Watson Analytics – classic

Announcement

Analytics Exchange and Watson Analytics

Recently announced at IBM Interconnect is the Analytics Exchange beta available in Bluemix. The Analytics Exchange gives you access to free and open data in categories such as economy and business, leisure, transportation, and others. The way to access the Analytics Exchange is by registering for BlueMix here with your IBM ID. Once you’re signed in with your IBM ID, click Dashboards. Then click Work with Data under “Data & Analytics.” Click Exchange in the left hand column. This will bring you to the Analytics Exchange. To access a data set, simply click one of the topics that interest you. I’ve clicked Environment, which returns 31 results. I’m going to select the data set Country Statistics: Refined Petroleum Products – Consumption. This brings up a brief description of the data, a preview, column details, and more information about where the data set came from. To explore the data set in Watson Analytics, click Explore Data in the top right. You’ll be asked to accept terms and conditions and then if you want to open the data in Watson Analytics. The data is now in your Watson Analytics instance. From there it can be analyzed like any other data set in Watson Analytics. We hope you enjoy access to these exciting data sets that can help you get up and running on Watson Analytics so you can find solutions to your business problems.

Resources

Analyst Report: IBM Builds Innovation into Watson Analytics

The Ventana Research report, IBM Builds Innovation into Watson Analytics is no longer available for download.

Resources

Taking a closer look at the new visualization types in Explore

Did you notice the new visualization types in Explore?  These are available to all users and are recommended to you based on your data. Watson Analytics will generate starting points for you, or you can pick them manually.  Take a closer look at the visualizations that have been added to Explore.   Packed bubble You can use a packed bubble visualization when you want to show relationships among columns that contain numeric values, such as revenue. It is similar to the bubble visualization but the bubbles are tightly packed instead of spread over a grid. A packed bubble visualization shows a large amount of data in a small space. The bubbles are in different sizes and colors. Because a packed bubble visualization uses area to represent numbers, it is best for positive values. If your data set includes negative values, use Color by to show the positive and negative numbers in different colors. If your data set has many negative numbers, consider using a bar visualization. For example, this packed bubble visualization shows how each product type is performing in terms of revenue. Each bubble is a different product type. The size of each bubble is determined by the revenue for that product type   Word cloud Use a word cloud visualization when you want to see a visual representation of text values. The size of each text value indicates its frequency or importance. For example, this word cloud visualization shows revenue for all product types. The Eyewear product type brings in the most revenue.   Summary A summary visualization can be used when you want to see the total for the measure that is shown in the visualization. This is ideal when you want to accentuate a very large or small number. For example, this summary visualization shows total revenue for all product types.   Network The network visualization is used when you want to see the connections among columns in your data set. A network visualization is a good choice when the data is hierarchical in nature. For example, this network visualization shows which positions are in each department.   And the categorical visualization is now called the heat map visualization. A categorical heat map visualization groups a column and uses a shape for each item in the column. It then uses size and color to show the relationship between two numeric columns. If you want a categorical visualization, add a column to the Points data slot. For example, this categorical heat map visualization shows revenue for each product line. You see that revenue varies within each quarter with revenue for some product lines being flat while the Personal Accessories product line is consistently bringing in more revenue. Use a heat map visualization to visualize the relationship between columns and you want it to be represented in a matrix type view. A heat map visualization uses color and intensity of the color to show the relationship between two columns. For example, this heat map visualization shows revenue for each product type and quarter. You can see how Eyewear compares to other products across quarters. It consistently brings in the most revenue every quarter. In Watson Analytics you can see the visualizations in the "Change the visualization" menu: And also in the starting points dialog: Let us know what you think, start a Discussion in our forum.

Resources

Watson Analytics' Intelligent Application on Financial Markets

When analyzing an asset such as a stock or a commodity, the conventional way is to draw it against time where the price of the asset is on the Y axis and the time is on the X axis as shown here in this graphic from stockcharts.com. However, sometimes, the investor does not know if this asset is expensive or cheap relative to other assets.    The time chart does not reveal this. This is where the power of Watson Analytics cognitive software can help.   It can easily produce intuitive multi-dimensional charts comparing assets relative to each other.   These kinds of charts reveal if the asset in focus is either expensive or cheap relative to 1 or more assets. The following example explains how an asset is very inexpensive relative to others when analyzed on the same chart.   Instead of having time on the X axis, we have another asset on the X axis.   The time is in the data, on the chart and it's labeled.   We added a third asset to the chart and it is the size of the bubbles. So basically, the next chart has 4 dimensions analyzing 3 assets relative to time.   We have gold stocks on the Y axis,  the Dow Jones Industrial average on the X axis and we have the gold price for the size of the bubbles.   Granularity is in years. The thesis of this analysis is to show how cheap gold stocks are relative to the general stock market (Dow),  the gold price and to time.  As you know, over decades, the correlation between the general stock market (Dow) and gold is almost zero.  This is why the following thesis is quite interesting. We will show the same chart 3 times with different time frames to show the history of the relative valuations of the assets.  This will help develop a pattern that is reoccurring again today. The above Watson Analytics chart shows how gold stocks, the Dow average and the price of gold fluctuated between 1990 and 2002.   In 1996,  there was a surge in the Dow as the internet age was born.   Investors were rushing to buy American equities leaving gold and gold stocks in the dust.   Investors sold gold stocks to all time lows, at the time gold touched a low of $255.80 in 2001.    The Dow was relatively expensive while gold stocks were very cheap.     However, as the Dow corrected and the internet bubble bursted in 2001 and 2002,  gold stocks started to surge as the gold price recovered. The same Watson Analytics chart between 1990 and 2011.   From 2003 to 2011,  gold and gold stocks surged when the Dow stayed relatively stable at expensive levels except for the financial crisis in 2007-2009.    The reason for the out performance of the gold stocks was because the US dollar depreciated against all currencies as the US Federal Reserve decreased interest rates to a low of 0%.   This fueled the price of gold to reach $1900 in 2011 as you can see it with the size of the bubble dwarfing the bubbles from 1990 to 2003. Hence, gold stocks went up because of this. In 2011,  gold stocks, gold and the Dow were all expensive as the bubble was located in the top right quadrant.   So here, an investor looking at this would rather be in safe investments like cash or bonds. However, the unthinkable happened afterwards and this leads us to where we are today. The same Watson Analytics chart between 1990 and 2016. The quantitative easing (QE) that the US Federal Reserve used starting in 2009, started to have a recovery effect in the stock market.  This effect was ironic as it was suppose to propel gold prices to new highs since QE has the property of diluting the US dollar.   However, it created a wealth effect to investors and the stock market started to surge leaving gold to correct resulting in a trend change for gold stocks. Since 2011, gold stocks have corrected from an all time high to an amazingly all time low. Surprisingly, the QE that the US Federal Reserve has injected into the banks has propelled the stock market (Dow) to all time highs creating a bubble in stocks, bonds and the US dollar sending gold stocks to absurd levels as gold corrected. Let's analyze this deeper from the above chart.   From 1998 to 2002, the Dow was at an all time high, gold was low at $250 and we had an internet stock bubble.   Today, the Dow is at a higher all time high,  gold is lower and we have a social media stock bubble. But there is a bonus here.   Look at the SIZE of the bubbles in 1998-2002 compared to today.   Today, the bubble is dwarfing the size of the 1998-2002 bubbles!!!!  This means that gold stocks are lower today with a gold price of $1200 US compared to when the gold price was at $255 in 2001!  Wow...   This is absolutely absurd if you think about it. What is occurring today is very similar to what happened in 2001 and we also know what the consequences were from that period. So, based on this chart, the gold stocks offer a terrific long term opportunity because it shows: 1) that they are very cheap relative to the Dow considering the current bubble is at the bottom right of the quadrant 2) that they are very cheap relative to gold as the 2016 bubble is very large compared to the 2001 bubble.  Also, the 2016 bubble is lower  than the 2001 bubble indicating that gold stocks now are lower than they were in 2001 considering that 2001 was a distress time for gold stocks 3) that they are very cheap over time as they are the cheapest in their history. 4) that the same pattern is reoccurring as it was in 2001 as the economy is experiencing the same investment conditions. Connecting history to this wonderful Watson Analytics chart,  it outputs very interesting and jaw dropping information when it is deeply analyzed.   This blog post does not represent IBM’s positions, strategies or opinions.

General Information

Everyone has an opinion - Starting a Conversation in IBM Watson Analytics

Have you ever seen something in one of your Watson Analytics reports that raises more questions than answers? That's where the Conversation functionality in Watson Analytics comes in. With Conversations, you can collaborate with other users inside of Watson Analytics by leaving comments and even creating polls that are saved with an exploration, prediction, or view. The conversations you create are only available for that specific asset. As with most social media services, within your comments you can also use the @ symbol to mention other users and the # symbol for keywords. For more information on using the Conversation functionality and getting the conversation going, see the How to add comments to analysis in Watson Analytics video on YouTube. For more information on creating a poll, see the How to create a poll in a conversation video on Youtube. Remember to subscribe to the IBM Watson Analytics channel on YouTube!

Resources

Using a secured gateway in IBM Watson Analytics

A Secure Gateway is required to connect Watson Analytics to secured on-premises data. The Secure Gateway provides secure connectivity by establishing a tunnel between Watson Analytics and the on-premises location that you want to connect to. Configuring your Secured Gateway in Watson Analytics could not be easier. In a few simple steps you'll be able to access your secure on-premise data using a client application such as Docker, IBM Installer or IBM DataPower. The required steps are outlined in this YouTube video entitled, How to Create a Secured Gateway.

General Information

How to ask a question about your data?

If you're wondering how to ask a question about your data, Watson Analytics will give you a few questions to start with or you can ask your own questions. And, you can always ask the coach. Yep, ask the coach. This new  YouTube video "How to ask a question about your data" will show you in a minute how easy it is to get started on your journey of data exploration. Remember to subscribe to the IBM Watson Analytics channel on YouTube!

General Information

Adding a visualization in IBM Watson Analytics

In this video you'll learn how to create a visualization by simply adding data to a dashboard using a sample data set.  Take a minute to watch this short video,"Adding a visualization in IBM Watson Analytics" then try it yourself using the sample HR Training Data set available right here in the Community. Remember to subscribe to the IBM Watson Analytics channel on YouTube!

Resources

SAMPLE DATA: Sales Win Loss

Description: Find the patterns in sales wins and losses. Understand your sales pipeline and uncover what can lead to successful sales opportunities and better anticipate performance gaps. Where to Access: Download: WA_Fn UseC_ Sales Win Loss.csv Also available directly within Watson Analytics as: Find Patterns in Wins and Losses Related Blogs and Videos: Flipbook: Closing the deal - Analytics for every sales person Blog: Discover Which Deals Are Most Likely To Close by Using Your Win-Loss Data Video: Sales Wins and Loses: Identifying and Understanding Your Patterns Suggested Data Analysis Use Explore to ask the question “What are the top supplies group by opportunity amount where result is won?”.  You will note on the insights bar after looking at this exploration “Top Route to Market by Opportunity Size”, you may want to add this to your exploration! Use Predict to determine what the main drivers are for Opportunity Result where there is a win?  Use the target “Opportunity Result” as a target for Predict” Use Assemble to gather the results of:  Comparison of the number of Wins and Losses in a bar chart, Bubble chart comparing Supplies Group by Opportunity Amount and a Bubble Chart comparing Competitor Type by Opportunity Amount in a single dashboard.  Once assembled click on Loss and Won to see the distributions change! Use Refine to create a data group for small, medium and large Rows and Columns This sample has the following format: # rows = 78,025 # columns = 19 Column Name Description Client Size by Employee Count Employee sized by number of clients. Values are: • 1: < 1k • 2: [1K, 5K] • 3: [5K, 10K] • 4: [10K, 30K] • 5: ≥ 30K Client Size by Revenue Client size based on annual revenue • 1: < $1M • 2: [$1M, $10M] • 3: [$10M, $50M] • 4: [$50M, $100M] • 5: ≥ $100M Competitor Type An indicator if a competitor has been identified Values: Known, Unknown, None Deal Size by Category Categorical grouping of the opportunity amount (OpportunityAmountUSD) • 1: < 10K • 2: [$10K, 25K] • 3: [$25K, $50K] • 4: [$50K, $100K] • 5: [$100K, $250K] • 6: [$250K, $500K] • 7: ≥ $500K Opportunity Number A unique generated number assigned to the opportunity Opportunity Results A closed opportunity is won or loss. Values could be Win/Loss Good example of a Target Field for Predict Region Name of the Region. Values could be : Mid-Atlantic, Midwest, Northeast, Northwest, Pacific, Southeast, Southwest Route to Market The opportunities’ route to market. Values are: Fields Sales, Other, Reseller, Telecoverage, Telesales Supplies Group Reporting supplies group Values are: Car Accessories, Car Electronics, Performance & Non-auto, Tires & Wheels Supplies SubGroup Reporting supplies subgroup. Values are: Batteries & Accessories, Car Electronics, Exterior Accessories, Garage & Car Care, Interior Accessories, Motorcycle Parts, Performance Parts, Replacement Parts, Shelters & RV, Tires & Wheels, Towing & Hitches Opportunity Amount (USD) Sum of line item revenue estimates by sales representative in American currency Sales Stage Change Count Actually a count of number of times an opportunity changes sales stages (back and forwards) Elapsed Days In Sales Stage The number of days between the change in sales stages. The counter is reset for each new sales stage Ratio Days Identified To Total Days Ratio of total days the opportunity has spent in sales stage: Identified/Validating over total days in sales process Ratio Days Qualified To Total Days Ratio of total days the opportunity has been spent in sales stage: Qualified/Gaining Agreement over total days in sales process Ratio Days Validated To Total Days Ratio of total days the Opportunity has presence in sales stage: Validated/Qualifying over total days in sales process Revenue From Client Past Two Years Revenue identified from this client in past two years • 0: 0 • 1: [1, 50K) • 2: [50K, 400K) • 3: [400K, 1.5M) • 4: ≥ 1.5M Total Days Identified Through Closing Total days the opportunity has spent in Sales Stages from Identified/Validating to Gained Agreement/closing Total Days Identified Through Qualified Total days the opportunity has spent in Siebel Stages from Identified/Validating to Qualified/Gaining Agreement Practice with this sample data and become an expert in no time.

Announcement

Curious about the world around you? Like to voice your opinions?

Then visit 'known' Powered by IBM Watson Analytics (https://www.ibm.com/blogs/known/). We've just launched known our brand new blog for exploring data within Watson Analytics.  Known is a unique digital experience where inquisitive minds can tap into the power of IBM Watson Analytics to find the often unexpected insights hidden in data. We've posted our findings in data narratives about a variety of topics including: The Business of Sports, Women in Leadership and The Fame Game.  But we want to know what you think.   We created known to start a dialog with our Watson Analytics users.  So we've also posted the data sources for you to download.  Do your own analysis then share the visualizations and your thoughts with the world via social media. New topics will be added on a regular basis and new articles will be added each week.  So check out 'known' Powered by IBM Watson Analytics and be part of the conversation.