales analyst?  “Well – Nancy . . .” – if your analytics cannot lay this new, unstructured data alongside your retailer supplied POS and inventory data – there is no value outside of water cooler talk.

This is one of the major limitations of rigid, outdated, DSRs built using data warehouse technology.  Rigid in the sense that to add new data feeds takes months – sometimes years.  The loading of the data – including current feeds – is long and cumbersome.  From a technological point-of-view these simply will break under the weight of increasing amount and type of data.

If you want to use these new – and exciting – data points (social media, weather, and et. al.) to draw insights and correlations to sales – then the only technology that can pull this off – today – is Big Data.

That’s what the buzz is about.  Will your brand be buzzing going forward?

We are Vortisieze.

Contact us today for a complimentary consultation on your BI strategy.

 

Sources:

Facebook July Leaderboard for CPG Dips and Dressings

http://blogs.imediaconnection.com/blog/2015/08/17/facebook-july-leaderboard-for-cpg-dips-and-dressings/

 

Ranking 25 of the most engaging brands on Facebook

http://zuumsocial.com/ranking-25-of-the-most-engaging-brands-on-facebook/

 

July Ranking of 25 Most Engaging Brands on Facebook

http://zuumsocial.com/july-ranking-of-25-most-engaging-brands-on-facebook/

 

#CPG

#Business Intelligence

#BigData

#AnalyticsInRetail

post

10 Steps for #CloudComputing #BusinessIntelligence Success

  1. Study Pricing and Total Cost of Ownership (TCO)
    Total cost of ownershipMany cloud analytic solutions offer per-user licensing on a monthly, quarterly or annual subscription basis. Be sure to work through the model that best fits your firm. Take into account the overall length of the project and expected growth to ensure the total cost of the project over time meets your goals, EMA asserts.
  2. Architecture and Features
    2. Architecture and FeaturesIn addition to studying current features, examine the release schedule and roadmap for your provider to better understand when new releases will be available. Some cloud analytic firms have a monthly release cadence. Others are quarterly. Depending on the architecture of your provider, you may be able to align these releases with your optimal IT schedule and business flow. Customization can be difficult in many cloud environments; be sure to investigate your specific needs and determine costs in advance. Some environments that are multi-tenant (many customers sharing a single architecture) can be restrictive with regards to custom features, EMA says.
  3. Service Management and Service Level Agreements (SLAs)
    3. Service Management and Service Level Agreements (SLAs)Determine how the provider supports new user provisioning, license management, system load balancing, and elasticity, as these are all critical components of a high performing system. Find out if the provider has assigned technical support and real-time chat support for customers, EMA recommends.
  4. Transparency and Communication
    4. Transparency and CommunicationCheck to see if the platform offers a real-time performance portal so customers are able to see system performance, uptime, outage scheduling, support issues, and scheduled maintenance, EMA suggests.
  5. Application Programming Interface (APIs)
    Robot with application programming interface sign. Technology concept. Isolated on white backgroundAPIs can ensure a cloud analytic environment doesn’t become an isolated silo. Additionally, some vendors will create an extensible environment by publishing and supporting Software Developer Kits (SDKs). If flexibility and connectivity are critical to your deployment, look for vendors who excel in this area, EMA asserts.
  6. Platform Security
    Safe cloud computing. Digitally Generated Image isolated on white backgroundLook for vendors that take the security challenge seriously. Leaders in the market are investing in their infrastructure and policies to deliver a high level of performance, EMA says.
  7. Proof of Concept (POCs)
    network_simple_commnunication12Deploy a POC before committing to a long-term cloud analytic partnership. POCs reduce risk and help ensure a smooth transition to the right vendor solution, EMA notes.
  8. Training
    8. TrainingMany SaaS solutions offer intuitive user interfaces, but training remains important. Research whether the vendor supports a community of users that you can leverage and learn from. Visit the forums and FAQ resources, research user groups, onsite seminars, and conferences designed to support you and your team post-purchase, EMA recommends.
  9. Vendor Strength
    9. Vendor StrengthResearch a company’s revenue and financial status (if the information is publicly available). For newer options, check the funding sources and venture capital positions, executive team, partner networks, and references, EMA says.
  10. Exit or Relocation
    Emergency exit sign glowing in the dark

    Emergency exit sign glowing in the dark

    Before committing to a vendor, understand the process for migrating to on-premises deployments. Some vendors are capable of serving their solution both ways, and at some point your circumstances may require moving the solution back behind your firewall. Be sure there is no impractical data transfer fees or penalties in the contract that make this process difficult, EMA concludes.

Source:  10 Steps for Cloud Business Intelligence Success

#BusinessIntelligence
#CloudComputing

post

How GoPro Is Using #BigDataAnalytics in The #CloudComputing to Kick Everyone Else’s Butt

We run around talking about how important analytics is and yet there are few really compelling examples of how well it is working. Part of this is because the vast majority of implementations are still in process and haven’t gotten to value yet, part is because they were done wrong and value wasn’t found, and part because firms don’t like sharing with competitors how they are kicking those competitor’s butts.

GoPro, however, is the perfect example of how analytics are being used competitively to out-execute much larger companies like Sony.

So starts the article on TechSpective.net published on August 3, 2015.  This article is a must read on how to use big data and analytics to out maneuver your competitor.  The challenge is adapting the technology to brick and mortar retail.

Interestingly GoPro implemented the same strategy that we at Vortisieze execute every day.  Data in the cloud, strong big data technology and a top-notch analytics engine.  Like GoPro, Vortisieze partners with Cloudera.  To round it out though, Vortisieze maintains its own data cloud and leverages MicroStrategy (et al.) for the analytics engine.

Contact us today for your complimentary BI consultation.

 

Source: How GoPro is using Amazon, BMC, and Cloudera to kick everyone else’s butt

#Cloudera
#BigDataAnalytics
#BusinessIntelligence
#DataAnalyticsTechnology
#DataAndAnalytics

post

#BigData and the 2016 Presidential Election #PredictiveAnalytics

What Nate Silver did for President Obama in the 2008 election cycle Deep Root Analytics, a media analytics company formed in response to the 2012 Republican loss in the presidential election, is working to do for several Republican candidates in this election cycle.

Deep Root Analytics partners with data-blending and advanced-analytics company Alteryx to merge voter file information, set-top box data and commercial data to optimize audience targeting and TV ad-space buying.

While this is very interesting to observe ultimately we in the retail space must learn from the lessons that these, and other big data exercises, provide.  As consumers move from one retail channel to another (actually many) channel using all the data available is paramount.  No longer will solely relying on the POS/Inventory data your retailer provides you be enough.

Wondering how to make this all come together for you?  Contact us today for a complimentary consultation.  And enjoy this election as you watch others use big data to move their candidate (brand) forward in the voter’s (consumer) eye.

 

Source:  How Data and Programmatic TV Will Dominate the 2016 Presidential Campaign

#DataForAnalytics

#DataAnalyticsTechnology

#PredictiveAnalytics

post

12 Major Mistake Companies Make With #BigData

HuffPost is not normally the go-to for business information but this is great online piece that asks 12 members of Young Entrepreneur Council (YEC) their opinion on the top mistake companies make with big data.  I’ll list a few below and the link to the article is at the bottom.

CPG companies working with retailer supplied data should make of note of this.

 

Not Making Data-Driven Decisions

Data equals power.

“I believe that most companies don’t realize how much you can pull out of your data. There are many tools out there that can help you make data-driven decisions, which in turn can give you more predictable results.” – Elliot Bohm, Cardcash.com

 

Not Having Data Scientists

“Collecting big data is easier then ever and implementing tools to work with big data has also become much more accessible. The problem is oftentimes companies do not have a qualified Data Scientist or someone who can interpret or map/reduce the proper dimensions of data. Instead they rely on non-qualified personnel to interpret data. The improper analysis of data can be very harmful to a company.” – Phil Chen, Systems Watch

OK – but you can focus on what you do best by hiring a third-party BI firm to handle this – contact us for a complimentary consultation.

 

Answering Trivial Questions With It

“The biggest mistake that companies make with big data is using it to answer relatively trivial questions such as “what.” Big data isn’t about “what” questions; it’s about “why” questions. Big data is about joining data sets that have never been joined before and asking questions that have never been asked. It’s about knowing why customers and employees are doing the things that they do.” – Dusty Wunderlich, Bristlecone Holdings

 

Focusing on Data Processing at the Expense of Analysis

“Half of the challenge of big data is finding the right algorithms and approaches to ingest the vast quantities of information you have. The second, and more overlooked, challenge is finding a way to present your findings in a usable fashion. Too many companies focus on the former (how do we process all that data?) at the expense of the latter (how do we make it actionable?).” – AJ Shankar, Everlaw, Inc.

 

Confusing Correlation and Causation

“When companies work with big data, a major and common mistake is to assume that correlation implies causation. While you can use data to understand correlation, equating it to “cause and effect” can lead to false results and fruitless decisions. Making the distinction between correlation and root cause is critical to utilizing data for best results.” – Doreen Bloch, Poshly Inc.

 

Not Using It to Answer Business Questions

“There’s so much data being generated and collected, it can be overwhelming. Successful organizations start with the business questions they want to answer and then assess the data they have to answer those questions. Just looking at your mountain of data and trying to figure out what to do with it is a recipe for a lot of wasted time and effort.” – David Booth, Cardinal Path

 

There are additional ideas at the online article.

 

Learn how to put big data insights to work for you by contacting us for a complimentary consultation.

 

 

Source: 12 Major Mistake Companies Make With Big Data

 

 

#BigData

 

post

27 Free Books on #DataMining

 

Part of our mission statement is to make CPG reports that strive for a higher level of insights. Consider the 8 levels of analytics below as coined from SAS. Where does your organization stand on this list?

1) Standard reports – Standard reports provide summary statistics and answer questions like “What happened?” and “When did it happen?” said Davis. “That’s analytics, but not enough.”

2) Ad hoc reports – Ad-hoc reports answer questions like, “How many? How often? Where?” he said. They provide a level of independence on desktops that allow an individual, for example, to see sales in a particular region or at a particular point in time without needing to go to an IT governance counsel and wait three months for the result.

3) Query drill-downs – Also referred to as OLAP, query drill-downs answer questions like, “Where exactly is the problem?” and “How do I find the answers?” said Davis. This is for when an organization wants to see not only the results, but what the results mean and what backs it up, he explained.

4) Alerts – Alerts answer questions like, “When should I react?” and “What actions are needed now?” said Davis. “This is when you reach a particular threshold … something changes from green to red, so you do something about it.”

5) Statistical analysis – Statistical analysis answers the questions, “Why is this happening?” and “What opportunities am I missing?” he said. “You begin to take the data … and you begin to understand why things are happening.”

6) Forecasting – A popular level, forecasting answers questions like, “What if these trends continue? How much is needed? When will it be needed?” he said.

7) Predictive modeling – Predictive modeling tells users what will happen next and how it will affect the business, Davis said.

8) Optimization – Optimization answers the questions, “How do we do things better?” and “What is the best decision for a complex problem?

 

Want to read up on how to make it to level 8? We’ve found a list of 27 free machine learning (think artificial intelligence, data mining, statistical inference, predictive modeling) books to whet your appetite.

Don’t pull a muscle! For a free consultation on your business intelligence strategy contact us.

  1. An Introduction to Statistical Learning: with Applications in R
    Overview of statistical learning based on large datasets of information. The exploratory techniques of the data are discussed using the R programming language.
  2. Data Science for Business: What you need to know about data mining and data-analytic thinking
    An introduction to data sciences principles and theory, explaining the necessary analytical thinking to approach these kind of problems. It discusses various data mining techniques to explore information.
  3. Modeling With Data
    This book focus some processes to solve analytical problems applied to data. In particular explains you the theory to create tools for exploring big datasets of information.
  4. Big Data, Data Mining, and Machine Learning: Value Creation for Business Leaders and Practitioners
    On this resource the reality of big data is explored, and its benefits, from the marketing point of view. It also explains how to store these kind of data and algorithms to process it, based on data mining and machine learning.
  5. Data Mining: Practical Machine Learning Tools and Techniques
    Full of real world situations where machine learning tools are applied, this is a practical book which provides you the knowledge and hability to master the whole process of machine learning.
  6. Machine Learning – Wikipedia Guide
    A great resource provided by Wikipedia assembling a lot of machine learning in a simple, yet very useful and complete guide.
  7. Data Mining and Analysis: Fundamental Concepts and Algorithms
    A great cover of the data mining exploratory algorithms and machine learning processes. These explanations are complemented by some statistical analysis.
  8. Mining the Social Web: Data Mining Facebook, Twitter, LinkedIn, Google+, GitHub, and More
    The exploration of social web data is explained on this book. Data capture from the social media apps, it’s manipulation and the final visualization tools are the focus of this resource.
  9. Probabilistic Programming & Bayesian Methods for Hackers
    A book about bayesian networks that provide capabilities to solve very complex problems. Also discusses programming implementations on the Python language.
  10. Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management
    A data mining book oriented specifically to marketing and business management. With great case studies in order to understand how to apply these techniques on the real world.
  11. Inductive Logic Programming Techniques and Applications
    An old book about inductive logic programming with great theoretical and practical information, referencing some important tools.
  12. The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition
    This is a conceptual book in terms of data mining and prediction from a statistical point of view. Covers many machine learning subjects too.
  13. An Introduction to Data Science
    An introductory level resource developed by a american university that presents a overview of the most important data science’s notions.
  14. Mining of Massive Datasets
    The main focus of this book is to provide the necessary tools and knowledge to manage, manipulate and consume large chunks of information into databases.
  15. A Programmer’s Guide to Data Mining
    A guide through data mining concepts in a programming point of view. It provides several hands-on problems to practice and test the subjects taught on this online book.
  16. Data Mining with Rattle and R: The Art of Excavating Data for Knowledge Discovery
    The objective of this book is to provide you lots of information  on data manipulation. It focus on the Rattle toolkit and the R language to demonstrate the implementation of these techniques.
  17. Reinforcement Learning: An introduction
    A solid approach to the reinforcement learning thematic providing solution methods. It describes also some very important case studies.
  18. Pattern Recognition and Machine Learning (Information Science and Statistics)
    This book presents you a lot of pattern recognition stuff based on the bayesian networks perspective. Many machine learning concepts are approached and exemplified.
  19. Machine Learning, Neural and Statistical Classification
    A good old book about statistical methodology, learning techniques and another important issues related to machine learning.
  20. Information Theory, Inference, and Learning Algorithms
    An interesting approach to information theory merged with the inference and learning concepts. This book taughts a lot of data mining techniques creating a bridge between it and information theory.
  21. Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die [Broken Link] A great predictive analytics book providing an insight about the concept, alongside with case studies to consolidate the theory.
  22. Introduction to Machine Learning
    A simple, yet very important book, to introduce everyone to the machine learning subject.
  23. Data Mining and Business Analytics with R
    Another R based book describing all processes and implementations to explore, transform and store information. It also focus on the concept of Business Analytics.
  24. Machine Learning
    A very complete book about the machine learning subject approching several specific, and very useful techniques.
  25. Think Bayes, Bayesian Statistics Made Simple
    A Python programming language approach to the bayesian statistical methods, where these techniques are applied to solve real-world problems and simulations.
  26. Bayesian Reasoning and Machine Learning
    Another bayesian book reference, this one focusing on applying it to machine learning algorithms and processes.  It is a hands-on resource, great to absorb all the knowledge in the book.
  27. Gaussian Processes for Machine Learning
    This is a theoretical book approaching learning algortihms based on probabilistic gaussian processes. It’s about supervised learning problems, describing models and solutions related to machine learning.

 

Original Source: http://www.dataonfocus.com/21-free-data-mining-books/

 

#DataMining

post

Demystifying #SelfServiceData and the Future of #BusinessIntelligence

Imagine that you wanted to see a review of the latest weekend blockbuster, but you couldn’t just Google its Rotten Tomatoes score. Instead, you had to submit a written request to an information technology department, wait five days, then sift through a binder-sized report that breaks down the reviews by publication type, readership size, and reviewer age.

As fascinating as that information might be, it took five days too long — you went to the movie days ago on a hunch it would be good.

Welcome to the world of traditional business information.  This is the beginning of a look at self-service data analytics which is the future of business intelligence.

Continue reading the article from the link below and contact us for a free consultation on how to put self-service BI into the hands of those who need it the most – your category managers and sales analysts.
Source: Demystifying Self-Service Data and the Future of Business Intelligence

 

#BusinessIntelligence
#DataAsAService
#BigData