Running my DataScience SG community, I often have hiring requests from employers directed to me. Unfortunately, when these employers shared their hiring demand, job description etc, it shows that they are not fully committed to building the Data Science capabilities. Why will you say they are not fully committed? Because the positions that they offer are usually contract positions (and usually about 2 years long). The lack of commitment can be a huge roadblock to building up the capabilities, because its a signal that leadership is unsure and the tendency is to go towards "NO".
Will Talents Join?
Most of these hiring requests that I get, they require the talents to have some years of experience. Here is the reality! Talents that have a few years of experience and are very good with their work, will have been GAINFULLY employed (i.e. they will have held permanent positions). The reason is the professionals in the field of data are in high demand.
Put yourself in their shoes, being gainfully employed, are you going to risk your current position and move to a contract role? I can only think of two reasons for moving, firstly, push factor: they HATE their current position and pull factor: the potential employer with contract position have very unique and interesting projects.
But HATING the job and taking on uncertainty in life are two different things, which means, yes the talent may hate the current position but its no reason to take on a contract role. Secondly, an organization that offers a contract position, very likely have just started out their data science journey, beginning to tap onto their data for more insights, starting to move up the learning curve. Organizations that just started have a VERY SMALL chance of having projects that are of interest to the experienced data scientist.
Since the contract position is not attractive to good talents, very like the "greens" will make up a big part of the application pool or..... Regardless, this will reduce the value that the organization can potentially tapped on given that good talents is needed to make data science works out.
Its a Journey!
In my opinion, Data Science will become a business function like Accounting, Operations, Logistics that business cannot do without. With the improvement in technology, businesses will continue to collect more data (or data will continue to be the by-product of business processes). Its unfathomable why businesses will not want to tap onto it to understand what is happening on the ground, so as to come up with better processes and strategy.
But its a LEARNING journey. Why is that so, you might ask? Think about it, technology keeps improving allowing businesses to capture more and varied data. If we claim that "data is the new oil", we need to learn the latest technology. New tools are being developed and become obsolete quickly. Best example will be Hadoop. Hadoop became very popular during the Big Data era, from 2012 to around 2017 but with the advent of cloud computing, there are very few mentions of Hadoop these days. Thus businesses need to learn and keep up with the latest tools.
Algorithms keeps being updated with variations or new algorithms comes up to replace the new one (we are now talking about machine learning algorithms that are suitable for quantum computing these days). The market is evolving as we speak. We need to constantly sent out "feelers" to understand our customers better, given they are businesses' lifeblood.
One can see there are many things to learn about in data science, things like technology, infrastructure & tools, algorithms, the market. As such, it is never a single project but a series of projects. Its a learning curve that the more experience you commit, the better you get. Climbing up the learning curve is a competitive advantage on its own.
How shall I start?
In a nutshell, make sure that the value from your Data Science initiatives is always greater than the costs of running it. No sane business owners going to run loss incurring initiatives. Simple as it may seems but its not easy, I have seen many initiatives failed because of overrun costs.
To kick start, perhaps is to hire an experienced talent (aka consultant) to scope relevant projects. One word on hiring consultants! Please ask around, get it through word-of-mouth. You want consultants who has dirtied their hands, scoped out projects before.
The projects scoped out will serve as 'low hanging fruits' to convince more stakeholders onboard the data science train. With something tangible and relevant (since its the businesses' data), more buy-in is gotten and from there we can create a virtuous cycle to build up the capability. If you are interested to find out more, check out my other blog post here.
Doing data science and building capabilities up is never easy because the field is very new and thus knowledge and experience are very limited, not forgetting the many myths that are 'swirling' out there like doing Data Science is expensive (which need not be). To make things worst, there are many "talents" out there which masquerade themselves as experts but if you manage to dig deeper, they just have a nice facade. Building data science capabilities is not easy and there are many factors involved. Your organization definitely need someone experience to assist. Where to find them? Go to the local tech community and ask around, seek their opinion over coffee, lunch or dinner. That will definitely be a good start. :)