What about data context?

Posted by Oscar Wood on July 12, 2018 at 9:50 AM
Oscar Wood

Your team of data analysts just spent two weeks churning data that your organization collected, and the culmination of their efforts has yielded a visualization or two.  Sure, ok, you have the “A Team” of data scientists and analysts, so they managed to produce three solid visualizations and two reports to provide answers to the leaderships’ toughest questions.  

Now you are sitting in front of your laptop trying to decide how to digest the information, determine its viability, and make an overall recommendation to the organizations’ executives for action. 

Which of these four images below provides the most significant and actionable information?

choose_answer

Hint: see the bottom of the article for the correct answer.

Today’s technology space is filled with BI tools, analytic and visualization tools that present data in every conceivable chart-wizard format.  Some even take an extra step to offer the data scientist or analyst a workbench to fine-tune and tweak the data to improve the images’ layout.  The tools provide templates and automated wizards to help you get the picture just right for maximum impact upon presentation.  That’s important, but tools don’t take humans into consideration.  At the end of the process, there is a human that must stand there and justify the end-product and the expected eventual outcome.  

Coach Pat Summit said, “Responsibility equals accountability equals ownership. And a sense of ownership is the most powerful weapon a team or organization can have.” With the scenario above and quote as a backdrop, I’d like to address 3 things:

  1. Responsibility
  2. Accountability
  3. Ownership of Data

Responsibility:

Someone must be responsible for the outcome. Eventually a decision gets made, actions were taken, and predicted outcomes were achieved or failed based on the analysis and report presented. When the questions come flying, the team that built the product often times isn’t there to explain how the conclusion was derived.  Worse yet, the team that built the product has moved on and there is no one to provide the data to back up the original conclusions.

So with any data problem set and solution – your tools must provide visibility into how the product was made, who made it, and access to the original source data.

Accountability:

Someone must have accountability, good or bad, for those same decisions and actions based on the visualization.  But far too often, data management and BI tools force decision makers to “trust or have faith” taking for granted that resultant information presented is devoid of inaccuracies. If there are elements within the data that are inaccurate…possession

Where and how do you begin the chain of custody discovery? 

Are you willing to make a $750,000 decision for your company or customer without knowing the details of how your data team created their end product?  If so, are you willing to risk your job because you choose to “have faith” in the accuracy of the reports? 

Ownership:

Months go by, you survived countless data calls and demands for additional analytic products. You are told you need to migrate your data or you realize there are several analytic products that would be useful in the upcoming annual report. Problem is that data is locked away in a proprietary data format that doesn’t merge with the other tools and work products your team is using or the larger organization has invested.  Who owns your data?

This is the final part of the data conundrum.  When using these tools, the vendor is locking away your data into a proprietary format that you can’t unlock. It’s like the Hotel California! You can always get data in but you can never leave.

The Solution:

At NNData, we built NNCompass (pronounced encompass) to mirror our mantra in how we help our customers with their data challenges. NNData has a culture of providing the ability to manage and administer data with BOTH accountability and responsibility as core tenants so that you can own your data.  No data is unaccounted for as it migrates from one system to the another.  When analysts or anyone works with that data and there is a change via a transform or enrichment, that change is detailed for all to see.  In other words, if a data manipulation has occurred to omit data that’s misunderstood, or more commonly the case…the data is manipulated to create a prettier picture, it’s right there for all to see!   

NNCompass provides the single most important asset for your data enterprise and analyst’s reports…context.  Because text without context, means all the pictures at the beginning of this article are of little value.

If you want to be able to:  

  1. Trust your data
  2. Derive insights quickly without coding
  3. Avoid costly proprietary solutions…

Then talk to the team at NNData and we will help you with a product and a team capable of delivering a quality solution.

Answer to the question on which visualization provides the most insight and actionability:

As you've probably surmised... it’s a trick question. There’s no context to the images.  That said, I’ve had blind faith enough to play the fortune cookie lottery numbers!

nncompass-export

Topics: Data Management, Data Preparation, Data Wrangling

Everyone Should Be Able to Create Smart Data.

You can do amazing things with data now. This is where everyone involved in bringing NNCompass to life talks about how your business can start using things like machine learning to create "Smart Data".

If any of this sound familiar to you, you're in the right spot...

  • I really want to take advantage of things like machine learning, and see what it can do for my business, but I don’t know where to start

  • I have a data problem. Too much, too different, coming in too quickly. I don’t have 4M USD to hire teams of people to overcome my data problems.

  • I wish I could ingest and get a handle on all of my data, not just beautifully structured CSV and XLS files, but all my emails, MS Word docs, Adobe PDFs.

  • I don’t have the resources or time to first hire data engineers, then data scientists just so I can start using machine learning algorithms to answer some of my most difficult business problem sets.

Subscribe Here!

Recent Posts