"Chart" new territory with your data

Amazing discoveries start with an innovative mind willing to look at things differently. Take Columbus, they said he was crazy for sailing the ocean blue in search of the “new world”. Well here’s another outrageous idea for you!  What if you could use your Big Data project as a way to make additional revenue? Here are some ideas so that you can begin to chart this unknown territory with your Big Data, and turn your discoveries into dollars.

3 ways to monetize your data

It is logical to use company data to save money and find cost savings internally. But what if you take another approach with that same data? Check this out-- U.S. News and World Report was able to make their own discovery.  They created the criteria and collected the data on college rankings for decades. And each year universities fight for the top rankings in their region or for a particular education track. They produced these ranking reports geared toward the prospective student. One day they stepped back and took another look at the rich data they had collected over the years, realizing they had another (big) market for this information. If they could package and sell it in a new way, to the colleges and universities, they could provide valuable insight and create new revenue streams!

So here are some tips to help you think outside the box with your Big Data.

1.  Make it unique

Think of ways that you can make the data unique to your audience and their needs. You have data that no one else has, and it can help users make better decisions. Think about who, outside your business, could benefit from this unique information, and how they can benefit. Then apply some additional strategies to really make your data a must have:

Mashup - combine your Big Data with a public data set

What would happen if you combined your data with a data set on or another public set? Perhaps you work in the public health sector as an executive of a health insurance company. You could overlay your Big Data with government census data to identify healthcare trends that a growing hospital needs to plan for. The hospitals could use your data product to set up their hospital for the future. Here’s a list of companies already using government data in creative ways.

Predictive Analytics - find the treasure in future trends

Can we apply an algorithm to our data to find some special meaning or make the data more helpful? Predikto is one company that has this down in the railroad industry. They have a great product to predict the breakdown of railroad track safety monitors. Their product analyzes a plethora of data from weather to train loads to provide maintenance crews critical yet simple health-check displays, so they can easily see when these monitors are likely to fail and preemptively send a crew out for repairs before any damage is done.

Composite Metrics - if you build it they will come

Sometimes a simple metric isn’t enough if it can’t fully describe a behavior or the performance of a system. That’s when you need to come up with a Franken-measure: a made-up metric that creates a comprehensive composite to capture complex concepts. Think Google’s PageRank or the NFL’s Passer Rating. PageRank combines multiple complex metrics on web traffic and trends in such a way that the end result is something we can understand and use.

2. Put your best efforts into the user experience.

By putting yourself in your user’s shoes, then you can design data products much more effectively. First, like we mentioned earlier, you need to really think about who your audience is and what your audience needs to get from the data. How does this impact the way you tell the story of the data, and how you design the product so that the users can see the value immediately?

More often than not, the heart of the designer’s message is lost among all the metrics and charts. In this flurry of enthusiasm to display tons of data, little attention is paid to the user and guiding them on how to consume the information.

Remember, your data consumers are not the experts in the data like you are.  Your users probably have responsibilities other than analyzing data. Give them the high-level path to follow, and let those users who need more info have the option to drill down into the details. Think about the delivery of data much like the way you tell a story, provide a beginning (starting point), middle (critical details) and end (decision points).

3. Start small, design one product first that solves a real problem easily.  

It’s better to prototype a data product that is ready to put in front of a user in six weeks instead of six months. This allows you to keep it simple and make adjustments quickly based on what’s working and what’s not. Think like Google. Put out a concept or idea as a beta, study the user responses and feedback and add more capabilities as you go. This kind of logic allows for a quick release, less investment in development of the product and the opportunity to grow with the consumer.

Now that you are ready to set sail and chart your own new data territory, here are more helpful leads to help you do more with your data products!

Join us in December for our webinar on Turning Data into Dollars.

Also check out DJ Patil’s (the U.S. Government’s Chief Data Scientist)  free e-book, Data Jujitsu, the art of creating a data product.

Finally, take a look at our own, Zach Gemignani’s slideshare on turning data into dollars.

For a demo of our product, Juicebox, schedule an appointment.

Not Knowing Where To Start

Books, movies and music all have a beginning. Data, when presented or shared, often does not have an intuitive starting point.  The challenge of not having a clear beginning is that when you see a dashboard littered with a dozen competing charts it’s easy to disengage. Tables of raw data can be even worse. Dashboards or reports are often designed to deliver everything and the kitchen sink.         

Here are a couple of examples of dashboards that miss the mark in terms of telling their audience where to start.  In both of these cases the user has to be familiar with the data and know how to read the information correctly.  Beginner or infrequent users will struggle to understand the value of this data.  Without guiding them, the users can lose interest and choose to avoid using the information altogether.





Good dashboards or reports start with a high-level summary and then let users progressively and logically drill into more complex details and context. They are also simple and uncluttered. They use white space and have a clear visual hierarchy.   Here are a few of alternative examples to get the wheels turning.                        

Even this more advanced interactive visualization, called a TreeMap, offers clarity on where to start and how to use it.

To have your audience follow your story it’s important to get them started on the right path.  Think Steven Covey’s, Begin with the End in Mind.  Just like a story your audience is along for the ride.  Carry them from initial explanation to a new, shared understanding.   Only then will they begin to value the effort you put into assembling and presenting the information you’ve given them.

For a demo of our product, Juicebox, schedule an appointment.

Find out more on effective data visualization from our book, Data Fluency. Excerpted here with permission from the publisher, Wiley, from Data Fluency: Empowering Your Organization with Effective Data Communication by Zach Gemignani, Chris Gemignani, Richard Galentino, Patrick Schuermann.  Copyright © 2014.

Decorating data

An early Christmas present has arrived from the DabbleDB team for the 100 million or so of us that have to work with data on a day to day basis.

They’ve created a do-what-I-mean web tool that lets you show how you want data to be restructured and bang! it’s done. Check out the video.

Cleanup data in action

It’s a great idea and a elegant, easy to use interface. There are so many directions I’d love to see them take this tool.

Cleanupdata is a great name, but they’re really giving you better ways to restructure data. This tool won’t help you find and fix errors and anomalies in data. At least not yet.

I also hope they extend cleanupdata to let people automate these data restructuring operations. If only you could apply a cleanup created in to 1,000 Excel spreadsheets or to a database table.

If you like this, it’s worth checking out DabbleDB. They have rethought the database with a database/spreadsheet/web forms/visualizer platypus of a tool. It lets your data be pliable in ways that databases don’t allow, while retaining structure that spreadsheets don’t recognize.

Added: Avi Bryant, one of the authors of the service notes that the example in the screencast is motivated by this post on cleaning data in Excel. Compare and contrast. I know most people would prefer to avoid ="("&MID(H2,1,3)&") "&MID(H2,4,3)&"-"&MID(H2,7,4) in order to format a phone number.

ANDs, ORs, and IFs: Comparing big lists in Excel

One problem we face when manipulating large amounts of data in Excel is checking to see if two lists of the same length contain the same items. For instance, we might be given a list of products that a company has for sale this month, running to thousands of items, then the following month, we get another list of products for sale and we need to see if there has been any change between those two lists. This isn’t too hard to deal with when you only have a hundred or so items, but it gets a little thorny when your list runs to tens of thousands.

What we do is line the two lists up, side by side, in sorted order.

Two lists

Use the simple “A1=B1” formula to compare pairs of items in the lists.

Formula for comparing elements in the lists

If the pairs are the same, this will be true, otherwise they’ll be false.

Results of comparing elements in the lists

Copy this formula down for all your rows. Then use the AND function and give it the entire range of comparison formulas.

Checking comparisons

This will only be true if every single one of the values in your list are exactly matches. If even one comparison is false, this big AND statement will evaluate to false.

This is a quick and dirty approach. For tougher problems, we use a slightly more complicated formula in the comparison where we evaluate it to 1 if the value is true, 0 if the value is false. This gives us more flexibility to combine comparisons, but that’s a topic for another post.