Outsourcing Data Science to Ukraine: 8 Do’s and Don’ts
Years ago, scientific researchers gathered and analyzed data by hand. They kept track of it in notebooks, and, in order to prove that their experiments (also called a “treatments”) had important results, inserted that data into statistical formulas, and crunched the numbers.
Computers changed all of this. Software programs would “crunch” those numbers and report significance. Out of all of this, data science was born.
Researchers and statisticians, in conjunction with IT specialists, realized that they could gather and analyze large amounts of data from a lot of different places to gain insights and make predictions. The concept of “Big Data” emerged next.
To the scientific community, this has allowed research to occur faster and, because of the volume collected, to be more accurate, and to allow for better predictions and solutions. Yet, Big Data is not just for science anymore.
Big Data Use Cases For Businesses
There is a growing understanding among business enterprises that take a data-driven approach to make decisions and find solutions which impact on the performance of their companies. And improved performance means increases in their revenue and profit.
The ability to digitally gather large amounts of data from multiple sources and perform an analysis on that data, allows companies to improve their effectiveness, predict customer behavior, and gain a competitive edge over others in their same niche. And when current software development includes machine learning and artificial intelligence (AI), big data collection and analysis become a highly-specialized IT niche. Data experts, also known as data scientists, are the people who perform all of these tasks.
Think about it. If you have been in business for 5 years, and you can take all of your customer information – demographics, purchasing behaviors and patterns, price points, etc. – and have that information analyzed so that it spits out conclusions about what you should do to gain and maintain a larger customer base, how valuable would that be? In fact, we’ve recently published a detailed case study outlining how e-commerce companies can increase revenue using Machine Learning solutions.
This is, on one small scale, exactly what big data analytics can do for you. And it’s probably already doing this for your competitors.
Data analytics, in fact, have become so popular that, during 2017, the big data market demand will grow to $33.5 billion in revenue for data science professionals and firms. And it will grow exponentially from there. Here are the predictions of big data market revenue for the next decade:
Image source: Statista
Trouble For the Big Data Science Environment
The problem for businesses is this: Data scientists come at high cost, and it is only the “big boys” who can afford to hire such specialists full-time. Add to this the prediction that, in the U.S. alone, there will be a shortage of data scientists in the range of 1.5 million, and lots of businesses will have to look to outsource this function. One of the primary sources for data analytics outsourcing is Eastern Europe, specifically in Ukraine.
So, let’s look at the reasons why Ukraine is so popular and then consider the Do’s and Don’ts when selecting a data science consulting firm in this region.
Popularity of Data Science Outsourcing to Ukraine – Facts and Benefits
Why Ukraine? Because it is one of the hottest spots for data science consulting in the world. Consider these facts for starters:
- A rapidly growing community of data scientists has been organized by Data-Science UA. Its goal is to make Ukraine the global hub for data science outsourcing services. To this end, is organizes coursework, workshops, meetings and more so that IT engineers can move into this field, under the tutelage and mentorship or some of the best in the business.
- The Science and Technology Center of Ukraine has a major focus on developing technology professionals and supports a large data science program.
- The Ukrainian Catholic University now offers a Master’s program with a specialty in data science.
- Ukraine is already known for IT expertise in developing the highest quality software. These firms have moved into data science with a passion and are providing training and development to their staffs through R&D centers that have popped up all over the country.
- Ukraine’s reputation for IT expertise in software development, AI, and now data science. As reported by Kyiv Post, in a series of articles, in 2015 alone, revenue from data science outsourcing services in Ukraine reached $2.5 billion. Figures from 2016 are expected to be at least double that. As well, several Ukrainian IT startups competed in the January, 2017 CES show in Las Vegas, and a co-founder of GitLab is now listed in the Forbes 30-under-30 success list.
- Several Ukrainian software development providers have moved into data science outsourcing services by employing full staffs of Ukrainian data science professionals. Romexsoft is on board too and we are offering on-demand data science services to our clients.
- Ukrainian IT professionals are, as a group, experts in Java and Apache Spark platform – perfect combinations for data science consulting.
- In looking at pros & cons of using Ukrainian IT outsourcing firms, most western countries prefer them because, according to Dou.ua, almost 80% of IT and data science professionals speak English.
- One of the biggest benefits of outsourcing data science to Ukrainian IT companies, of course, is cost. With the same level of expertise as their Western counterparts, Ukrainian data scientists cost far less. According to StatisticBrain, in fact, of all countries to which the West outsources services, Ukraine’s cost index is 6.3 on a 10-point scale. And given that 43% of all U.S. companies outsource IT services, from the purchase of software products to customized development services, big data outsourcing is a natural extension of already-formed relationships with firms and their developers.
Choosing a Data Science Consulting Firm – Do’s and Don’ts
There is a host of Eastern European firms now offering to consult in the data science industry. As you consider potential partners for your BDaaS (Big Data as a Service), here are your do’s and don’ts:
1. Do your Due Diligence – ask the important questions
- What are the backgrounds and experience of the team that will provide your big data service?
- Will you have a dedicated qualified team specifically assigned to your partnership? You don’t want to have to develop new relationships with new people along the way.
- How long has the company been in business and what references can they give you to check out? (and do check them out!)
- What experience does the firm have in your industry niche?
- Ask for a detailed proposal based upon your goals for data gathering and analysis. Do you want predictions of customer behaviors? Do you want analysis that will suggest new products/services? Do you want a predictive analysis of fraud and/or risk? Identify these before any discussion, and listen to what they say. Be wary of huge promises made without enough information from you.
- Do they work with all types of data storage systems and specifically yours?
2. Do Think in Terms of “Testing”
Before you launch into a comprehensive, long-term contract with legal and financial obligations, divide your gathering and analysis goals. Start with just one, and see how the outsourcer performs. If you are pleased, then obviously, you can move forward with a larger scope.
3. Do Be Totally Open
Just like you should never lie to your lawyer, you must be totally open when discussing your goals and needs. If your enterprise is experiencing falling sales, decreased revenues, etc., be honest about it all. A professional consulting team can then recommend the specific data to aggregate and analyze based on those issues. You can then compare their recommendations with those of other firms you are considering.
4. Do Involve Members of Your Team
It’s a huge mistake to ignore your sales department, for example, if you are going to use big data analytics to alter their work environments and the methods by which they operate. The same goes for your customer support department. If analyses will alter their operations, they need to be involved. In addition, they may know better what questions to ask.
1. Don’t Be Looking for a Magic Bullet
Data science is a process. Data is aggregated; it is then analyzed; recommendations and suggestions are made. Those outcomes are then tested through your implementation. More data is gathered and analyzed; recommendations and suggestions may be altered or tweaked as a result. You may be pleased with the results of initial recommendations, but do not stop there. If you want continued growth in sales, for example, there should be points down the road where data gathering and analysis should occur again.
2. Don’t Outsource to a Firm That Just Aggregates
There is no value in just gathering and aggregating data. There are tools that allow you to do this yourself. What you are looking for is a team of consultants who can dig deep, interpret that data, discuss the meaning of it in terms you can understand, make recommendations and justify those recommendations to you.
3. Don’t Give Away the “Whole Farm”
When you are giving access to your data, there is a joint responsibility between you and your partner outsourcer to keep that data secure. Don’t ever deliver data to a firm that has not explained to you exactly how it will be protected. And, you can protect some of it upfront. For example, by changing the actual customer names to assigned IDs, and after signing up the NDA, opening access to complete records.
4. Don’t Sit Back and Wait
The whole idea of a “partnership” is that there is mutual participation and involvement. Stay involved in the process as it moves along; get regular updates and reporting. A reputable firm will insist upon this anyway. If you are not involved in the entire process, you can’t make effective decisions about how far you want the analyses to go and how you can develop a timeline for onboarding the recommendations.
You are running a business. You are not a data scientist, nor do you want to be. What you want are qualified data scientists who can gather, synthesize, and analyze the data you have. You want professionals who can explain it all to you in simple terms so that you understand the recommendations and the “why” of those recommendations.
Have a conversation with Romexsoft, let us listen to your goals and needs, and then explain to you how our data science teams will be able to meet those.