Data Science is exciting. From the moment we learn our first machine learning algorithm, it becomes necessary to apply it to data, make predictions, and uncover deep relationships.
But the business side is even more important and interesting. The author promises you that by spending some time understanding how a business works and what separates good companies from bad ones, you will receive significant dividends as you advance in your data career.
Why business needs Data Science
Companies use Data Science regardless of the size of their business, statistics from Kaggle (the professional social network for data scientists) show. And according to IDC and Hitachi estimates, 78% of enterprises confirm that the amount of information analyzed and used has increased significantly lately. Business understands that unstructured information contains knowledge that is very important for the company and can affect business results, the authors of the study note.
And this applies to a wide variety of areas of the economy. Read article about to explore how industries are using Data Science to solve their problems:
- online trading and entertainment services: recommendation systems for users;
- health care: forecasting diseases and recommendations for maintaining health;
- logistics: planning and optimization of delivery routes;
- digital advertising: automated content placement and targeting;
- finance: scoring, detection and prevention of fraud;
- Industry: predictive analytics for planning repairs and production;
- real estate: search and offer of the most suitable objects for the buyer;
- public administration: forecasting employment and economic situation, combating crime;
- sports: selection of promising players and development of game strategies.
And this is just the shortest and most cursory list of Data Science uses. The number of different cases using “data science” is growing exponentially every year.
Every Internet user and just a consumer every day, dozens of times come across products and solutions that use Data Science tools. For example, the audio service Spotify uses them to better tailor tracks for users according to their preferences. The same can be said for offering movies and series on video streaming like Netflix . And at Uber, data science is seen as a tool for predictive analytics, forecasting demand, improving and automating all products and customer experiences.
What is Big Data and why is it called “new oil”
Of course, data scientists cannot accurately predict the future of the company and take into account absolutely all possible risks. “All the models are wrong, but some of them are useful,” British statistician George Box sarcastically . Nonetheless, Data Science tools serve as good support for companies looking to make more informed and informed decisions about their future.
Growth stage of the company and the need for Data Science
While Silicon Valley is dependent on startups, the goal of a business is to be profitable. How this happens depends on what stage the company is in:
1. Hyper-growth
At this stage , sales growth is more important than profit , so the company is likely to suffer significant losses in order to attract more and more customers. The fuel for this growth is venture capitalists, whose goal is to scale revenues to such an extent that the company is attractive enough for an IPO (initial public offering).
Since growth is what it takes, data scientists at hyper-growing companies are focused on customer acquisition and product development. The problems they can work on include:
- Quantifying the most attractive leads.
- Development and optimization of marketing strategies such as coupons, referral campaigns, social media advertising, etc.
- Research and development of algorithms that form the core of a company’s product or service.
2. Public, Growing (Less Fast) and Profitable
Some private companies fall into this category as well, but for now we will focus on public ones (because Wall Street is the driver of the firm’s actions at this stage). When a company goes public, earnings per share (net income divided by outstanding shares) becomes the key metric . Government investors reward firms that can consistently increase profits and punish firms that cannot (Uber is a recent example of this). Thus, data scientists at these firms need to focus on more than just growth. Often these firms are large enough that optimizing expenses in the income statement is as important to profit as increasing sales…. Some of the challenges data researchers may face:
- Developing customer loyalty programs or other ways to engage customers and spend. The goal of this, and most of the next few points, is to maximize the customer’s lifespan.
- Quantitative identification of customers who are most prone to churn.
- Creation of recommendation systems for effective cross-selling of new products and services to existing customers.
- Use data and analytics to help identify new markets to enter or even companies to acquire.
3. Decline
While no one aspires to work for a declining company, these companies often have as many or more data needs than growing ones. Usually a company is in decline due to managers or its industry is in structural decline.
If a company is poorly managed, experienced people may have opportunities to use analytics to fix the situation. There are often so many areas for improvement in these firms that even simple analytics (a few Excel PivotTables) can have a significant business impact . Given the abundance of room for improvement, the bigger question is whether the board and management have the ambition, courage, and willingness to change. If not, then stay away.
Companies in structurally declining industries are putting their data scientists to the test. And it can be a test in which there really is no right answer, but only those that are less wrong. I would definitely recommend not associating your career with an industry in decline. But if you need to, as a data scientist, you can try to quantify why your industry is in decline and if there is any leverage to reduce the damage to your company compared to its competitors.
Also, is there room for consolidation? Consolidation (buying up your competitors) is often the only hope experienced by industries in severe downturns as it eliminates supply and regains some semblance of price power. Of course, most of these decisions are probably above the pay level of the data scientist, but I do believe that an objective, data-driven approach can be quite useful in such situations (the ability to effectively communicate quantitative results is also important).
