Toggle Menu

Episode 10

Data Analytics: Hard Skills vs Soft Skills and the Gift of Thinking Different

Analytics manager, and Women in Data podcast producer and host Karen JEAN-FRANCOIS walks us through the differences between Data Science and Analytics. Join her and Sam as they discuss valuable skills you’ll need when transitioning to a career in Data Analytics. Hear Karen’s perspective on the benefits of thinking differently and having a mentor to guide you through transitions.

Published January 28th, 2021  |  27:22 Runtime

Episode Guest

Karen JEAN-FRANCOIS

Karen JEAN-FRANCOIS

Senior Data Analyst at Cardlytics, Host & Producer of Women in Data Podcast

Episode Transcript

Karen JEAN-FRANCOIS:
I was able to find a mentor who helped me understand that my differences were my strengths. That thinking differently is actually a gift. And I felt more confident about talking with stakeholders, helping them solve their problems with data, and most importantly, I stopped second guessing my opinion, and that was the most rewarding thing ever.

Sam Ramji:
Hi, I'm Sam Ramji and this is Open Source Data. Today, we're interviewing Karen Jean Francois. While studying to be a math teacher in 2010, Karen fell in love with data science, which prompted a career change leading her to a master's in applied statistics. Her career in data started in predictive analytics in Paris, where she was analyzing payment defaults.

Sam Ramji:
She then moved to London and is now working for Cardlytics as an analytics consultant and manager. There, she manages the analytics related to all financial institution partners. Previously, Karen trained as an international athlete, running 100 meter hurdles for France, admiring athletes, such as Marie-José Pérec, a promoter of peace, which gave Karen a focus to continue and strive for more. Today, Karen is bringing more transparency on how data is used across industries and shining light on women working in the field, through the Women in Data Podcast, which she hosts and produces. The podcast covers topics such as values, career challenges, team structure, the transition from having a technical role to being a leader, as well as transitioning to data science. Karen, welcome, and thank you so much for joining us.

Karen JEAN-FRANCOIS:
Thanks Sam. And thank you so much for having me.

Sam Ramji:
I really enjoyed our conversation back in October, as we explored a little bit about what our audience might want to hear from you. But I'd like to start with asking you what does Open Source Data mean to you?

Karen JEAN-FRANCOIS:
I listened to a few of your podcasts and I thought, "What am I going to answer to that question?" Because you have so many great answers in there. And what I settled on is that for me, Open Source Data is just a synonym to the community. So yes, we're sharing data nowadays, so you have open data, you have open softwares that are used in the data industry. But the community for me, is amazing. So it's all the data professionals you find that are open to sharing tips, sharing their knowledge, mentoring others, but also, opening their network to everybody who could benefit from it. And this is what it is for me.

Sam Ramji:
That is pretty awesome. I've heard the term open source expanded into open sources. Sometimes they talk about that in the intelligence community or in other similar areas. The idea that you can just publish all of the things that you've learned and let other people analyze it. So, that focus on community on people, is really very inspiring to me.

Karen JEAN-FRANCOIS:
I think this is what's moving the industry forward, really because things change so fast in data, very important to be able to talk to each other and share what we know.

Sam Ramji:
And you end up doing a lot of that in your day-to-day work as a leader, but also as a community leader and a podcaster. So I'm wondering if we could start by exploring what's the difference between a Data Analyst and a Data Scientist? And I think our audience would love to get a sense of your personal experience and what you've learned along the way.

Karen JEAN-FRANCOIS:
That is such a difficult question, actually. I feel like you could ask that questions to 100 data professionals and you would have as many different answers. And sometimes I'm not even sure we know what data science means. So if you take the example of a few years ago, Monzo, which is a mobile bank in the UK, they did some job title testing. So they posted two job offers on their website. One saying Data Analyst, one saying Data Scientist. Same job description, same salary, same set of skills, just the job title being different. And they've got more than twice more answers for the ones with Data Scientists. And the quality of the people applying there was also much better than the ones applying for a Data Analyst, but yet it was the same job. So this is one of the things that make me feel like there is a confusion there.

Karen JEAN-FRANCOIS:
Data Science roles and Data Analyst roles are still being defined. And you could take one data science or even one Data Scientist and compare that professional to another one and they would have completely different roles. So even within data analytics, you're going to have Data Analysts doing completely different things. And I think it's both scary and exciting at the same time. So exciting because it gives you an opportunity to shape your role and explore and do things that you want to do. Scary because we grew up in a world where people know what a profession entailed. So if you are a teacher, if you're a math teacher, you know you're going to teach math and this is what you're going to teach for the year. If you want a career as a data professional, you don't always know what you're going to do.

Karen JEAN-FRANCOIS:
But if I try to say what I think is the difference between a Data Analyst and a Data Scientist. If you take apart Advanced Analysts, because there are some Advanced Analysts that would do something that is closer to data science. So removing them from the equation, I would say that Data Analysts would work more on quick ad hoc and quick target run projects, than Data Scientists who will tend to be more on the longer projects, focusing on mobilization, clusterings and things like that. So this is the biggest difference I see between the two roles, but both of them will do a lot of data wrangling. Do you have a new opinion on that?

Sam Ramji:
I think it's a fascinating development of the two roles. I've seen certainly Data Analyst, if you think about the semantics, and the semiotics, sort of the bigger meetings around analysts, it feels like someone who's deployed in the service of doing exactly what you said, quick turn projects for somebody else. So there's somebody who's got the business insight, who's got the question and they say, "Hey, go crunch the data for me." So Data Analyst feels a bit more data crunching, closer to reporting, how do you get things procedural? Whereas science, the semiotics of science are all around research and exploration of self-direction and agency. So you see people going on farther excursions to explore and find new things. And then some bring something really valuable back from the edge of practice that could surprise people. Personally, I've seen the change over time as we've sourced Data Scientists and Data Analysts differently with different degrees, different specializations.

Sam Ramji:
I think we're starting to see Data Scientists graduating with degrees in data science. But a good friend of mine, Nick Goffeney, who is the head of Data Science for Facebook AR actually has a PhD in biophysics. So we saw this movement of people with these advanced science degrees, moving into industry to bring their scientific, computational, quantitative numerical analysis skills into realms of data. And that's kind of my sense of the early days of data science and now it's becoming a bit more of a commonly known profession, but there's still a lot more standardization ahead, which personally I think is pretty exciting. It tells us about the decade of data to come. 10 years from now, these roles will be really well understood. The infrastructure will be really well understood, but right now it's kind of a new frontier.

Karen JEAN-FRANCOIS:
Yeah, I agree. And I'm looking forward to seeing that. Not to be 10 years older, obviously, but to see what's going to happen in the data industry. And I think also, I don't know how it is in the US but in the UK, some businesses are creating a lot of gray areas on that, because there are businesses that will just call everybody Data Scientist, even if it's a reporting analyst for example. And that doesn't help, so I'm really waiting for that to change. And for people to stop asking me if I'm going to get promoted to a data science position, because data analytics and data science are different career paths. And me being a Data Analyst doesn't mean that the next step is being a Data Scientist. I could get there, but it's not the next step. And I really want people to understand that.

Sam Ramji:
There's often, in these kinds of waves of change, there's title inflation. I remember when 20 years ago, some advanced developers started being called Architects and all of a sudden, everybody wanted to grab an Architect title instead of Software Engineer with DevOps. I've seen cases where people who are really operation professionals, who are focused on just keeping machines up, they're not doing a lot of automation. They want to be called DevOps engineers because they type some part of the CI chain. It's almost the hype of the new thing is signaled by the people who are doing the old thing, but want to be called by the new thing. So when I was younger, I was more frustrated by these kinds of things. But now I see it as a signal. Hey, something neat is happening. There's a change in the industry. So I would say that don't fear being 10 years older, getting older beats the alternative. And it's actually quite lovely.

Karen JEAN-FRANCOIS:
Yeah. I will keep that in mind.

Sam Ramji:
So there seems like there's a lot of room for growth in the industry. What should people expect when transitioning, as you did over to a career in analytics?

Karen JEAN-FRANCOIS:
There is definitely a lot of growth and things move fast. And with data being democratized, there are new roles who are going to get created. So it's not all about Data engineers, Data Architects, Data Analysts, and Data Scientists, it's going to get bigger than that. And it's going to open doors to people who did not think they could have a career in data. In terms of what to expect when you're transitioning, that's another tough question. Different industries will have different requirements. Every company will take a different approach to data. This is what they're doing right now. And hopefully in the future, there will be more clarity on who's doing what and what you need to get there. But I think the questions people who are transitioning into data should really ask themselves right now is, "What kind of data professional do I want to be?" And then make that happen. So we spoke about the community that shares tips and knowledge and opened their network, use that. So open your network to be able to find out who's doing what, what it is that excites you, and then get in that direction.

Sam Ramji:
I think because one of the things that people want as they prepare themselves to go there, partly reach out and get in the community. They also want to know, what are the standards here? As we know, there are a lot of skills required to be a data professional. And as we just said, the roles are not yet as clearly defined as they are in other domains. What are the skills required to be a data professional that you value? You're a leader, you're a hiring manager, you're also an experienced and competent practitioner. So could you lay out kind of the core set of skills that you believe are important that you look for when you hire and that you value on your team?

Karen JEAN-FRANCOIS:
Yeah. In data analytics, I think just like data science, there are two types of skills that are very important. So you have the hard case that are going to be coding and understanding statistics, and you also have the soft skills and other things. So I think in terms of coding skills, when you are into analytics, something that we tend to ... not under value, but we don't talk about it as much. So everybody talks about Python or R, but SQL is extremely important when you are a Data Analyst because most of the time you work on structured data and you need to be able to manipulate that data and put it in the right format, clean it before you export it and put it in whatever, or the software or programming and language you're going to use.

Karen JEAN-FRANCOIS:
And I think people who are transitioning into the industry, if they want to go into analytics, this is definitely something they need to look at. I would love to say SAS, but I saw a LinkedIn post yesterday, I think from SAS saying, "Why you SAS for AI?" And I was thinking, "Really are we trying that now?" I need to upskill now.

Sam Ramji:
Well, it sounds like those are some hard skills, R, Python and SQL. Often the SQL components in larger organizations, I've seen being pulled out towards data engineering, building out large scale pipelines, using tools like Qubole and others. But knowing at least some SQL seems really important. What are some of the other hard skills or what are some of the soft skills that you hire for and assess to make sure someone's going to be successful on your team?

Karen JEAN-FRANCOIS:
Curiosity is extremely important. Data is a field that moves really fast. And I've had lots of people who are trying to transition into data careers telling me that it feels like they are in a never ending learning loop. That every week they learn something new and it feels like they will never stop learning. They will never be able to apply for a job because they can't absorb that much information. And you need to understand when you're getting there, is that going into a career in data means that you're never going to stop learning, because everything changes. So eight years ago, everybody wanted someone who could code in SAS, now everybody wants someone who can code in Python and R. So it means everybody needs to up-skill or learn new things, new technologies, even if few coding SQL, for example, your company can move from SQL server to Vertical, something like that.

Karen JEAN-FRANCOIS:
And you have to learn then your language. Still similar, but it changes a bit. I think this is a blessing, because always having to learn to mean that you're always developing and career development and personal development is something I am very passionate about. So showing that you can learn, that you're happy to learn, that you are a self-starter and showing curiosity is extremely important. And other skilling is having some business awareness. So being able to take a business problem and transform it into data. So take something that is okay, the business has this question. How can I solve that with data? And that people will be able to understand what I mean and being able to apply it and solve their problem with that. So, these are skills that we look for when hiring.

Sam Ramji:
That was really helpful. So you've gone all the way through the curve from being the hands-on individual contributor and all the way through your current level of leadership. No doubt there is way more ahead. We had the privilege of talking with Margot Gerritsen recently, and she emphasized that one of the things about a career in data is you'll often be promoted outside of your comfort zone because there's so much growth and demand for more people in the teams to do more work. You'll probably keep getting promoted above your level of competency and you'll be uncomfortable. So thinking about that arc. As a Data Analyst, can you tell us what are the most rewarding and most challenging times in your career with open source in mind?

Karen JEAN-FRANCOIS:
I actually have a story that we'd cover the both at the same time. More and more you hear about people questioning their career choices. And I think that the pandemic had made that even bigger. So I've had, for example, a lot of people are coming to me asking me about transitioning into a career in data. And a few years ago, it was my turn to ask myself, what am I doing here? And to question my career. And one day what happened is I actually walked into a bakery and I asked the pastry chef, "Can I come and bake with you on Saturday? I just want to know if this is a career I would like to consider." So for two months I woke up at 4:00 AM on Saturdays to do full shifts, slicing cakes, glazing tarts, and baking sponge cakes. I did get some wicked baking skills out of that.

Karen JEAN-FRANCOIS:
But I'm still in data, and I am really glad I did not change. I believe that this happened because there was a lack of role models I could identify to and I was struggling to find my space and my voice in the industry. So I remember sitting at my desk and thinking, "What am I doing here? I don't fit. I have different interests." As you mentioned, in my introduction, I was an international athlete and no one was interested in sports around me. I would rather be on the track running then learning Python, for example, sometimes. And I was thinking, "No one is like me. I don't fit, I think differently. I shouldn't be here." And Margo mentioned that in the podcast we did with her, actually. She mentioned the fact that sometimes you feel as a woman in the industry, you feel alone and you can feel like you don't belong. And I feel like I was exactly there.

Karen JEAN-FRANCOIS:
And actually according to women in data in the UK, only 20% of data professionals are women. And it's very problematic when it comes to finding a role model, right? But through the community, through all the people that were surrounding me, I was able to find a mentor. So that is Victoria Pike who helps me understand that my differences were my strength, and that thinking differently is actually a gift. I should not feel guilty about the fact that I think differently. And once I understood that I actually started to own my career, which is something that never happened before. And I felt more confident about talking with stakeholders, helping them solve their problems with data. And most importantly, I stopped second guessing my opinion because others were louder or just having a different opinion. And that was the most rewriting thing ever, because all the beautiful things that happened after that, so that was my podcast, that was me being promoted at work, that was me taking on management responsibilities and leading on other projects. These all came from that. And I do believe everybody should have a mentor, because it makes such a difference.

Sam Ramji:
That is just fantastic. I'm smiling from ear to ear, listening to your story. I think it's the best career advice in a way that I've ever heard.

Karen JEAN-FRANCOIS:
Thank you.

Sam Ramji:
And I'll remind you what I said about getting older. Getting older is the sum of those kinds of experiences and living through them and kind of growing into your feet, growing into your paws. Like a puppy becoming a full dog. And I remember very vividly those moments for me of letting go of some self-doubts and moving forward with a little bit more confidence. And those are always those moments where all of a sudden, it's like the room lights up and you can see where you're supposed to go.

Karen JEAN-FRANCOIS:
Exactly.

Sam Ramji:
I would love to hear anything unique you learned in 2020 from your guests on the Women in Data Podcast.

Karen JEAN-FRANCOIS:
All right. On the podcast, I tend to have very diverse guests. So they're all women working in data, both working from different fields. So I have been mind blown by the extent of information they had to share, all their career stories were amazing. But because we've been talking a lot about transitioning into data and data analytics versus data science. I'm going to take the example of Sumedha Menon, who works for Google in the UK. And she was talking about the steps she took to transition from data analytics to data science. And basically she asked herself three questions, which I believe are questions that anyone should ask themselves at any time in their career, if they want to change jobs or if they want to transition. So it doesn't have to be transitioning from data analytics to data science. Any time you want to change something, these questions are really great.

Karen JEAN-FRANCOIS:
So the first question she asked herself was, "What do you want to be?" The second one was, "Where do you need to upskill?" So, is it in programming languages? Is it in going deeper into statistical modeling and clustering techniques and things like that? And the last question was, "What type of company / culture do you want to work with?" And that was very important for her because she has young kids and she wanted to be able to finish work early, for example. So a company culture that was understanding of that and understood that, because she leaves early doesn't mean that her job is not done. That was very important for her.

Karen JEAN-FRANCOIS:
And that led her to her job at Google. And she was very happy about it when she spoke about it, and I found her story very inspiring. Another thing I learned is, people are trying to transition into data and all the overwhelming information that's out there / available to them and how they handle it, and how they can sort through it and make their transition successful. So I learned that many people were getting discouraged because they thought it was too hard, that there was too much to learn and they couldn't do it. So this is something that I actually learned about myself through that experience or through talking to people who listened to the podcast. I discovered that helping people transition into data careers is something that I would like to focus on in the next years.

Sam Ramji:
Well, it seems like you're doing a wonderful job with that.

Karen JEAN-FRANCOIS:
Hopefully.

Sam Ramji:
There's a bridge between data analytics and data science that's I think maybe coming. I had the opportunity to talk with Simba Khadder, who is the founder of StreamSQL.io ... and he helped me understand where AutoML might fit in this equation. He felt that when you think about data science, there's a lot of machine learning specific skills and a very particular set of infrastructure that you need to be able to do your training runs and figure out how you can get a good neural network trained to produce particular assessments on a set of data.

Sam Ramji:
And that production cycle of those is pretty difficult because there are so few people who can do that and so much work waiting to come in. And so his view was AutoML would be a technique that would enable Data Analysts to cut way, way, way far into that backlog so that they could say, "Hey, let's take a look at this, I've got a concept of a feature. Let's run AutoML on it and see if something comes back." Because that might get you 90% of the way and prove whether or not your thought about a feature is valuable or not, and kind of shorten the gap between the two sets of skills. What do you think about that?

Karen JEAN-FRANCOIS:
It's a tough one because I think because I come from data science background, so with my master's being in applied statistics, I feel like I like to play with the data and explore it myself. But it will definitely help people who would be a bit less into the statistics side of things to get the information out of it can make things much faster for them as well.

Sam Ramji:
It's going to be interesting to see how all of these technologies move around. It seems like there's a lot of development ahead. What is a technical resource that you might offer our audience given where we are? We're entering 2021. It's been a wild ride through COVID for so many reasons, but one of the things obviously is that there's been a much higher rate of digital transformation. And what I see is there are way more jobs and money flowing into the tech side of businesses and away from let's say the retail side as companies try to adapt themselves to succeeding with data. So as you look at the technologies and techniques that are coming into the mainstream, or maybe just emerging, what's a resource that you might point our audience at?

Karen JEAN-FRANCOIS:
There's so much resource out there. Sometimes I even feel like there is too much because you have to sort it out and find out which one works for you. Something that I'm doing right now, I am learning Python, for example. So using all the online courses is something that I do a lot. I found that different courses will work for different people. So DataComp is one that most people talk about. I'm not entirely sure it works for me. I like it. There is also Udemy. But I think where you would get the most out of it is there are a lot of podcasts and YouTube videos about people who are in the work and talking about things that they're doing. And this is something I find very amazing, because getting the knowledge from someone who is on the job at the moment and is talking about what they're doing is just priceless for me. And in fact, when it's on YouTube or on podcast, it's all for free. So head there, yeah.

Sam Ramji:
Well, that's part of the commitment we all make as a community to open source, right? We want to share the knowledge freely so that everybody can get on board. And then in return, they can all enrich the whole community again, and it gets bigger. So the beauty of these positive sum games that we play in open source is pretty exciting. And it's neat to see it growing in pretty much every demand in every industry.

Karen JEAN-FRANCOIS:
Yeah.

Sam Ramji:
I would then direct people to the Women in Data podcast, as a great place to get a jumping off point into what's changing. It's up-to-date, you can subscribe, you can listen to all of the new episodes and get a sense of what's coming and what you might want to adapt to. So, that might be my answer to the one resource question. Hopefully it's okay with you, Karen.

Karen JEAN-FRANCOIS:
It is definitely.

Sam Ramji:
That's awesome. Well, thank you so much for your time and thank you so much for taking time to talk with us at Open||Source||Data and I wish you a very happy birthday. We won't say the number of course, but you're celebrating your birthday this weekend and we wish you the happiest, healthiest, safest, most peaceful birthday you can imagine.

Karen JEAN-FRANCOIS:
Thank you, Sam. And thanks for having me. Now, it's my turn to smile ear to ear.

Sam Ramji:
Thank you so much.

Karen JEAN-FRANCOIS:
Thank you.

Narrator:
Thank you so much for tuning in to today's episode of the Open||Source||Data podcast, hosted by DataStax's Chief Strategy Officer, Sam Ramji. We're privileged and excited to feature many more guests who will share their perspectives on the future of software, so please stay tuned. If you haven't already done. So subscribe to this series to be notified when a new conversation is released and feel free to drop us any questions or feedback at opensourcedata@datastax.com.