Data SaaS

With the recent explosion of hype around data science, there has been an explosion of courses and offers that promise to transform anybody into a data scientist. Is it possible, or, differently put, is it science or science fiction?

If you ever landed on a page where Facebook lists its Software Engineering openings, then you can see a common theme in the requirements listed for a Core Data Science position:

  • Ph.D. in a relevant technical field, or 4+ years experience in a relevant role.

Want to try Google instead? Besides your degree you also need at least 2 years of experience as a minimum requirement!

  • MS or other advanced degree in an analytical field [...] or equivalent practical experience;
  • 2 years of experience in graph or network theory;
  • 2 years of experience as an analyst or in an analytical role [...].

What about Amazon? Hmm, preferred qualifications are kind of high there as well!

  • PhD degree in a quantitative field.

You may think that I'm cheating by only mentioning A-players in the market. But try finding a good job where you don't need either a MS in a relevant field or some years of field experience.

Then why do we find courses (I'll try not to reveal who's offering them) that promise after finishing them you'll

Speak the language of a Data Scientist in terms of statistics, data transformation, tricks (!), algorithms and theory!

all only after a 4 days course? I have another theory. And this was corroborated when I looked at the description of data science courses from more prestigious institutions, like for example Coursera: after an investment of 12 hours/week for eight weeks (equivalent to 12 full days) only promises that course promises you only this:

Introduce yourself to the basics of data science.

So what is going on here?

Science as a scam (not your ordinary SaaS)

Often, these miraculous courses appear offered by consultancy firms. These firms also sell data science services, so I imagine interactions such as:

Client: Hey, we would love if one of our analyst would be able to do some machine learning magic, do you have anything for that?

Consultancy firm: Of course! He should follow our "Become a Data Science ninja in 4 days" course. It just costs a couple of grand.

Client: Great! I can't wait to have my ninja back.

The analyst takes the course, gets shown how Google/Facebook/Amazon are doing, and is given enough theory to make data science even more dark magic than before. Then the client asks analyst:

Client: Hey, how was the course?

Analyst: Good, really nice, I learned tons of exciting things.

Client: Great. Can you build a real-time recommender for our website?

Analyst: Ehr, no, I barely know what a recommender is; let alone building a real-time one.

Analyst: But the consultancy firm should be able to help us, those guys are rock solid!

Client: Ok, let me call them.


Client: Hi there, I was wondering if you could build us a real-time recommender system for our website. Is it possible?

Consultancy firm: Of course we can, but a real-time recommender system is so old-school [Translation: we don't know how to build one]. Let me tell you what you really want and how we can achieve that for a small daily fee. It may seem like a lot of money, but it will do wonders for the revenues of your company. You'll earn it back in a week.

Client: Cool, let's do that.

We all know how these things end: the client usually gets locked into a service provider and the analyst does not really learn anything. And that's because this business model, the services company's interests are reached when the analyst will not learn!

The GoDataDriven™ way

At GoDataDriven, we believe things should be different. We believe in short, to the point courses that do not over-promise. And that's also why carefully choose our training partners. We also want to empower the participants to apply what they've learnt.

Even more, we know miracles do not happen to aspiring data scientists, so we are always willing to help the client with training on the job. Still, we are reluctant to do so for long period of time: the median length of our projects is about 6 weeks. And in these six weeks we at least teach the ins and outs of the area we are working on. That is far from making you speak the language of a data scientist, but at least you will be able to continue on the right track once we leave. We lay a solid foundation, allowing you and your people to keep on growing.

This is without doubts harder (learning data science take years!) but the long term benefits can't be overstated. Do it the right way, do it the GoDataDriven™ way!

Stay up to date on the latest insights and best-practices by registering for the GoDataDriven newsletter.
Follow us for more of this