P L E A S E  W A I T...

How to build Data Science capabilities within your organization

Modern businesses are flooded with a huge amount of data. And it is this enormous data that holds the key for better decisions to accelerate business growth. Simple! What more could any organization need? But here is the problem. To get the most out of this Big Data, it has to be analyzed , processed and this can only be done by experts who understand this data and handle it well.

Enter data scientists!-The new superheroes who use cutting edge technology to sieve data for actionable insights, and consequently make data science a highly sought after skill set by CEOs, management and HR officials for their workforce. Although learning data science seems complicated, hard and a never ending process, a dedicated, consistent effort by any small, mid-sized organizations to upscale their employees by teaching them the skills required for Data Science can bring in great results.

This is how. And I ‘d like to call it The five point plan for building Data Science capabilities

These five points will help your organization pave the way for running the data science program successfully and trust me, these are tried and tested.

Spread the word

Familiarize and prepare your employees with the shift towards data science. Reiterate the data science program in town hall and senior leadership meetings with emphasis on the transformation. The launch should be announced by the senior most person in the organization like the CEO or a director to show the commitment of your organization towards the data science program and it significance.

Get them started

Signing up for a data science program is half the battle won. But only a strong, steady commitment and effort will take it to completion and yield amazing results. You as an organization may be clear on the ‘why’ of the whole endeavor. You know that more self-sufficiency and expertise will bring in more revenue. But without communicating the benefits learning data science has for your employees, you are unlikely to see genuine involvement. You can encourage a buy-in from employees by showcasing the future career path, rewards of upskilling, higher payouts for working on advanced projects or even the fear of being left out( I hate to say this but this is how the cookie crumbles). Of course, the seniority in your organization needs to weigh the pros & cons of such a transformation and accordingly roll out the mandate to selected groups as there may be employees who may not be sold to the idea of building the skills required for data science at all.

Give them a digestible data science program

Below are the three must haves.

Simple syllabus- This is vital. You can begin with an eye-catching name for the program for data science that not only attracts attention but clearly portrays your intentions. Stress on quality rather than quantity learning with inclusions of a few essential concepts and facts. Hire experts to develop a good, comprehensible syllabus with assessments.

I personally like Udemy for their bite sized learning syllabus for the data scientist course.

Short-term modules– Breaking the program into short-term modules of about 2 -4 months works well for regular and timely deliverables given the short concentration spans. A final client related project to work on serves as a wonderful icing on the cake for practicing the data science skills.

Gratification- One of the most vital components is instant gratification. An appreciated mind always gives unexpectedly excellent results! This can be rolled out as packages of monetary rewards, leaderboard display, certificates, points, badges and gamification. It may facilitate increased participation with an enhanced display of talent, pull the employees out of their comfort zone and filter the ‘not interested’ ones during the data scientist learning course

Use the right tools and platforms

Monitor progress– A great deal of time, energy and effort is saved by a wide variety of platforms that provide a bunch of tools and services for the data science monitoring. They track and test the employees progress during the data science program. This can keep your employees on their toes. The task can be efficiently carried out though a business account that showcases user learning movements and which can be reviewed by the management. This enables them to identify and poke sluggish learners to improve their pace. Smaller organizations who can’t afford a business account can manually create a Google sheet which each learner can update on weekly basis with the progress of the skills required for data science.

Test knowledge- This is as essential as monitoring the progress. You can design multiple customized tests as per your need or hire freelancers to do the same, for data science practice. You can foster a healthy learning environment where employees can come together and solve problems as a team and share their collective knowledge. However, individual tests should also be conducted during the data science program to discourage a ‘crowd thought scenario’ and an unfair delegation of work within the group to complete assignments.

A controlled examination process at the end of each module is also a good idea. ‘ClassMarker’ platform does this by hosting all your assignments, tests and examinations. To obtain a qualitative and immediate feedback at this point, you could reach out to the same freelancers to get this task executed.

Practice! and practice! and practice!

Practice makes perfect. Repetitive tests in the data science program may give you long faces and weary glances, but stands to prepare your employees better for upcoming challenges, tasks and troubleshooting. Their confidence gets a considerable boost when they actually practice something hands on rather than just watch it passively on a video about the skills required for data science.

Oh heck! Why can’t we just hire ready data scientists and save ourselves all this trouble!

ROI, I’m sure is a common and crucial consideration in your company. If you have a data science program like the one I’ve suggested above, it should be running at approximately INR 800,000 for 100 users (8,000 per user). And here’s the deal. This is lesser than the amount you shall spend on hiring a single data scientist. And the program usually guarantees a 60% success, which means you are preparing 60 internal data scientists in 8 months considering a sample size of 100. You can do the rest of the math.

However I’d like to stress on the appointment of one senior coordinator for this program who can drive initiatives with persistence. He needs to stitch all the parts of the data science program together and make minor alterations during the course to ensure its success.

In a recent Harris Poll of 515 U.S. human resources and business managers conducted for Glassdoor, 48 percent of respondents said they are unable to find enough qualified candidates to fill open positions, and 26 percent of respondents anticipated this to become a larger problem in the coming months.

Hence, a smart company will put its business on the fast track by developing its own people with skills required for data science, through a comprehensive data science program, rather than spending a huge amount of money, time and effort in trying to hire someone who fits the gap ideally. A fully customized, internally trained workforce in data science is the bright future and it sure can metamorphose the company for an overall celebrated success.

Author Details

Praveen Dalal is the Director of HR & Partnerships at Nabler & drives change management processes. He has a decade of experience in working small & medium sized companies & bringing transformation through training & performance management

Drive better results by understanding customer data