What it takes to build a data science capability
AUGUST 16, 2017
There is more data than ever and the companies that make the best use of this data will win the race. Data scientists will be the key to achieving this victory. Having the right culture will help recruit and train the best data science team.
Top companies have realised that the key to sustainable competitive advantage is the ability to use data more effectively than their competitors. This will help them make better decisions and see opportunities before other people. Companies like Amazon, Starbucks and Netflix have shown the world that being with good at data can lead to unassailable lead over your competition. To get all this benefit, you need help from data scientists. So the simple solution is just to hire them? Unfortunately, it’s not that simple.
Firstly, there are not enough data scientist to meet the demand. There currently is a shortage of data scientists around the world, even in the US where data science has been recognized as a key to future growth for many years already. Even with so many university and online courses teaching data science, the world is still not producing enough good data scientists. This shortage is even more acute in Thailand where people has only recently realized the importance of data science.
Secondly, you need to screen for data scientists that have the capability and experience to produce high impact work to the company. Similar to other professions, there are both very good and not so good data scientists. If you are not in this field, it is not easy to tell the good ones from the bad. In our experience, not everyone who calls themselves a data scientist can perform at the level we expect of our data scientists.
Lastly, how do you retain your data scientists once you’ve hired them. Data scientists are in demand everywhere; they can choose where they work. What they are looking for is to be continually challenged by new problems that will help them improve their skills. Data scientists want to be surrounded by other data scientists who they can learn from. This means you need a diverse team of data scientists with experiences in different areas, for example, AI, statistics and language processing. Obviously not an easy task.
To overcome these challenges, a company aiming to build a data science team will need the right culture. These people are experts in their fields and need to have the freedom to explore new ideas and try different options. Because of this, there shouldn’t be a strict bureaucracy and a more relaxed culture will help boost productivity. They also need an opportunity to continually learn and improve. Support for further education is important. Data scientists should also be exposed to different projects to improve their skills, so management needs to rotate them to work on different topics.
If you are operating in a competitive field and don’t have the luxury of time to build such a culture to hire and train your data science team then the easiest route is to hire external help. This way, you get the benefit of having a data science team quickly and you know they have the experience to deliver the work you need. For example, when companies work with Sertis, they get access to nearly 50 experienced data professionals without the hassle of recruiting and training them. Companies also get to improve their teams by getting their employees to work alongside these external consultants. This way companies get to improve their internal data science capability.
Another option is to acquire companies with a strong data science capability. This is the favored route of Silicon Valley tech giants like Google and Apple. Siri that is on every iPhone was built by a company that Apple acquired. Obviously, this is an expensive route, but for many companies, it is worth the investment. Certain fields like AI is quite a niche and it is faster to buy the company than to wait to build the technology yourselves.
Whatever path your organization chooses to create your data science capability, it is best to decide soon. The risk of falling behind is too great that if you don’t have a strategy for this, it might not be possible to catch up with your competition. The future number one in your industry will be the one who knows how to best use their data asset.
This article first appeared in Thai on Think Data Science Column at Bangkok Biz News.