Boost Your Analytics: The Rise of the Citizen Data Scientist

Are you struggling to hire talented data scientists to glean insights from your corporate data? There's currently a lack of big data talent hampering corporate analytics and causing nightmares for CIOs, but I have good news for you: you may already have all the data scientists you need!

There are alternatives that could help you boost corporate analytics. The rise of visual data discovery and approachable analytics (read more here) are making room for a new hybrid role within organizations: the citizen data scientist.

With easy-to-use visual analytics tools, the citizen data scientist is able to create models that use predictive or prescriptive analytics, even though they're not statisticians.

It's a new hybrid role that will prove extremely valuable to your organization. These citizen data scientists can use their business knowledge and embed it into the models they're testing, prototyping or building. And in collaboration with different stakeholders, they can help drive initiatives from all the different business units within the enterprise. Their data and analytics undertakings will give a quick start to new and innovative projects.

Your Citizen Data Scientist Strategy

So, where do you start? Begin by determining where you already have pockets of expertise, where more talent is needed, and where talent could be better used.

Next, decide which analytics initiatives should be further developed, which should be maintained, and which should be cancelled. The key here is to work in smaller iterations and fail fast when necessary.

Leading organizations that are fostering a citizen data scientist culture and cultivating their analytics teams, have a much better chance at acquiring deep analytical talent. Keep this in mind when you start developing your analytical teams and the way they collaborate.

Given enough time to grow, your junior hires can be taught the skills they need for your company's business and industry. In return, they're provided an upwardly mobile career path.

One training strategy that firms use is to rotate talent through different business silos, allowing employees to combine their analytical knowledge with domain knowledge to gain a stronger understanding of how actionable insights from analytics can impact the business.

From a technology perspective, you will need to equip your newly assigned citizen data scientists with the right tools so that they can combine their business expertise with analytics to achieve the best possible results.

Implementing a citizen data scientist strategy is a decision that should not be taken lightly. It needs to fit into the overarching goal of building a firm foundation for the enterprise insight platform that will support your data strategy for years to come (learn more about future proofing your data strategy).

To summarize:

  1. The shallow pool of deep analytical skills should not hamper your corporate analytics.
  2. The rise of visual data discovery and approachable analytics have forged the hybrid role of citizen data scientist.
  3. Executing a good strategy regarding this new hybrid role and collaboration with different stakeholders within your organization will boost your corporate analytics.
  4. An enterprise insight platform (the technology) should be built to enable different types of users to capitalize on data insights.
  5. What are your experiences in boosting your corporate analytics? Where do you see opportunities and what are your challenges?

Learn more about how to find and equip citizen data scientists.

This content was reposted from the SAS Learning Post. Go there to view the original.

Natan Meekers, SAS Senior Associate Systems Engineer

Natan Meekers is Digital Spearhead, Data Discovery & Reporting, at SAS.

Re: Bridging the gap
  • 2/22/2017 1:29:58 PM

Agreed that the interface will need to be intuitive enough that the creativity of the end user isn't hampered.  What I see a lot of software packages doing now is giving you dozens of canned templates that get close enough 80% of the time.  It's that 20% that you want your users to feel comfortable in.  To be successful in the future I think these tools will have to cater more to that 20% as not to drive users away once they get past the surface level of the software.           

Re: Bridging the gap
  • 2/21/2017 3:56:19 PM

@ SaneIT. That sounds like a good case for open source anaytical software where companies can tweek things to their indiviudal needs.  No one software can solve it all. 

Re: Bridging the gap
  • 2/21/2017 3:55:05 PM

And not only must the interface be easy to operate and understand, but the user probably needs some education and incentive to make the efforts necessary to take on the tasks as well. It's one thing to have the tools ready to go, but if the staff isn't convinced of the need or necessity they aren't going to be implemented fully.

Re: Bridging the gap
  • 2/21/2017 11:35:00 AM

Ha! Great point, Joe. I'm no fan of Excel either, but at least there's plenty of online help available to steer me through calculations or processes I find cumbersome and confusing.

Re: Bridging the gap
  • 2/21/2017 10:10:30 AM

I'm curious about how quickly the tools that might be suitable for the citizen data scientist can reach a level of reasonable maturity. It isn't surprising that the first generation of tools are suited only for the data sets that might be considered low hanging fruit.

Until those tools do mature a bit we probably will be stuck with having citizen data scientists go through crash courses in data science so they will have some degree of proficiency.

Another thing that will be interesting to see is how quickly the technology providers and the higher education systems respond to the likely need to train business managers in some level of data science/analytics. Some MBA programs are already integrating analytics elements into their curricula. Will those managers be content with an involvement in analytics that is limited to considering what data can do for their organization? Or, will they want to get hands-on with manipulating data and experimenting with their own models? If it's the latter, then those "citizen" tools better be ready.


Re: Bridging the gap
  • 2/21/2017 8:27:30 AM

Sadly, what I find is that the new tools while easier to create very simple graphical depictions of data as soon as you step outside of a preconceived reporting model it is not only difficult but often impossible to get what you need. I've had some talks with companies lately that I'm using their product three different ways to get sets of data out.  Out of the box it works very well if you want to look at the data they have decided everyone wants.  Getting deeper into the data means making connections using another version of the tool or an external tool.  If you can keep reporting simple, then yes we're in a much better place than we've been in the past but if the tool's developers didn't think of things that you need it can be even more frustrating.


Re: Bridging the gap
  • 2/18/2017 3:55:36 PM

I agree that most in the past have not been too user firendly. The new visual analytics tools , however, are much more intuitive and are easier to use.

Re: Bridging the gap
  • 2/18/2017 2:25:09 PM

@Terry: Meanwhile, many in the field are still saying that Excel is the real problem.  My take: The tool is the tool.  How are you making it accessible?

Re: Bridging the gap
  • 2/18/2017 1:54:57 PM

I'm in total agreement with simplifying the interface, Seth. This has been an ongoing complaint I've had with many analytics packages -- expensive to acquire, cumbersome (and expensive) to learn, challenging to manage over the long term.

Who's doing it?
  • 2/18/2017 1:50:11 PM

Interesting post, Natan... and one certainly to draw second and third looks from those who run major departments or lines of business.

One easy question: Any anecdotal examples of organizations that may have already benefited from data collected and analyzed by citizen data scientists?

Page 1 / 2   >   >>