SDS PODCAST EPISODE 89 WITH CHRIS DUTTON
Transcript of SDS PODCAST EPISODE 89 WITH CHRIS DUTTON
Kirill: This is episode number 35 89 with Best Selling Data Science
Instructor Chris Dutton.
(background music plays)
Welcome to the SuperDataScience podcast. My name is Kirill
Eremenko, data science coach and lifestyle entrepreneur.
And each week we bring you inspiring people and ideas to
help you build your successful career in data science.
Thanks for being here today and now let’s make the complex
simple.
(background music plays)
Hello everybody and welcome back to the SuperDataScience
podcast. Today we've got a very exciting guest, Chris Dutton
on the show. So if you haven't met or don't know about
Chris yet, then you are very likely to encounter him on
Udemy, a platform where I also teach. Chris is the top
instructor on Excel courses. And so what do we talk about
in this show? So first of all, just to make everybody feel
comfortable, we're going to answer the question, why data
scientists should still learn Excel. Because a lot of the time,
you hear comments that Excel is not a data science tool, and
data scientists should be using other tools, and so on, which
are fair in some cases, but there are some valuable benefits
of actually learning and knowing Excel, and we'll dig into
that. So if you've wondered the question, "do you need to
know Excel, and to what extent," this is going to be a great
podcast for you.
Also, Chris runs his own business. He's got a website,
excelmaven.com, and he's running both an education
business, and a consulting business, so he'll tell us all about
that and how he transitioned in his career from working in
an agency in Boston to doing his own thing, doing freelance
work, doing consulting work. So that will also be very, very
valuable. And plus we'll dig into some of the things around
how you can start into teaching if you like, how you can
explore different avenues of careers, and we'll get some of
Chris's thoughts on what is coming for data science.
So a very exciting podcast ahead, can't wait for you to check
it out. And without further ado, I bring to you Chris Dutton,
CEO of ExcelMaven and top data science instructor on
Udemy.
(background music plays)
Hello everybody and welcome to the SuperDataScience
podcast. Today I've got a special guest, my buddy instructor
from Udemy, Chris Dutton. Chris, welcome to the show.
How're you going today?
Chris: Chris: Thank you very much. Excited to be here.
Kirill: Awesome. And where are you calling from today?
Chris: I am calling from Boston, Massachusetts.
Kirill: Nice, nice. How's the weather? It's summer, right? So it must
be good weather in Boston.
Chris: It is. It's actually a beautiful day. It's got 76 degrees. No
complaints.
Kirill: Fantastic. I've got this thing with Boston, because when I
was there in May, it was raining and cold and everything, so
whenever I get a guest from Boston, I'm always like, "how's
the weather?" Because for me, it's like London, pretty much.
Chris: It happens.
Kirill: Pretty lucky day today. Alright, so Chris, it's fantastic that
you're on the show. You're the best selling instructor on
Udemy for courses relating to some of the top Microsoft
products, such as Excel, PowerBI, and others. So tell us a
bit about your background. How did you get into teaching on
Udemy?
Chris: Sure. So I'm a marketing analytics consultant, and also the
Founder of Excel Maven, which is all about providing online
training and consulting with a focus on Excel and PowerBI.
So going back to the college years, I studied quantitative
economics and multimedia art, so kind of a strange balance.
A lot of the left-brain-right-brain balance. After college, I was
looking for a role in analytics, found a job in the strategy
and analytics group at a large Boston-based advertising
agency, which was kind of the perfect fit for me because we
were doing really interesting data science and analytical
work, but in the context of this really creative marketing
field. So I got to work with a really wide range of clients
across just about every vertical: automotive, healthcare, e-
commerce, insurance. And my role in that job was to
essentially help them develop measurement plans, track
performance, better understand their customers, and
ultimately optimise how they're spending their marketing
budgets.
So it was in that role that I really started to get exposure to
Excel, and specifically some of the more advanced Excel
tools and techniques. After a couple of years there, I pretty
quickly became the Excel guy at the office, so three or four
times a week I would have co-workers coming by with their
Excel challenges, their toughest problems they were trying to
solve. And, you know, some people would be annoyed by
that, but I actually loved it because it was a chance for me to
kind of continue upping my game and continue trying to
solve progressively more challenging problems with Excel.
That eventually developed into producing and teaching the
advanced Excel training materials for the incoming analysts’
classes at the agency and that was really my first taste of
teaching analytics and I loved it. I don’t know what it was, I
think it was just that excitement of seeing people’s eyes light
up when they get a complex function to operate or build
their first tool. It was just really inspiring and I found that I
really loved teaching and had a lot of fun with it. So, fast
forward about six years working with the agency, I decided
to go out on my own, start my own consulting business, and
at the same time offer this Excel training on a broader scale
and that’s what brought me to Udemy and this whole e-
learning world.
Kirill: I’m glad you mentioned your website, Excel Maven. It’s
actually very well-made. Congratulations on that. I was
looking at it today and it’s a very, very professional website.
I’m very impressed. Did you code that yourself? Or did you
get someone else to help you out?
Chris: No, that was one of the areas that I outsourced to a friend of
mine who I’ve done some work with in the past who does
some really, really high quality work, so I’ve been really
happy with it as well.
Kirill: Yeah, it looks very good. I can really relate to your story, the
whole being the Excel guy in the office. When I was at
Deloitte, I wouldn’t say—like, I knew the basics of Excel, but
at some point I set myself a goal to know it really well
because in consulting, you use Excel for pretty much
everything, especially at the start of the project. At some
point, everybody—not in my department because I was in
the data science department and everybody knows Excel
very well there, but people from other departments were
coming up to me and saying, “How do you do this? How do
you do that?”
You know, I remember once when one lady, she was using
the keyboard to move up and down through the cells, but it
was actually moving the whole sheet and she was like, “How
do you undo that?” And I looked everywhere and you just
have to unclick scroll lock on the keyboard and it was so
funny, that was like a 5-second thing but you needed to
know that. I have an interesting question for you, for me it’s
always been the metric of how well I know Excel. What
percentage of your work in Excel can you do without
touching your mouse?
Chris: That’s a great question. The keyboard shortcuts and the Alt
key tips are huge timesavers, so I actually have a whole
section dedicated to using shortcuts to work efficiently in
Excel. As far as the percentage goes, if you get really good at
it, I would say you can probably do 75%-80% of your work
without touching a mouse.
Kirill: Yeah, and that for me was the main selling point when I
personally started learning Excel. I remember I was in a
project somewhere in the middle of nowhere in Mackay in
Queensland, it’s the middle of the desert pretty much, and
these consultants flew in from America and one of them was
doing everything on his computer, on Excel just by using the
keyboard. I was so impressed that I set myself a goal, “I’m
going to be like that.” It took me like a year to master that,
even though probably if I put in more effort I would have
mastered it earlier.
But then one day I was flying on the plane and doing some
work on my laptop and the person sitting next to me was
like, “Oh, wow, you can use Excel without your mouse,” and
at that point I was really happy about it. Yeah, so it’s really
cool that you teach that in your course.
Chris: Yeah, for sure. It’s definitely a little tricky to get used to, but
it’s worth the effort for sure.
Kirill: Yeah, definitely. All right, I’m keeping the listeners, or we are
keeping the listeners here, in a bit of a suspense because
this is a podcast – and I warned you about this question –
this is a podcast for careers in data science and a lot of the
time in data science literature, in data science courses, in
data science conversations, you hear that Excel is not a
great data science tool, like data scientists should not be
using Excel.
You know, the whole point of that stems from the fact—like
most of it, there’s lots of arguments for that, but I think the
main one is that Excel combines data plus function in the
same space. In SQL, for instance, you can’t put a formula
inside a cell, you have data separate and you have the
command separate. In Excel they combine and that can get
confusing if you then try to do more advanced data science.
So let’s break that myth. Let’s completely destroy it and let’s
tell our listeners, give our listeners a reason why Excel is
still important to know for data science.
Chris: Yeah, that’s a really good point and I’m happy to talk about
it. I think you’re right, Excel does get kind of a bad rep,
especially in the data science field. Just to preface, I would
say that there are certain projects where Excel is absolutely
not the right fit for the role, but others where it is—a lot of
times what I see is this issue of people not knowing what
they don’t know, it’s people who are very familiar with a very
small fraction of Excel’s functionality and in their opinion
that’s kind of the entire program so they don’t really expose
themselves to some of the more interesting advanced
capabilities of the program.
The other thing I’ll say is that a lot of people who share that
sentiment about Excel have that perspective because they’re
trying to use Excel for the wrong purposes. People who are
trying to use Excel as a database tool or a data storage tool
are going to run into problems because that’s really not what
Excel is built for. That said, I have been exploring some of
the newer business intelligence tools that Microsoft has been
rolling out, things like Power Query and Power Pivot, and
more recently Power BI, that are really starting to break
down a lot of these walls that have kind of prevented people
from using Excel for more serious data science projects or
things that involve larger datasets.
I’ll give you an example. Just yesterday, I loaded up a raw
dataset and connected to a table with 30 million rows,
stored it in Power Query within Excel itself, it’s a brand new
compression engine where you can store way more data
than you ever could before, and then I used tools like Power
Query and Power Pivot to basically create relationships
between tables, build data models, and then develop
completely custom visualizations on top of that model. That
kind of stuff, it’s things that you really couldn’t do 3 or 4
years ago in Excel, but you really can now. So, it’s opening
up a lot of doors and a whole new world of Excel capabilities.
Kirill: Okay. That’s really cool. So, Microsoft is bringing on all
these capabilities and even new tools such as Power BI. I
totally agree with that. How about this question: Why would
somebody learn Excel, being a data scientist, beyond the
basics if they can do all those same things in tools such as R
and SQL and Python and so on? Are there any other reasons
for people to take on Excel? You know, why would it, in
some cases, might even be better or more advantageous to
know Excel instead of those tools or maybe in addition to
those other datasets?
Chris: I think you bring up a good point which is, ‘in addition to
those tools.’ I would never say that Excel is a replacement
for R or Python. I would say that those tools do a really nice
job supplementing each other. I also recommend for newer
data scientists or people who are first getting into this
analytics world, Excel is a really great way to learn and
master the fundamentals, so there’s something about seeing
the data in front of you and seeing the output as you
manipulate, transform and shape that data in a way that’s a
little bit less tangible for programming languages and tools
like R and Python, for instance. I think that’s one benefit of
Excel.
The other thing that I found out as I’ve started to push the
limits of Excel further and further is that the level of
customization with Excel is actually pretty outstanding,
especially once you start getting into the data visualization
side of things. You know, I’ve used Tableau, I’ve used Domo,
I’ve used Custom R Visuals, and honestly, I often go back to
Excel strictly because I can personalize and customize my
visuals exactly how I choose and can kind of hack together
these interesting visualizations that quite honestly I wouldn’t
be able to build elsewhere.
Kirill: Okay, very interesting. Actually, one of our other guests on
the podcast, previous guest Nadieh Bremer, she said that
she had a similar reason why in terms of visualization she
moved in the opposite direction, to move to a more advanced
tool, which is D3, because she wasn’t able to make the
custom visuals that she wanted in Tableau, etc. That’s a
great point.
And also I like what you mentioned, you know, ‘in addition
and also starting out.’ I completely agree with that. For
somebody who’s just starting out into the field of data
science, the whole notion of what Chris just mentioned, of
seeing what you’re doing with the data is invaluable. Like, it
might not be a tool that can handle any data science
problem in the world, but at the same time, Excel is really
good for seeing what you’re doing with the data and
therefore understanding how to speak the data language
better on an intuitive level. That, for me personally, has been
an invaluable journey that I went through to get through the
whole space of data science to where I am now. That’s some
very good insights. Thank you so much. Tell us about how
many courses you have total on Excel.
Chris: Let’s see. I have three courses on Udemy. I launched the first
one just under two years ago, and those are kind of the
comprehensive full-scale courses. I’ve got one covering
formulas and functions, second one covering data
visualization, charts and graphs, and the third is data
analysis with pivot tables. And I’ve got three others on
Lynda.com and LinkedIn Learning. Those are more project-
focused, shorter, more niche courses on that platform, so six
total.
Kirill: Cool. Congratulations on getting on Lynda.com, I heard they
have very stringent selection criteria for the instructors that
they select. Do you know that LinkedIn Learning—they do
courses, videos on whole flights between Europe and
Australia? Like, I can watch these. Did you know about
that?
Chris: No, I didn’t.
Kirill: Yeah, I’ll check next time on the flight. I’ll check if I can see
your courses there.
Chris: Sounds good.
Kirill: That would be really cool. Okay, I want to ask you this
question. Your lowest rating for any of your courses on
Udemy is 4.7 out of 5. This is incredible, this is with like
thousands of reviews. How do you manage to have such high
student satisfaction? What is your secret sauce?
Chris: Oh, that’s a good question. Going back to when I first
published course number one, obviously I did the research, I
looked at supply and demand for Excel courses on Udemy,
and even two years ago it was an extremely competitive
category. You know, if you search Excel on a platform like
Udemy, there’s something like 40 or 50 pages of courses. So,
for me, trying to break into a category like that as an
instructor with no current student base, no following to
speak of, no e-mail list, I knew that I had a pretty steep
mountain to climb ahead of me.
So, from day one, it was really just about producing the
highest quality content that I possibly could, including
content that you really can’t find in any another course, so
things like custom datasets, really interesting unique
examples. I like to use kind of fun and interesting datasets.
In my pivot table course, for instance, there's a whole
section at the end with just different types of case studies to
take what you’ve learned in the course and then apply it in
all sorts of different contexts.
So I’ve got a dataset on San Diego burrito ratings, I’ve got a
dataset on all shark attack records over the last 100 years,
I’ve got salary data, Major League Baseball data, social data,
all different things. I think that just makes it more
interesting and it makes people want to learn and want to
stay engaged.
The last thing I’ll note on that is I take a much more serious
focus on student engagement and interaction and I think a
lot of instructors do. I take pride in the level of one-on-one
attention that I give to my students. I’m there answering
every single message that people are sending me, I’m offering
support and one-on-one guidance for anyone who posts to
the course discussion board. And I think that just goes a
long way over time and I think it helps students trust me
and it helps me offer more value as an instructor than
anyone else.
Kirill: Okay. And what are some of the most common questions
that you get, like the most common question that you get
from the students?
Chris: “Can I take this course on a Mac?”
Kirill: (Laughs) And what’s the answer?
Chris: Most of the content, yes, although there are some caveats
because the user experience is frustratingly different across
platforms in certain cases. With tables and charts and
graphs specifically.
Kirill: Yeah, I can totally imagine that. We’ve had similar questions
on our other courses. It depends on the platform, but you
can still have Excel on Mac, so I think it’s a worthwhile thing
to learn it.
Chris: Absolutely.
Kirill: You know, I’ve never taken your courses before, but this is
what I’m gathering from what you’re saying and I’m actually
looking at your course right now. Even if somebody has
Excel skills, they will still get a lot of value from your
courses not only because they’ll get the tips and hacks that
they’ve missed out on and maybe some shortcuts and so on
and some ways to do things that they haven’t thought of
before, but also they will get these case studies, these great
examples of how to apply Excel to the real world. Given that,
where do you think somebody should start – out of your
three courses that are available on Udemy – where would
somebody get started?
Chris: I usually recommend the formulas and functions course
first. That was the first one that I produced and I think it’s
kind of a good starting point. And then there, depending
on—if you’re more interested in the analytics side of things,
I’d recommend the pivot table course next. If you’re more
interested in data visualization, I’d say the charts and
graphs course as a good follow-up. But honestly, you can
take them in any order and within each course – to your
point – I try to create content that’s appropriate for students
of all skill levels. So I do try to cover the fundamentals and
basics relatively quickly early on and then kind of progress
into more and more sophisticated and complex examples
and case studies.
Kirill: That’s really cool. I’m looking at your Excel formulas and
functions course and I can see you have array formulas and
that’s a really powerful tool. For me that was one of the
latest things I’ve learned when I was at Deloitte and it really
changed a lot in terms of working with clients. It was really
helpful. Yeah, that’s really cool that you’re going through
some very advanced topics there.
Okay, we’ve talked about some of your courses, that’s great,
and some of your teaching methodologies. For somebody
listening who wants to start into maybe teaching something
online themselves, maybe not as advanced as you having a
whole business in that space, but maybe just giving back to
the world and contributing and explaining some sort of skill
that they have or they’ve developed, in a certain tool or a
methodology or something relating to data and analytics.
What would you say is the best place to start? How do you
take that first step?
Chris: I think, number one, you have to ask yourself two questions.
Number one, am I truly an expert and do I feel like an expert
in this topic? And number two, do I love to teach? If you
don’t answer yes to both of those questions, it’s just going to
be a struggle and an absolute grind. So that’s why I really
only teach Excel and now Power BI courses because that
really is my true area of expertise. You know, I have working
knowledge in other tools and programming languages, but
my opinion is that if you want to be the best teacher and
instructor that you can be, you really need that deep, deep
expertise in the topic that you’re teaching. So, if you do
answer yes to both of those questions, the next step would
be to evaluate the landscape and go on Udemy and type in
some search terms related to that topic and see what the
competitive landscape looks like, make sure there’s demand
for that topic, and then just lay out a roadmap for building
content and starting to produce your course.
Kirill: Really interesting. I can see your point, but I’ll have to
disagree with you on that first one about the expert. I would
say that if you have that working knowledge, you can still
start to teach. I just want to encourage our students here as
well that you can still start to teach, you can start a blog,
you can start something basic, something like explaining
things on maybe YouTube or in a blog or so on, and still give
back to the community.
My personal opinion – again, different opinions, I totally
understand – is that you don’t have to be an expert to teach
something, but at the same time I can see where you’re
coming from with this and you have that integrity that you
have to give the absolute best. And to your point, you have
4.7 stars on all of your courses, so you definitely are the
expert in all of those areas and you obviously love to teach
so that stands as a testament to that.
But at the same time, to encourage our students—in my
personal opinion, I’ve taught subjects where I’m not an
expert on something, but I learn something for myself and
while I’m learning it it’s just easier for me to learn it even
better if I teach it to other people. So that’s also an
approach. Would you agree with that kind of sentiment?
Chris: Oh, definitely. And I’ve had this conversation with a number
of other instructors too, and I totally appreciate your
perspective and viewpoint as well as one of the top overall
Udemy instructors yourself. I’ll make a couple of points
about that. For one, I think you bring up a great point about
learning and teaching yourself as you’re teaching these
courses. I became a much stronger Excel user and expert
through the process of teaching so you’re right, I don’t want
to discourage people from starting if they don’t feel like
they’re at this ultimate level of expertise.
The other point that you mentioned was YouTube. I think
that is such a great platform for people to test the water on a
slightly smaller, more informal scale. And that’s a great
place where you don’t need tons of technical equipment or
recording gear, you don’t need a huge 6-hour curriculum for
a full-scale course. You can just pick little topics or
individual lectures and just throw them up on YouTube and
see what kind of response you get. I think that’s a really
awesome testbed for people who are interested in eventually
teaching larger courses.
Kirill: Yeah, I totally agree. We’ve tested out a few things that way
as well. For instance, somebody could just google Sankey
diagram. You know, before creating a course about these
things, we tested these things out and you can really see if
there’s demand or not for certain topics. Thank you for that.
That was an interesting discussion. Let’s move on to
something else that you’re very passionate about, and that is
consulting. Tell us a bit more about the consulting side of
your business.
Chris: Sure. Basically, my area of expertise is marketing analytics
consulting. Basically, there are thousands of companies out
there that are sitting on huge amounts of data with no idea
what do with it. My focus as a consultant is really just
helping them collect, transform and visualize that data, and
then, most importantly, actually translate it into something
meaningful. You know, I started at the advertising agency in
Boston, and after about six years, I kind of broke off and
decided to do this on my own, similar types of work working
with other companies individually, working with other ad
agencies, but generally speaking, the majority of my projects
involve helping companies blend together data across a
number of different sources, help them define those
relationships between their sources, and then building the
tools and dashboards to help them explore and visualize
that data and ultimately optimize the way that they’re
spending their marketing budgets.
Kirill: Fantastic. And do you educate the teams as well?
Chris: I do. That’s another part of my consulting role, is obviously
leveraging the training content that I’ve developed on Udemy
and through the Excel Maven stuff, offering that as kind of a
service as a consultant as well.
Kirill: Okay, I understand. Moving from your previous career—this
is a question which some of our listeners will find
interesting, who are contemplating whether or not to stay in
the space of being employed or moving into freelancing and
having their own business. If you’re one of those students,
yes, I’m talking exactly about you. So how would you
describe the difference between when you were working in
the agency in Boston, and now that you have your own
business? What are the pros and cons and why would
somebody pick one over the other?
Chris: Yeah. It’s not the right fit for everyone, but for me it’s been
pretty liberating. I haven’t looked back since. I’ve been doing
my own independent consulting for about 3 years now, and
the biggest pro I think is really just flexibility in terms of
creating your own schedule, managing your own time as you
see fit, and also the beauty of consulting on your own when
you really have the reins is the ability and the option to get
exposure to a really, really broad range of projects.
Generally, people who are working a traditional full-time job
tend to be on a relatively narrow path or have a relatively
narrow scope of work; whereas consulting kind of opens up
that door and allows you to have a lot more flexibility to
explore many, many different types of projects. That’s been
far and away the number one pro for me.
As far as cons, you lose a little bit of stability, you lose a
little bit of predictability of that 9 to 5 salaried role which
some might see as a downside. You do have to sell yourself a
bit more, you’re constantly looking for new opportunities,
selling your services to new clients, and that requires a
different skillset that not everyone has, which again is why I
say that this path is not for everyone. But once you establish
yourself, and once you get some clients, and really start
getting exposure to some of that work, and start adding
value as a consultant, it really is a wonderful thing – at least
in my experience.
Kirill: Okay. And tell us a bit more—I can feel, I can sense that a
lot of people who are thinking about this, right now they
have this question, “How do you get the client?” How do you
go out there and find the clients who are going to pay you for
your work, for your consulting engagement?
Chris: For me, I was fortunate because I made a lot of connections
during my six years working in client services, both on the
client side and among colleagues who I’ve worked with here
in Boston at the agency. So, a lot of those existing
relationship helped turn into some of my initial contracts
and projects. That said, that’s not the case for everyone.
The other thing that’s been really helpful for me has been
kind of getting my face and my work out there publicly.
Honestly, that’s a big reason why I decided to become an
instructor. You know, it validates my skillset, it provides
unbiased, objective proof of my expertise, you know, just
looking at the student counts and the student reviews and
the content that I have published out there. That has
actually created a number of new relationships that have
then turned into consulting projects as well. So, the online
learning and the teaching has really been a nice—almost like
a lead gen source for the consulting work, and vice versa.
Kirill: Yeah. And to your point, more than anything, it’s a
testament to your skills. If you can teach something, it’s
obvious you can perform it and run a consulting project in
that space. It’s a no-brainer to hire you at that point.
Chris: Yeah, exactly. It’s one thing to put a tool or a language on
your resume, and it’s another to prove that you can actually
teach it.
Kirill: Exactly. It’s a great inspiration for those listening. Maybe
some are thinking of going into consulting, and going on
your own. And if you already have a solid plan on how you’re
going to keep that cash flow coming in, then go for it when
you feel confident. And if you don’t, then this is a good
solution on how to build that plan. You know, start building
an online presence, whether it’s through Udemy or through
a blog, or through YouTube, or something, so you have
something to stand for you, so that it’s not you going around
saying, “Hey, do you want me to do some consulting work?”
but people are coming to you because you are the expert or
one of the experts or one of the influencers or teachers in the
space.
Yeah, that’s a great place to start. Thanks a lot for the quick
rundown on your consulting business. What are your plans
going forward for your business? How do you plan on
expanding and growing it, if you can share that with us?
Chris: Yeah. So, you know, I really would just consider my journey
as an online instructor just starting. Like I said, I just got
into this space less than two years ago, so for one, I’m really
interested to try to build more content, really see how far
that path can take me. To be honest, instructors like you
and Phil Ebiner and Mark Price, these are guys who’ve built
these pretty impressive followings and have really proven the
potential that there is in that space.
Kirill: Thanks.
Chris: That’s exciting, you know, the exposure that you can get as
a top instructor on a platform like Udemy is phenomenal.
The other benefit I think is that it really keeps you honest by
forcing you to constantly keep learning to stay relevant. So
I’m excited to kind of continue pushing forward in that path
and producing new courses and potentially partnering up
with some other instructors to see how far I can take that
route.
And, you know, also the business model that I’ve created
with Excel Maven, of which the online self-paced course is
one component of it – the other components being on-site
group training and project consulting – that business model
is really starting to prove itself. So, my other focus is
eventually trying to scale things up on the Excel Maven side
of things, and hopefully broaden the focus, find some
partners, and potentially expand to a broader range of
analytics resources.
Kirill: That’s some really solid plans. And with the expansion, just
out of curiosity, are you starting to hire people, are you
starting to build a team, or are you planning to do this on
your own for some time?
Chris: I think I’m at the point now where I will be looking for some
partnerships. You know, going back to the whole concept
of—for me, I’m really only comfortable teaching what I feel
I’m an expert in, so rather than me trying to become the
teacher for all other courses, I think I would try to identify
the experts and build some partnerships and start to grow.
Kirill: Oh, yeah. Yeah, totally, but I mean more the administration
side of your business, because then you don’t have to take
care of the courses, the website and everything like that. So
are you planning on getting some admin staff on board?
Chris: Yeah, that’s the plan. I’ve been doing a pretty bad job about
outsourcing some of those roles, to be totally honest. I’ve
really tried to wear too many hats up to this point. So, yeah,
I’m definitely starting to look for people to help support other
aspects of the business to help things grow.
Kirill: Yeah, I’ve been there, I’ve done that. I got to the point where
it was just too much. I was answering close to maybe 30
questions per week on Udemy – which doesn’t sound like a
lot, but in addition to all these other things, it was just
driving me crazy so, yeah, at some point I had the same
realization and—once you start adding people to your
business who are helping you and who are assisting you in
your goals and mission, you get to focus on the things you
actually love, and it’s a great feeling.
Chris: Yeah, definitely.
Kirill: Okay, we’ve talked about learning and teaching at the same
time, you have these plans for growing your business. What
are your plans for learning new stuff? What are you excited
about learning yourself in the coming months or maybe a
year or so?
Chris: I’m really, really excited with some of the stuff that Microsoft
is coming out with in their BI stack. I referenced some of
those newer tools earlier: Power Query, Power Pivot and
Power BI. Really just in the past year or so, I’ve been
integrating those tools more and more into the work that I’m
doing for my clients, and I’ve been incredibly impressed by
the capabilities of those tools. I’ll be teaching a Power
Pivot/Power Query course next, followed by a Power BI
course. I’m looking for opportunities to practice and learn
those tools every chance I get. So, it’s really exciting stuff
that Microsoft is doing in the BI and the data science world.
Kirill: Fantastic. That sounds really amazing. I can attest to that.
I’ve worked with Power BI and I also have a course with
Power BI and I have seen how Power BI has grown. Like, in
the Gartner Report, it was somewhere in the middle of the
quadrant a year ago, or a year and a half ago, and then last
time they released it in February it’s now at the top, near
Tableau. They’re releasing updates literally every month,
major updates as well, so they’re really focusing on this
analytics space, and indeed it’s very exciting to see what
they’re coming up with.
Chris: Yeah, absolutely.
Kirill: Okay, thanks a lot for sharing that. Let’s do some rapid fire
questions about your career. Are you ready for this?
Chris: Sure.
Kirill: Okay. What’s been the biggest challenge you’ve ever faced in
your career?
Chris: Oh, man. I’m going to give you a kind of general answer to
this, but in general, the biggest challenge that I’ve had
personally is just trying to keep up with everything. Data
science is one of those fields where it’s really easy to feel
inadequate. You know, you ask yourself, “Am I a slacker if I
don’t know both Python and R? Am I falling behind if I
haven’t learned TensorFlow yet?” You’ve got people throwing
around these acronyms and these tools left and right, you’ve
got new things showing up what feels like every single day,
so honestly, I think one of the biggest challenges of working
in this field is, a) trying to keep up with what’s relevant, and
b) reminding yourself that it’s okay to ignore some of the
stuff that isn’t, which can be easier said than done, but at
the end of the day, no one has the capacity to learn all of it.
So it really comes down to picking your battles, which has
certainly been a challenge for me.
Kirill: I totally agree. I think that choice has to be guided by
everybody’s—their own passion. Yes, there are lots of tools,
but don’t just get carried away running after the latest,
greatest, newest, biggest thing if your passion lies
somewhere else. There is always going to be space in this
field of data science. There is always going to be space for
you to realize your passion if you’ve really focused on it.
Chris: Right. And you don’t want to end up learning ten tools at a
very shallow level, as opposed to one or two really well.
Kirill: Interesting. Yeah. Okay, thank you. Next one is, what is a
recent win that you can share with us that you’ve had in
your role, something that you’re proud of?
Chris: One project in particular has stuck with me. And when you
say recent, this one was a few years ago, so not the most
recent, but—
Kirill: That’s totally cool.
Chris: It was probably my favourite project that I’ve worked on,
which was actually back in college. I’m a big baseball fan as
a player and a fan of the sport and the Red Sox, and also a
fan of the data and the statistics behind the game. So, back
in college, I actually started a group called “Baseball
Analysis at Tufts.” I went to university outside the city, and
our goal as a group was basically to come up with really
interesting questions and hypotheses about the game of
baseball and then try to answer them with econometrics and
statistical models and data analysis.
And one of the projects that really took off for us was a
project to try to quantify luck, so how lucky was a given
hitter in a specific season. And the way we did that was we
essentially tried to quantify and identify every single factor
that contributes to a hitter’s batting average, so things like
their power, their speed score, how well they can spread the
ball across all fields.
So we took all of these individual elements, these
independent variables, and we fed them into this regression
model that essentially would spit out an expected batting
average. And what we were able to do then is look at actual
hitter’s performance, compare it against their model’s output
and call the delta, something called the ‘luck factor’.
So, that was really interesting and the best part was we were
able to take a given season’s worth of data and identify the
list of players who outperformed the model by the widest
margin, those were the ‘lucky’ ones, and the players who
underperformed their model by the widest margin, those
were the ‘unlucky’ ones, and then track how their
performance changed year over year.
And what we found was that it was actually remarkably
predictive of which hitters would improve the next year and
which hitters would regress. It was really awesome to see
and it ended up getting a good amount of coverage in some
baseball blogs and websites, and there was a feature in the
‘New York Times’ about it, which was very exciting.
That was a really meaningful project to me: a) because it was
just a lot of fun, and b) it really made me love analytics and
really appreciate its ability to expose these patterns and
trends and stories in the data that you otherwise never
would have seen. It almost feels like becoming fluent in a
new language, except the language is data. So that was a
really meaningful project, and really I would say one of the
biggest influences in guiding my career into the analytics
and data science space.
Kirill: That’s so cool, such a cool story. I’m burning to find out—so,
players who had higher luck, in the next season they
dropped down, and who had lower luck, they usually went
up. Is that correct?
Chris: Yeah. And as a follow-up to the project, we actually
partnered with a really popular baseball researcher who
compared our model against seven or eight other predictors,
other predictive tools, and at the end of the day ours was the
winner by a pretty wide margin, which was really satisfying
to see.
Kirill: So you might even say your model was lucky?
Chris: Oh, yeah. There you go. (Laughs)
Kirill: So how old were you then?
Chris: I was 21 at that time.
Kirill: 21? That’s really impressive. It’s a great example. I’ve also
had stories like that in my life where I was passionate about
physics and I would go do a physics project on my own and
build this magnetic thing that I thought was the first one in
the world or do something in programming, create this
programming algorithm with 10,000 lines of code just in my
free time. I actually totally agree with you that these are the
projects that—I wouldn’t find a better way to put it—that
shape your career. These are the projects that shape your
future. It doesn’t matter what you really do at work. That’s
all great and that’s what you’re told to do, but when you’re
really passionate about something, and you go and you
spend your free time, your blood, sweat and tears on that,
because time ultimately is the most valuable resource we
have; if you’re spending your time on something, it means
you have to love that thing so much.
And when you spend a lot of your free time on something
and you really, really work on it and you get that final result
which you’re working towards, that, my friends, really
shapes where you’re going to go in life. So if you haven’t
done one of those projects—I’m sure everybody has at some
point in their life—but if you haven’t done one recently, I
would highly encourage you to go and do that and find some
time to invest into something that you’re passionate about.
Pick some problem, pick some challenge and solve it. No
matter how long it takes, no matter how complex it is, you
will be super satisfied at the end and it will reveal to you
more what your passion is all about and how you can dig
deeper into it. Thank you so much, Chris, for sharing that.
That’s a great testament to it.
Chris: Of course, couldn’t agree more.
Kirill: Okay, so next one is, what is your one most favourite thing
about working in the space of data? What excites you the
most?
Chris: Data visualization. Yeah, I love data viz. It’s always been my
favourite part of the job. I think there’s just something really
powerful about turning a mountain of raw and unstructured
data into something beautiful. And, more importantly, into
something that has meaning and insight and can actually
guide decisions. I think data viz is an underappreciated skill,
to be honest, and I think it’s one that tends to be
surprisingly uncommon among data scientists. But I love
data viz, I love getting creative with it. I love constantly
looking for new and interesting ways to present my data.
Kirill: Fantastic. Thank you, I totally agree with that. It’s a very,
very exciting and powerful skill to have in your arsenal. And
slowly wrapping up the show, a very philosophical question
which I’d like to get your opinion on: Where do you think the
field of data science is going, and what should our listeners
prepare for so that they’re ready for the future?
Chris: Yeah, it’s a great question. For one, I think we’re going to
start to see data play a much more critical role in industries
and in scenarios that we haven’t considered to be very data-
driven to this point. We have things like sensor generated
data, IoT, wearable technology, and all of those things are
creating data in places that it hasn’t existed in the past.
That’s really exciting to me and I think that’s going to lead to
some really fascinating developments. Thinking about the
field of medicine, for instance, being able to predict health
issues before they’re even diagnosed just based on patterns
of behaviour. Or manufacturing and using data to replace
components before they actually break. Or personal health
and fitness, getting real-time feedback through things like
biometric monitoring.
These are the types of possibilities that are already becoming
reality today and it’s only going to continue down that path
in the future. So that’s number one that I think is really
exciting, just to see data play a role in places where it really
hasn’t in the past. And second, I think we’re going to see a
lot more accessibility to advanced tools and techniques,
things that up until this point required years of training and
even a PhD to deploy. And we’re certainly seeing this already
with the rise of self-service BI tools and with open-source
libraries, but it’s now kind of getting to the point where, in
some cases, some guy off the street could build a pretty
decent predictive model using some free software and a few
clicks.
Now, whether that is a good thing or not, I think that’s a
different question altogether. Personally it’s a little bit
frightening to me, but I’m trying to be conservatively
optimistic about it. But that certainly feels like the path that
things have been going. It’s just this concept of self-service
BI and this accessibility to very advanced tools and
techniques.
Kirill: All right. That’s such a cool description. And how do we
prepare for that? How do the listeners of the podcast prepare
for that future so that they have careers that are aligned
with this future?
Chris: So, as far as preparing for a future in data science, there are
two things that I would recommend. Early on, I think it’s
really important to get exposure to as many different types of
diverse projects as possible. You know, having a role in a
consulting firm like Deloitte, for instance, or with an ad
agency, anything that’s client service-focused where you get
exposure to different types of projects, or even just exposing
yourselves to different types of Kaggle competitions or
exploring personal projects like the ones you and I talked
about, just try to get exposure to a really broad range of
projects early on. I think that’s really important.
But eventually, what I personally would recommend is
starting to focus really on becoming an absolute badass in
one or two particular areas. There’s something called a T-
shape skillset which I personally believe in, which is
basically having solid working knowledge of a pretty broad
range of skills. So if you think about listing those skills
horizontally and then really having one or two where your
level of expertise goes really deep, that’s like the vertical line
of the T.
So, I believe in developing a T-shaped skillset, and in my
personal experience, I found that the strongest teams tend
to consist of complementary T-shaped people, each with the
ability to speak intelligently about a very wide range of
topics, but who have one or two or even three specific world
class skills. That’s recommendation number one as far as
preparing yourself: Get a lot of exposure early on and then
think about starting to focus on what you really feel
passionate about.
And then number two is just learning how to constantly
adapt. We’ve talked so much about how fast this field is
changing, be it the tools, the techniques, the best practices.
So at the end of the day, those who evolve along with it are
the ones who are going to thrive. And there’s really no
excuse these days, given the accessibility of educational
content out there today. You and I know Udemy is a great
example of that, Coursera, edX, Lynda.com, the list goes on.
So, I think learning how to adapt and evolve and learning
how to learn is a really important skill for someone who
wants to get into a field that’s changing as fast as data
science.
Kirill: Very cool, Chris. That is really very cool. You made me think
about this now and personally I think that learning how to
learn has been a killer skill in my arsenal. Without that, I
really wouldn’t be able to be where I am right now. And
speaking of your T-shaped approach, personally I think my
vertical one, the one that I’m really deep into, is probably the
communication side of things. So it’s not a technical skill,
it’s more the people side of things that I—like, when I need
to, I can communicate the complex insights and so on.
Yeah, that’s a very, very good overview. Thank you so much.
I hope that will make other people listening to this podcast
also think about their approach right now. Thank you so
much for coming on the show. How can our listeners contact
you, follow you, find you if they’d like to learn more about
how your career develops from here?
Chris: Sure. You can find me at ExcelMaven.com, contact me
through the website. I’m also on LinkedIn, happy to connect
with anyone who wants to get in touch. And if you’re
interested in the coursework, or the training side of things,
you can find me on Udemy or on Lynda.com.
Kirill: Perfect. Fantastic. We’ll definitely include all of those links in
the show notes. And one final question for you today: Do you
have a book that you can recommend to our listeners to help
them become better data scientists?
Chris: Instead of a book, I am going to give you two blogs which I
actually have become a huge fan of and I think everyone
should become familiar with. Blog number one is
InformationisBeautiful.net. It’s a collection of some of the
most unique and powerful data visualizations that I’ve ever
seen. So if you’re into data viz and you’re into charts and
graphs, and really unique ways to present data, check out
InformationisBeautiful.net. That’s a great one.
And then the second recommendation that I have is
FiveThirtyEight.com, which is Nate Silver’s blog. It’s really
just about taking an extremely analytical approach to
popular stories and politics and economics and sports. It’s a
really, really entertaining read. Those are my two
recommendations.
Kirill: Yeah, fantastic. Thank you so much.
InformationisBeautiful.net and Nate Silver’s blog,
FiveThirtyEight.com. Once again, Chris, thank you so much
for coming on the show and sharing all your insights about
business, education, consulting, Excel and so much more.
Chris: Thank you very much, Happy to be here.
Kirill: So there we go. That was Chris Dutton, a top instructor on
Udemy and also the founder and CEO of ExcelMaven.com. I
hope you enjoyed this episode. Personally, for me, probably
the biggest takeaway was this whole situation which we
talked about at the very end about the different things that
you need to focus on going forward and one of them was the
ability to learn all the time, which I personally love doing
and I know love doing because you’re listening to this
podcast. And, of course, that T-shaped skill personality – I
think that’s what it’s called – that was very, very valuable as
well and it made me think about my skills from a different
perspective and in a way I haven’t thought of it before.
Hopefully some of these elements on this podcast made you
think as well and maybe now you’re a bit more excited about
learning Excel if you weren’t previously. For me, personally,
once again, Excel has been kind of like the foundation on
which I built my future data science career, so it was a
necessary step, and I’m really glad I did learn Excel to the
extent that I did. So, thank you very much to Chris for
sharing his insights today. And of course, you can get all of
the show notes at www.superdatascience.com/89. Make
sure to follow Chris on LinkedIn and check out his website,
ExcelMaven.com, it’s very well-made. And of course, you can
find him on Udemy as well. And on that note, I look forward
to seeing you next time. Until then, happy analyzing.