Because when I realized that Google AdWords (I know, Google Ads now) consisted of staring at wide tables of statistics all day long and getting paid for it, I fell in love (Weird, I know.
To each their own!).
But after five years of Digital Marketing I realized this wasn’t for me.
This wasn’t the thing that was going to make me giddy about going to work.
This isn’t the subject I can keep harping on about to people who may or may not have asked for it.
So I decided to take a break and start a Data Science MOOC to see if it’s as interesting as I thought it would be.
And, spoiler alert, it is.
So I decide to kick it up a notch and get certified in the Data Science Starter Program at the oldest engineering school in France.
“I have a plan…”“Ok.
So now you’re a Data scientist, Lai.
What was the point of your story?”Not quite, you skeptical reader.
Even with my certification in the bag, I decided to continue with my career by working as a Data Analyst at a consulting agency.
It’s somewhere between my Digital Marketing past and my Data Science future.
It makes sense.
It’s a steady path.
The certification gave me a steady basic understanding of Data Science projects and Python programming so I could go like this with my sales engineer: “So, Clara, I have a 2-to-4-year plan.
I’ll keep learning Python online while gaining knowledge in data projects by being a Data Analyst, and when I feel secure enough about my Python skills, I’ll be a Data Scientist.
”Since I’ve said that, I’ve already made a bit of adjustment by changing 2-to-4 years to 1-to-2 years.
But still, Dan Becker came around and crushed that plan with all the benevolence in the world.
Time for a change of strategy“On the spectrum, there are two extremes: Short term Sal and Long term Lee”Dan explained it this way: there are two types of approaches to get into Data Science.
On the one hand, you have short term Sal who gets just the knowledge they need to get a job, dives in, and then lands a job in Data Science, learning the rest while doing it.
On the other hand, you have long term Lee who learns and learns and learns and finally gets a job as Data Scientist.
That’s right, reader, that’s me.
And that’s what Dan said should be a “fair”career trajectory.
But let’s be real, Sal gets on the job sooner, gets hands-on experience quicker (meaning practical knowledge that Lee doesn’t get from text books), and earns himself a network in the field, something lonely Lee didn’t get from his learning hermit life.
So this is what really happens:At this point, I’m staring at the webinar, annoyed.
“Oh Dan, how dare you crush my dream like that!“ but Dan has thought of us, the poor Lees watching all the Sals attending Youtube’s live stream pointing their virtual finger at us and laughing in the comments section.
He has one piece of advice that might be simple for some of you, but has a large impact as I think about my approach.
“Find your passion project, work on it, publish about it, tweet about it.
”“Ok, hold on, Dan.
So far I only managed to copy a pre-prepared Titanic survival analysis, tinker with a few classification algorithms, and make a very mildly successful text mining project for my certification.
Do you really think I’m going to put any work of mine out there to be slaughtered by top notch data scientists?” At this point, I firmly believe that Dan reads minds because he then explains that yes, first projects may not be good but it’s ok, putting them out there might attract some constructive feedback that will make the next one better.
And that’s how a Lee becomes a Sal.
“Fine, Dan, fine.
Jeff, it’s time to take on a Kaggle challenge.
”Thank you, Jeff, for proof reading this article and for being the Sal to my Lee.
Follow UpTo this article, I would like to add Dan’s answer to the original post on LinkedIn.
Again, I greatly admire the spot on insight.
(Fangirling much, Lai?)Kaggle Competitions are a good way to learn.
Especially because you can try a problem, and then see how your approach compares to others (since so many people publish their work in easily accessible “kernel notebooks”).
But I don’t think they are a quick path towards a portfolio project to land you your first job (which I think is the real turning point for most people).
Here’s why: Competitions have a lot of skillful participants.
You need to score exceptionally well for employers to take notice, and (tautologically) most people will be average or below average.
Instead, I’d suggest finding a topic that’s interesting to you and doing some data exploration.
Search for relevant data, and explore it until you can write a nice blog post about it (and ideally present it at meetups).
It could be about news, a hobby of yours, whatever.
Make your project visual (potentially through graphs), communicate about it clearly.
Heck, maybe even brainstorm on whether you can leverage your digital marketing background to help it get more exposure.
But the idea is to do something creative.
This is more likely to get you noticed (that yellow scarf trick I mentioned yesterday), and more readers (or meetup attendees) will find it interesting.