Broadly, the 5 ways to make money with data science are to:
- Get a data science job
- Consult on data science projects
- Build a tool for external consumption that leverages data science
- Build a tool for your own consumption that leverages data science
There’s no objective way to define “best”, as it varies by person. I’ll step through each way to make money, and break down how it rates on three main axes:
- Ease of doing
- Amount of control
- Moneymaking potential
This breakdown should let you make your own decision on which one is “best”.
1.Get a data science job
This can involve a significant investment in learning and interviewing. It often takes 1–2 years to learn enough data science to get one of the more desirable data science jobs. The monetary rewards can be high, but in most cases will be around what a top-tier software engineer makes.
This can be an easy option if you land at a large company with low expectations, but can be very hard if you’re in a high visibility position.
There often isn’t a lot of control over your work, although this varies by company. In a smaller startup you might be working longer hours, but in a bigger company, you might be doing less interesting work.
This is the option most people choose, and it’s a good default.
2.Consult on data science projects
This is more difficult than getting a data science job, simply because you have to learn, then put in a lot of work to build your profile and authority. It’s also a lot of work to constantly build your portfolio and gather good reviews so you can up your rates.
Although the eventual financial rewards can be high, they’re about on par with the top-paying data science jobs. The big advantage here is more control and freedom. You can pick your clients and set your hours. The downside is that clients may expect ongoing maintenance, and you’ll have to constantly manage existing clients while finding new ones.
This is a good option after you’ve had a job, and have a network of contacts who you can ask for consulting work.
3.Build a tool for external consumption that leverages data science
This generally manifests as starting a company. An example would be a tool that analyzes a company’s website traffic, and tells them what to optimize on their site. Your goal would be to charge for this tool, and get revenue.
The initial effort is very high, and you won’t be paid a lot. You’ll probably want some money saved away before doing this. Although the eventual rewards can be high, it’s no guarantee, and it can take years.
The benefit is that you get a lot of control over what you’re doing, and you get to build your vision. Even still, you’re still accountable to customers.
This is a good option once you’ve had a data science job, and have a good idea of the problems in the industry.
4.Build a tool for your own consumption that leverages data science
An example of this would be creating a tool that automatically buys stocks low and sells high, or a tool that tells you what houses to buy so you can flip them for a profit. This can be a nice way to make money, particularly if you find a good niche.
It can be very hard to identify that niche, though, so it usually takes a lot of effort to find and tweak. It also requires a good amount of upfront money, as you’ll usually need to spend money upfront, then see it returned later.
There is a lot of control if you choose this option. As long as you’re making enough money, you aren’t accountable to anyone, and can do whatever you want with your time.
This is a good option once you have some money saved, and understand problems that could be solved with data science.
As I’ve done with Dataquest, and others have done on Udemy, or with writing their own books, teaching data science is another way to make money. In order to teach, you’ll need to build up authority and credibility, so you’ll probably need to have a job or consult beforehand.
It also has a lot of the upfront risks of a startup in that you won’t make much money for a while, as you refine your curriculum, and find the right audience.
There is a good amount of control here, as you choose how you teach, but you’re also accountable to students, and want to see them succeed.
I’d recommend this after you have a data science job, and only if you enjoy teaching.
6.The bottom line
There are quite a few ways to make money with data science, but they all involve good amounts of time investment, both upfront and ongoing. I’d think hard about what kind of lifestyle and income you want, then pick accordingly.