Is Big Data Buzz Distracting you from the Data Movement Fueled by Tableau & Qlik?
Guest Blogger – Jakki Geiger , Senior Director, Solutions Marketing, Information Quality Solutions at Informatica
Are people at your company using next-generation analytical tools such as tableau and Qlik?
If you answered yes, then your organization is one of the growing list of 32,000 organizations using QlikView and 24,000 organizations using tableau to organize data, visualize data, analyze data, and take action based on data insights without any technical skills or enterprise IT involvement.
Have you heard of the term “Citizens Analyst”?
If you answered no, you are not alone. During the Informatica World tour Boston on Oct 20, about a handful of data experts confirmed they were familiar with this term.
“The fact that not many people have yet heard of this term points out that data practitioners are too caught up in this big data buzz, which is drowning out what is arguably an even bigger change happening in the data industry,” said Kurt Rosenfeld from Corporate Technologies during our onstage interview.
Here is a summary of our discussion, where we talked about what a citizens analyst is, what it means for data practitioners, and how to get started on meeting the data needs of an organization where everyone is an analyst.
Jakki: Kurt, how would you describe the Citizens Analyst?
Kurt: Let me describe the change by stepping back for a moment.
We’re all used to that pyramid view of BI – the one that shows information consumers are a hierarchy with the executive team wanting dashboards, analysts wanting ad-hoc self-service, and the hordes getting reports. And we’d design the BI architecture to fit this pyramid.
What’s happened is that now everyone is an analyst –business managers, supervisors, accountants, sales reps. They’re all tapping into data to understand and improve performance. And these new BI tools, such as tableau and Qlik, are the catalyst.
I also think there is a demographic aspect. This trend is coinciding with a new generation of professionals – the kids that grew up with the Internet – for whom “search and investigate” is a very natural way of working.
I am using “Citizens Analyst” to reflect this democratization of BI. It’s incredibly empowering and it is already having a deep impact on businesses.
Jakki: It’s exciting to think about everyone in the organization being empowered to be an analyst. How do you build a data architecture to support this new reality?
Kurt: Yes, it’s all about the data. There’s another industry change coming right behind it. We might even call it Citizens Data Integration.
A new breed of tools allow that same citizens analyst to connect, prepare, and load the variety of data they need to analyze. Sales data, service data, finance data, HR data, whatever it is. On premise or in the Cloud, Informatica has a jump in this space.
So, now you can see the workflow as follows…
- Connect to data
- Select what to use
- Mash in other data needed
- Visualize the results
- Do some investigation (i.e. analytics)
- Refine the visualizations and refine the data until you “get it right”.
Iterative thinking and iterative doing working together. This how people are most productive. You can see why it’s so compelling.
Jakki: Are there business issues arising because of this new way of analyzing data?
Kurt: Yes, there are business issues. For example:
- How do you handle sales territory changes and yet keep the numbers accurate across such changes?
- When mashing up data, how is the customer service call related back to the product purchase?
- How is your customer churn ratio being calculated and why is it different than mine?
And these are big issues. They translate into loss of trust in the numbers and the process, which are all very damaging.
Jakki: What about technical issues?
Kurt: A critical one that relates back to the business problems is the erosion of a good data foundation. We all know data is messy and complicated. Traditionally powerful BI analysis required a dimensional model – and that is where we’d sort out the mess and complexity.
But the visualization tools remove this need. A very clever part of their technology is an internal data engine that can figure out the relationships, hierarchies and summarization grain on the fly. And they can handle millions and millions of records. Just feed the tool flat transaction data, and you’re instantly analyzing your business. So all that hard dimensional model work is bypassed and goodbye to that good data foundation.
There are also serious systems issues that rapidly multiply:
- More and more and larger and larger extracts running all the time
- Demand on the data sources becomes unsustainable
- Old extracts and visualizations are saved all over the place
- Tons of nearly identical data is being backed up and copied
- And source data changes have unknown downstream impacts to “everyone”
Jakki: What’s the solution to these business and technical issues?
Kurt: The solution is data architecture and data governanceframed as an evolving balance with the citizens process.
Too much governance = you choke off the citizens analyst and it is bypassed
Too little governance = things run amok and trust in the data is los
I like to think of the Internet as a good example of this balancing act. There is no process and control. It avoids chaos because of helpful services and protocols that people like to use – and the decentralized internet works.
We need the business analytics equivalent.
And the good news is that the users know there is a problem and they want the help.
Jakki: Can you give us an idea of how that might look?
Kurt: Well, we need to deliver a mix of “helper” services that provide the following:
- A way to find the right data with uniform definitions, business codes, hierarchies and relationships (data mastering)
- A way to explicitly know where the data has gaps or errors or inconsistencies, and help correct it where appropriate (data quality)
- A way to control access or mask sensitive data (data security)
- A way to keep track of where everything is (to find it and to manage change)
These are all services that IT brings to the table. And IT should be looking across the vendor landscape for players that do this well – clearly Informatica is one.
Jakki: What’s your advice on getting started on the journey to supporting an organization where everyone is an analyst?
Kurt: First of all, IT must acknowledge this is happening. It’s not going to stop. And IT needs to commit to figuring out how to engage and support it. And a willingness to be open minded about throwing out traditional thinking is fundamental.
For example, reconsider the role of the dimensional model. It shouldn’t always be the first stop on the analytics journey. Indeed, we advocate it is often not the first stop. We just worked with a telecomm firm that needed to understand how network performance was impacting customer churn – especially new customers captured through marketing promotions. Millions of records analyzed every day with no dimensional model. We did this by creating data architecture strategies within the visualization platform itself.
And IT must plan out the roadmap for these “data helper” services.
To learn more about meeting the data needs of Citizens Analysts, read:
If you enjoyed the insights shared in this interview with Kurt Rosenfeld from Corporate Technologies, join us for an upcoming Informatica World Tour in New York or Chicago! Click on the links to learn more.
- Informatica World Tour New York, December 8, 2015
o Terri Mikol,Director, Data Governance Office, UPMC
o Jeff Semonovich, Senior VP Managing Director – Enterprise Services, First Niagara Financial Group
o BJ Fesq, Chief Architect and Chief Data Officer, CIT Group
o Larry Keyser, Solution Architect/Technical Lead, United Technologies
o 20 minute interview with Deloitte
o 10 minute interview with Corporate Technologies
- Informatica World Tour Chicago, December 9, 2015
o Mike Hart, VP, Director of Data Management & Technology Americas Research, JLL
o Jer Thorp, Co-Founder, Office for Creative Research
o 20 minute interview with Accenture
o 10 minute interview with Cloudera
To view the original blog post, click here