SDS 179: How a Data Science Recruiter Thinks

SDS 179: How a Data Science Recruiter Thinks

Data Science Recuiter Welcome to episode #179 of the Super Data Science Podcast. Here we go!

Feeling stuck in a career slump? On today’s episode, we talk to Matt Corey, a Data Scientist Recruitment Provider, to guide you. Though patience and positive mindset really help to get you out of it, it might be better if we take an advice from the expert!

Subscribe on iTunesStitcher Radio or TuneIn

About Matt Corey

Matt Corey is the leader of Change Force, an exclusive Data Scientist Recruitment Practice which helps clients find the best data scientist for the specified industry. He is also an advisor, a speaker, and the author of Data Scientist’s Book of Quotes.

Overview

Matt’s job can be considered unique and very relevant to the data science community. On the previous episodes, we’ve talked about how important it is for data scientists to present and convey analysis to the target audience. We want to get actionable plans and improve an industry’s operations. To make that work, Matt acts as a mediator between the employers and the data scientists. He learns first what his clients (e.g. hiring managers, entrepreneurs, etc.) demand. Then, he looks at his own network of data scientists. Most of them reach out to him through LinkedIn. He submits 3-4 resumes of people with a set of skills and experience which would prove suitable for what they’re looking for.

What if you’re a data scientist who has already set an eye on a certain employer and don’t know how to present yourself? If you have issues on how you communicate, Matt helps in these kinds of preparation. Be confident. Matt suggests you tweak your CV or resume to showcase your achievement that would matter to the target industry. Highlight the tasks you’ve done that could be similar to what they’re looking for.

Most of the time, we rely on the job descriptions for vacant positions. We look into the essential and desirable criteria. Sometimes, the prospect jobs require years of experience that you couldn’t satisfy even though your CV gives a perfect impression to other employers. When Matt himself sees this kind of issue, he makes recommendations to the client. They could change what is written on the job descriptions or Matt advises them how similar skill sets could also work on what they demand. Matt advises the aspirants to evaluate where they are, gain experience, and learn more if they have to. It’s a never-ending journey. For employers, fully understand what your organization needs to improve operations. Be flexible, be open-minded, and learn to adjust your expectations.

There may be times that we see no direction and satisfaction in our jobs. We just have to continue looking what suits us, and that is what data scientist recruitment provider can offer. Let’s try to avoid shutting doors to once-in-a-lifetime opportunities unknowingly.

Let’s end today with Matt’s favorite quote from his book:

“Your work is going to fill a large part of your life and the only way to be fully satisfied is do what you believe is great work. And the only way to do great work is you love what you do. If you haven’t found it yet, keep looking. Don’t settle. As with all matters of the heart, you’ll know when you find it.” – Steve Jobs

In this episode you will learn:

  • What does a Data Scientist Recruitment Provider do? (05:30)
  • After years of working in human resources, Matt decided to niche in the data science recruitment practice. (07:06)
  • “Data science is a community unlike anything that I’ve ever seen.” – Matt (08:29)
  • His job is beyond what a normal recruiter does. (10:37)
  • The difference between passive and active candidates. (13:30)
  • When it comes down to specific roles, take a thorough look into the essential and desirable criteria in the posted job description. (16:15)
  • How do you prove your relevant skills if you don’t fit the job description? (18:15)
  • For hiring managers and entrepreneurs, it’s important to be flexible when looking for the right person for the right job. (22:20)
  • Certain data science skills are flexible and transferrable in industries. (34:20)
  • Temporary vs. Permanent Roles of Data Scientists. (41:00)
  • Matt talks about his favorite quote from his book Data Scientist’s Book of Quotes. (43:35)

Items mentioned in this podcast:

Follow Matt

Episode Transcript

0

Full Podcast Transcript

Expand to view full transcript

Kirill Eremenko: This is episode number 179 with Data Science Recruiter, Matt Corey. Welcome to the Super Data Science 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.

Welcome back to the Super Data Science Podcast, ladies and gentlemen. Super excited to have you on the show today. And we've got a very interesting and insightful guest joining us. Matt Corey is a data science recruiter. And what I found very interesting about Matt was that he actually specializes only in data science recruiting, specifically just that niche. And that's what we were talking about in this episode. You will find out many interesting tips and insights into what recruiters look for when finding candidates for a data science position, you will understand what to kind of expect from recruiters.

Also on the other hand, Matt will share some insights on how he works with his clients, the companies that are hiring. And you'll understand more about their thinking, what are they looking for, what are their fears, what are their desires, what is driving them. And moreover, we'll talk about the intricate role of a good recruiter in data science. Not just the person who puts people in positions, but a person who acts as a bridge between the candidates and the clients. A person who works with expectations of clients, because we all know that data science hasn't been around for that long, and yet a lot of companies have huge expectations. They're looking for unicorns, they're looking for people with 10 years in data science experience, and lots of different tools and techniques and methodologies, and industry knowledge, which just physically don't exist. And so Matt will share his insights on how he goes about those situations, and how he works with the clients themselves to manage their expectations.

So if you are looking to hire data scientists, this episode is also going to be valuable for you. And finally, Matt has just recently published a book. You can buy it on Amazon. When we were recording the podcast, only the ebook version was available. But when this is gonna go live, probably the hard copy's gonna be available as well. It's called, The Data Scientist's Book of Quotes. And I can't wait to get my hands on that book, because it's got some very valuable quotes. It's got over 300 quotes in there, categorized by different areas of data science and different topics. So I'm looking forward to getting that as well. And we'll talk about the book and some, he'll share some insights from there too. So on that note, can't wait for you to check out this episode. Without further ado, I bring to you Matt Corey, a data science recruiter. Welcome ladies and gentlemen to the Super Data Science Podcast. Today we've got a very exciting guest on the show, Matt Corey. Welcome, Matt. How are you doing today?

Matt Corey: I'm fine, Kirill. Thank you so much for inviting me. It's a real pleasure.

Kirill Eremenko: The pleasure's all mine. Matt, where are you calling in from today?

Matt Corey: I'm calling from beautiful London.

Kirill Eremenko: Amazing. And you-

Matt Corey: [crosstalk 00:03:42] warm at the moment, about-

Kirill Eremenko: You-

Matt Corey: ... 30 degrees.

Kirill Eremenko: That's fantastic. 30 degrees Celsius?

Matt Corey: 30 degrees Celsius, yes.

Kirill Eremenko: Just for our U.S. listeners, that's ... I should find out. I'll find out what is in Fahrenheit. Which is 86 degrees Fahrenheit. Quite a lot for London. Quick question, you mentioned that it's been warm for quite a while now. And you've been in London for 20 years. How warm has it been before ... How long has it been warm for now?

Matt Corey: It's been warm ... Yeah, it's a great question. Thank you. Yeah, it's been warm now for about almost a month. And I'm talking about maybe one day where it rained possibly, like a couple of days sort of in the evening. But in general, it's been a good month of just solid sun, really.

Kirill Eremenko: Fantastic. That is totally, totally fantastic. The first time I went to London was last year, I got there, first day it was sunny. And I thought, "What is everybody talking about? Why the rain, the bad weather? It seems lovely." But then on the second day that's when the rain started, and it was like four times in the day it was raining. So yeah, I'm a bit ... I'm actually very excited for you right now that it's such a good time of the year.

Matt Corey: Yeah, no. It's amazing. It's kind of expected, because it's obviously the sorta Summer period. But it's not always like that. And this is really really a treat this year. So I'm talking about sorta climate change and all that, it is happening. And it is a lot warmer now than ever. I mean this is, we're in London, and it's kind of Mediterranean weather.

Kirill Eremenko: Yeah.

Matt Corey: So we're blessed.

Kirill Eremenko: Yeah. Okay. Yeah, something to be concerned about as well, I guess. Well okay, so Matt, you are a data scientist recruitment provider. You're an advisor, speaker, and now a book author. And we'll talk about that in a second. Tell us-

Matt Corey: Thank you.

Kirill Eremenko: ... quickly from a high-level perspective, what do you do as a data scientist recruitment provider?

Matt Corey: Wow well, very simply I ... First of all, it is a niche. It is only data scientists that I provide to clients and organizations. So it is exclusively data scientists unlike others who choose to do the whole sorta data science sort of portfolio in terms of analysts and engineers, and architects. I felt that there was a real importance and need for data scientists to have that very sort of special role in terms of providing the insights. And I think that with over the years what I've seen is that more and more, it's a role is gonna take a sort of predominant role within changing business, and providing sustainability. And also really being able to maximize the data that is already inherent within organizations. So that's why I chose to only focus on data scientists.

Kirill Eremenko: Mm-hmm (affirmative). Gotcha. And how long have you been doing that for?

Matt Corey: It's been a little bit under a year.

Kirill Eremenko: Mm-hmm (affirmative). Okay. So you've been helping data scientists get roles in the past year. And where did you come from into this space? Where was ... Where were you recruited?

Matt Corey: So my background is within HR, Human Resources. And I started off my career as a generalist HR person. Then focused within recruitment. And at some point I then became an independent contractor. And there were a few changes in the market, and I decided to set up my own recruitment practice. And I initially started off within change and transformation. But within as I mentioned, about almost a year ago now, I felt that that was a bit too broad. And I wanted to really focus on, and zero in on one certain position that was so very very important. And I had seen a film called Money Ball, which you might've seen with Brad Pitt.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: And there were certain things that were sort of ... That were coming up into my life, and seeing the film, and then reading a few articles. And then suddenly it was like, "Wow. Data. Data science. That's the real change. That's really what's happening." And then I then just dived and just read as much as I could, ask people about it, and just eventually just set it up as a data scientist recruitment practice.

Kirill Eremenko: Mm-hmm (affirmative). Gotcha. And so how's it been going? You've been doing it ... The whole transformation and change you've been doing for quite a while now. But the data science part that you've been doing for the past year, how's that been going? Have you been able to help many people?

Matt Corey: Yeah, I have helped many people. Either in terms of placements, or in terms of advice, or in terms of helping them with their CVs. Get a lot of people from across the globe. My LinkedIn connections have just sort of skyrocketed. I'm currently doing a promotion for as a sort of Summer promotion for one person to get a free CV and sort of LinkedIn profile rewrite. And it's just been massive in terms of the response and people being interested in, and thanking me. And it's ... You know, the thing is what's happening is that the data science, it is a community. It really really is a community, unlike anything else that I've ever seen.

Kirill Eremenko: True.

Matt Corey: I mean, I was in HR before, in change and transformation. But data science is a real community. They really join. They really help each other a lot.

Kirill Eremenko: Yeah. Yeah. I totally agree with that. And they share, and they comment, give feedback like in a positive was on what can be improved, resources like LinkedIn articles that people are sharing and writing about their learning pathways. Or GitHub code that people are sharing with each other. Or comments on Tableau public dashboards, or-

Matt Corey: Yes.

Kirill Eremenko: ... Kaggle competitions. And how data scientists collaborate on Kaggle competitions. It's very exciting to see people from different parts of the planet actually come together to do these projects. So couldn't agree with you more on that one.

Matt Corey: Yeah, it's incredible. It's a very giving community.

Kirill Eremenko: Mm-hmm (affirmative). Very excited to be part of it. And so in terms of like ... You mentioned a couple things. You mentioned you help people with placements, mentoring, also rewriting CVs, or advice of how to write them-

Matt Corey: Mm-hmm (affirmative).

Kirill Eremenko: ... LinkedIn profiles and so on. Could you give us a bit more insights into like the different aspects that a recruiter does? So what is the job of a recruiter in the space of data science? Like those items that you help people with, and maybe a bit more details on those if you can?

Matt Corey: Well I mean, what I do is a little bit beyond what I would say a normal recruiter does. I think that's where ... Because of my specialism in being exclusively a data scientist recruitment practice, or a recruiter.

Kirill Eremenko: Yep.

Matt Corey: I have a sort of a greater sort of insight, a greater focus on what I'm doing. And I also want them to succeed. So I mean I know that personally, apart from some people who work with me and work for me, they ... And I try to also tell them that you need to also coach people a lot of times, you need to help them. They need sort of some preparation in terms of their interviews. At times there's an issue of confidence.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: They need to maybe at times improve their communication style. So it is about ... It's not just about sending a CV or a resume, it's also about helping this person. This person is having ... This will have a major impact in their life, on their family, on their whole sort of circle within either their family, their friends, their life, their children. And it creates a major impact. And that's why I think one of the reasons why I'm in recruitment is because when you help that one person get a job, you make a major impact in their lives.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: And it goes way beyond just getting them a job, just [inaudible 00:12:13] in. It's about also helping them. I've seen so many people change not because of the job, but because of the process of getting to the job.

Kirill Eremenko: Mm-hmm (affirmative). It's not the end destination, it's the journey that matters, right?

Matt Corey: It is the journey. It's the process. Yeah. Definitely.

Kirill Eremenko: Fantastic. And an interesting question I had in mind while you were speaking came to me, how often do you see people looking for a new job because they're unhappy in their current job? Not because simply data science is the trendy thing to be in, but actually because when they were choosing their original job, they found something with a high pay, or something that was available, something that sounded really interesting, but they didn't do enough research to understand is this the right thing for them? How often does that happen that people are really unhappy in their role, and therefore they're looking for a new opportunity?

Matt Corey: Interesting question. I mean, I have to answer it in a slightly different way just to sort of so I can see how-

Kirill Eremenko: Sure.

Matt Corey: ... I can best answer this. I would say that there's passive candidates and there's active candidates.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: I would say that because of the way I work and others with me, the market is primarily, we always approach mt passive candidate as much as possible. It's not just about the active candidates. The active, when I say the word, "Active candidates," they are the ones out there saying, "Here's my CV. I'm leaving in a week."

Kirill Eremenko: Oh, okay.

Matt Corey: And you have the passive ones who are in roles, so who are either happy, going back to your original question. Or possibly unhappy and have accepted it.

Kirill Eremenko: Oh, okay.

Matt Corey: But we always approach passive candidates as well, because we're looking for certain people with certain experience from certain industries.

Kirill Eremenko: Gotcha.

Matt Corey: Because our client is looking for that.

Kirill Eremenko: Yeah. Okay. Gotcha. So you kind of act as a head hunter for the businesses, for the clients that need those skills?

Matt Corey: Yes. I mean, it is a matter of also looking at ... You know, we have our own database-

Kirill Eremenko: Yeah.

Matt Corey: ... of course. We have our own network. So I have my own network that I know. I have then my LinkedIn sort of network. I then have the database. I then also have people who know people, who I then seek out let's just say as an example, a data scientist who's worked in retail and I have a client who's like saying, "I definitely want someone from this company, X, Y, Z company in retail." Or, "I definitely don't. I want someone completely different. I don't want anyone from retail. I want someone from banking or financial services. And then who has so much experience in this specific area."

So it's about then looking for that person. Now that individual again what I mentioned earlier, may be happy with where they are, they may not be that happy. They might be happy with their salary, but they don't like their boss, they don't like their manager. But they're also then weighing it up and saying to themselves, "The salary is good. My boss is so-so." But people normally leave not because of the money necessarily, of the salary or the package. They usually leave because of the environment within the company.

Kirill Eremenko: Mm-hmm (affirmative). Gotcha. No, no, I agree. Okay. Interesting. Interesting. All right. And then so, on our podcast, and just in our community of students, we have quite a large portion of listeners and data scientists. Or not actually data scientists yet, but listeners and students who are in adjacent fields, are in either IT, or something similar like system administration. Something to do with technology. And they want to move into data science. What would your advice be for them? What is the current status of the job market in data science? Is it a good idea to move from IT, business intelligence, system administration and so on into the space of data science?

Matt Corey: Absolutely. Of course. It's going to be ... I mean, I think that every business out there in the future, if that's in five years, if that's in 10 years, will be talking about that they spoke to the data scientist consultant, or their data scientist within the company. It'll be an absolute norm in future. So do I think it's ... Yes, absolutely. Anyone who wants to be get out of their position, if it's from business intelligence, or from IT. Or whoever has an interest, this passion about data science, or to become a data scientist, do it. If they're to do courses with you, or other providers, absolutely. Definitely.

Kirill Eremenko: Gotcha. Okay. And so you mentioned courses. What are the best steps to make this transition? What is even the starting point? I get this question a lot, where would somebody start if they want to transition into data science? The thing is that there's a lot of demand for data science skills, right? And some people have already a lot of experience in something very similar to data science. Some kind of field that they can leverage their experience from. But at the same time, they're not technically qualified to apply for data science jobs that require five years experience. So somebody might have 10 years of experience in IT, or programming, or database design. But there's a job that requires five years of data science experience. What would you say is the first step? And how should people thinK about their prior experience? Should they be like, "Okay, well that prior experience that I have is actually now irrelevant. And I should start from scratch." Or should they find ways to demonstrate the values that they've provided and actually show that it is relevant to the role that they're seeking in data science? And how can they do that?

Matt Corey: Well again, a very good question. It come down to the specific role. It comes down also how long ... Not how long. How many years they're actually looking for. So when you have a job description, you have a role. You have a job description and it says, "Essential." And this is where we split things that we say, "Essential criteria," and, "Desirable criteria."

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: As an essential criteria, it is something that they essentially want you to have. And at times it's flexible. It could be, "We want experience ... " No, "Experience in, or proven successful experience in." That makes it quite broad. If it's, "Essential three years experience in," and you don't have it, and you only have one year, then you're completely then crossed out. And you'll not be considered at all. So it is about evaluating where you are, "Can I go for this role?" If they're asking for ... I mean, let's put this more specific. I mean, I mentioned the example earlier about retail. If they're asking for a data scientist who have experience of three years experience within a retail environment-

Kirill Eremenko: Yeah.

Matt Corey: ... of successfully implementing projects, et cetera. Predictive analytics, et cetera. And then you don't have it, then you can't really apply for it. Unless the field is not ... Hasn't really ... They can't find someone to have the three years experience. But they have someone who has two years experience, or a year and a half. They might then either rewrite the job description, and allow that person to apply. That's what normally ought to be done.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: But it comes down to what the employer's looking for, and where that person is. And to what extent the employer is willing to be flexible, and to what extent the prospective employee is willing to either train themselves up, go off and do a course, reapply in future, or would be considered today because the company's willing to train him or her to reach the level that they are expecting. I hope [inaudible 00:20:46] sort of answers your question. I'm not [crosstalk 00:20:48]-

Kirill Eremenko: Mm-hmm (affirmative). Yeah. That's ... I appreciate your comment on that. I just wanted to see, what about this scenario? For instance, the job description says, "In retail. Data science application, prediction and modeling in retail," and so on. "Three years of experience." And the person applying has let's say, three years of experience, but not in retail, not in data science. They have three years of experience in business intelligence and reporting in the healthcare industry. Something kind of like technologically relevant. But not exactly the same thing, and not even in the same industry. But now this person instead of completely foregoing this opportunity, and completely giving up on it.

What they do is they go and do an online course in data science and retail. They go to [Cagle 00:21:44] and download datasets about retail datasets. They go to the World Bank, or some other sources of data science and retail, and datasets relevant to that role. And actually do projects. They demonstrate their capacity. So over the next six months, they do six major projects, they write articles on LinkedIn, they write six blog posts on LinkedIn. They share their code on GitHub, they do [inaudible 00:22:10] dashboards on Tableau Public. They do a Kaggle competition and they take 17th place, and so on.

And so they demonstrate that even though they don't have the three years of experience, they are capable of producing the results that the employer want. What will happen in that situation? I know it's a case by case basis depending on the employer. But do you think that strategy, that approach actually has a chance with the right employers for that person to get hired? Or is it like-

Matt Corey: Yes.

Kirill Eremenko: ... [crosstalk 00:22:40]?

Matt Corey: Personally I believe that they have ... Yes. The answer is yes. And it would be, they have a very good chance. It also depends on how flexible the employer or the hiring manager is.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: If they are savvy enough in terms of all these mediums, and are aware of the value of that, and are willing to consider all these things, fantastic. But I think a non-data science person or someone who is not immersed enough in, not sort of involved enough, and not aware enough, they would take it very very strictly and just cross it out. I mean, very open about it. But it comes down to what extent who's actually going to be shortlisting for this role.

Kirill Eremenko: Yeah.

Matt Corey: And how strict the criteria is.

Kirill Eremenko: Mm-hmm (affirmative). Okay.

Matt Corey: So if the hiring manager tells for example HR, "I only want this. Do not consider anything else. I don't wanna see anything from ... I just wanna see exactly that. I can get 500 people tomorrow who have something which is slightly different. I don't want that. I'm looking for exactly that." So it really comes down to how flexible and how open-minded they are to accept other related transferable experience.

Kirill Eremenko: Gotcha. Okay. And then flipping the coin onto this other side, what would you say to hiring managers who are listening to this podcast? Or to entrepreneurs, or business owners who are looking to hire data scientists? Should they be flexible, or should they look specifically for that type of person from that industry with that experience? I'll tell you my opinion on this. I think that there's so much demand in this space of data science, that being inflexible can be costly in terms of time and in terms of the talent that you pass by. But I'm really interested to hear your opinion, because you're in this space. And you might say, "No, look you have to. Like if you really want something specific, you gotta stick to it and go for it." So what are your thoughts on it?

Matt Corey: Thank you. I mean, you're obviously an entrepreneur. And you understand that one has to be flexible, one has to be open-minded. And I think that's a certain mindset that not everyone has.

Kirill Eremenko: Yeah.

Matt Corey: Personally I do make recommendations. I do at times, depending on my relationship with the client, I would then adapt and say, "Look, I think that this person is hitting the mark. They're not hitting the mark in exactly the way the job description has been written. Maybe this person doesn't have the three years experience. However, you are looking for this and this, and this. And in order for this project that you have at the moment that you want someone to have ... It doesn't necessarily require three years of experience, because he or she has actually done this Kaggle project, has been on GitHub, and has done very relevant things here. If you go on their website and you look at the projects they've done, or on Kaggle, you'll that they've been quite high up of in terms of where they rank. And they've done really really well. And a lot of these comments are actually very relevant to what you're looking for." And it's also my reputation as well on the line, because I'm then talking to a client who trusts me.

Kirill Eremenko: Yeah.

Matt Corey: And I also don't want to put forward a person who I think cannot do the job. I'd rather just say, "You know what, I'm sorry. I can't find someone." And my role is also to send very few resumes over. I don't like to send, if someone is a client of mind is looking for one person, I don't send 10 CVs or resumes, I send maybe a maximum of four.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: Because I want it to be the absolute best ones. And if I send only one, then my client knows that, "Matt has sent me the best CV, because that is the only one that he really believes in enough."

Kirill Eremenko: Yeah.

Matt Corey: So it's at times you have to challenge in a nice way, your client. Because you're there to also inform him or her that, "I think in this case, you ought to see him or her, because they have relevant experience. And I think you'll very quickly find out. If you don't wanna see him or her in person, and you'd like to have maybe like a 10 minute chat with him, I'd recommend at least that."

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: But I would say ... I would definitely say, let's say for example this person is another country. I mean, for obvious reasons we're gonna have a Skype interview, we're gonna have some sort of a chat online. Don't fly this person in if you have some reservations. Have a chat with him first. No, because I say that because I have lived the experience of where we would ... When I worked as an internal recruiter where at times maybe we would maybe not thoroughly check it. And that's something which I had my own views. But you have to also work with everyone. And at times, not everyone's perfect.

Kirill Eremenko: Yep.

Matt Corey: So it's about also ... It goes back to what we said earlier, about having that flexibility.

Kirill Eremenko: Yeah. Yeah.

Matt Corey: And being sort of understanding what the objectives are, and seeing if this person can actually make this happen for you.

Kirill Eremenko: Mm-hmm (affirmative). Yeah. I agree. And so basically that's a great transition to the ... I guess, the over your ... Of your role, a role with data science recruiter, it's not just to find, head hunt the right people. And it's also not just for the client. It is also not just for the individuals who are looking for a job to put them into jobs. Your role as I see it is much bigger than that. It's actually being that middleman, and being that advisor/negotiator who guides this flexibility. And it's exactly what you said, that you need to ... Sometimes clients, especially in this space of data science which is so new, they're looking for something that is like a unicorn, that doesn't exist. That person with 10 years of experience in data science, and they could do this, and this. When some of those technologies haven't even been around for 10 years. And so that's where this advice and like kind of shaping up the expectations of the client comes in.

And I wanted to draw on my own experience in this matter. And this is going back to when I was leaving Deloitte, I was looking for a job. And sometimes I would get contacted by companies directly. Like for instance, two banks contacted me about potentially working with them. And sometimes I would get in touch with recruiters. And I remember this specific day a recruiter went onto my LinkedIn, and I saw that they look at my profile, but they didn't message me or say anything. So I hunted them down, messaged them myself and said, "Hey, like I noticed you saw my profile. Is there anything I can help you with?"

And they said, "Well look, your profile looks interesting. But the job we're recruiting for," or, "Job I'm recruiting for is not ... It requires more work experience." So this was a role in a pension fund that required six years of experience. And I only had two years of data science experience at Deloitte. And some work experience prior to that not in the field of data science. And in total it wasn't even close to six years. And so as you can imagine, that's quite a large difference. Six years in data science versus two years in data science plus a bit of work in an unrelated field, or not in a specifically data science field.

And nevertheless, what I told them was like, "Let's catch up, and I'll send you my CV. Tell you about the projects I've done. Bring you a portfolio of the projects I've done," like a desensitized portfolio of the projects I've done, "Just to showcase all the projects that I can do. To showcase my abilities and show you that I can actually deliver for this plan." And in the end after we caught up, they really thought that I can do the job. They recommended my CV to their client. And when I went for the interview, I got the job.

Matt Corey: Fantastic.

Kirill Eremenko: Yeah. And that's where I worked for a year after that. And so yeah. Just stands to show that sometimes, or actually quite often especially with larger corporations where this processes of recruiting are standardized, they are still not entirely adapted to the situation in the data science job market, and just the profession as a whole. And so they need people like you, Matt, to adjust their expectations, to be more flexible, and eventually to get the candidates that might not meet the criteria exactly, but that will get the job done, or that actually maybe even get the job done better than who they thought they were looking for.

Matt Corey: Yeah.

Kirill Eremenko: My question would be to you here is, how often does that happen? How often does it happen that you help the client be more flexible?

Matt Corey: Yeah. I think this is an excellent question for many reasons. Because I think that is also a reflection of not just the data science world, but this is also a reflection ... What you're entering into an area which is fantastic, because I think it's something which isn't really discussed enough. And I think it's something which the industry ... Or when I talk about the industry now, I'm gonna talk about the recruitment industry as recruiters, I think really this is a major, major issue that exists I think for recruiters.

Because it comes down to the recruiter being confident enough to ... So the recruiter in your case for example was open enough and flexible enough, and adaptable to allow your CV, your resume, to be taken onboard. To allow your experience. And then have the confidence to discuss this with their client. Because your background was not straightforward in terms of ... That recruiter had to actually to some extent convince the client to see you.

Kirill Eremenko: Mm-hmm (affirmative). Exactly.

Matt Corey: And that comes down to ... I'll use the word, "Backbone," or, "Confidence," or to say, "Actually, you know what? I'm going to ask the client," and say, "Mister or Miss client, you know, I know that your job description says this. I know that this is what you're looking for in terms of the essential criteria. However, I have met someone who is meeting this, but in a slightly different way. Doesn't meet the ... Here's however their experience is such that I think we ought to consider him," in your case.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: But that comes down to the recruiter being confident and flexible enough, and being able to in a way, in a nice way, challenge their client. And also then the second party, which is the actual client to be again, open enough and flexible, and adaptable enough to allow another resume or CV to come forward, which is not exactly the way the job description has been presented.

Kirill Eremenko: Mm-hmm (affirmative). Okay, gotcha. And so how ... Like what would you say is it? Is it 50% of your clients that you advise that way? Or is it 80%, or is it 20%? I just wanna get a gauge for how is the industry shaping up? I know that a few years ago, that would've been predominantly the case, like people getting these job descriptions very wrong. How is it right now?

Matt Corey: I think it's changing, because we're now having a lot more people who are ... The hiring managers are usually data science professionals. So when I deal with head of data science professionals who are hiring for their team, they are aware of what the role is, because they are essentially data scientists themselves. The difference is that they're also head of data science, so they are running a team. It is rare so far for me to have people who are unrelated to data science be hiring data scientists. So hence they know-

Kirill Eremenko: Yeah.

Matt Corey: ... the nuts and bolts of what is required. So if you were to hire a data scientist tomorrow, you know what you're looking for, because you also have that inside track of knowing A, what you're looking for. And B, you've been there before.

Kirill Eremenko: Mm-hmm (affirmative). Yeah, yeah. Gotcha. Okay. So any ballpark estimate? How often?

Matt Corey: I'd say at the moment, the majority ... I mean, to give you an exact percentage I would say at the moment for me at least here in London, it's normally about 70 to 80% of people are data science professionals. And I don't necessarily need to challenge them in that sense, because they know what they're looking for. I can ... There's always gonna be some flexibility I mean, if they say three years. But it's not so much years, it's more about having a certain experience. But I do admit that there's a very strong industry preference. So I do have clients who are very specific in terms of having that industry experience. So if it's retail, they want retail. It's quite rare you hear, "I don't want ... " If they're in retail and they say, "I don't want anyone in retail."

Kirill Eremenko: Yeah.

Matt Corey: That's quite rare.

Kirill Eremenko: Yeah.

Matt Corey: Because there's a certain comfort. And I'm gonna say that that is disappointing, because I have also been on the other side of the fence as a candidate where when I finished off with a client years ago who was in FMCG, Fast Moving Consumer Goods. And at that time in the market, there was a real boom about having financial services or banking, which I didn't have. They were so, I'm gonna use the word, "Fixated," on having that for the extent that it's like, we were all kind of ... Anyone who wasn't in that ... Didn't have that industry experience, was just not invited.

So I've lived it as a candidate. I know how that feels. And it can be very frustrating, especially when you have so much experience that as a recruiter, I personally have done so many different areas, that a recruiter is a professional. And they adapt. And if you want me to find you a sales manager, or a sales director, or you want me to find you a fundraising director, or you want me to find you a head of data science, or you want me to ... There's a point where a recruiter becomes so adept that he or she is going to learn the industry, learn the role or the roles-

Kirill Eremenko: Yeah.

Matt Corey: ... and be ... And also know the competitors as well, well enough. I mean, a true professional that's what one does. You immerse yourself so much in understanding what the role is, you even go and do ... You spend a day, in this case for example today, with a data scientist. You go and you ask your client, "Can I sit in within a meeting and understand things, how they work here?"

Kirill Eremenko: Yeah.

Matt Corey: So it's about immersing yourself. And yeah, it is about ... But to go back to your original sort of question, it really comes down to the person. The majority of them in my case are data science professionals.

Kirill Eremenko: Mm-hmm (affirmative). Yeah. Gotcha. And what you mentioned about the industry focus, I agree with you. It is disappointing, because in addition to your points, it is such a flexible profession. Knowing how to deal with data in the health industry, and then taking that skill and learning how to deal with data in the entertainment industry, or in the public services industry, it takes a couple weeks maximum for somebody to gain all that domain knowledge, the core domain knowledge. Of course there's gonna be details that you will learn along the way. But the working of the data part of the skill is extremely transferable. And I know that coming from consulting where at Deloitte, one day I was working on a railway. Another day I was working like analytics for a railway company. Another day I was working on a healthcare industry. Another one I was working for a mining services company.

So very very transferable skills. If I was recruiting for a data scientist right now, and I was in a specific industry, the last thing I would put on my job description is, "Industry specific experience." Because ultimately that is not relevant at all. What are your thoughts? Do you agree with me on that, or do you have a different opinion?

Matt Corey: I agree, and I'm gonna say both. I'm sort of on the fence with it, because I'll tell you why.

Kirill Eremenko: Okay.

Matt Corey: It comes down to as an individual, I'm absolutely 100% behind you. Because I want to give everyone a chance.

Kirill Eremenko: Yeah.

Matt Corey: I think it comes down to also how pressing it is, because if the industry is quite complex. And if for example there's a project that involves someone to know the expression, "Hit the ground running," and really be able to very very quickly be knowledgeable enough to such an extent that they would have to really know the industry well, because the project is for three months, the project is for six months max. And it really requires someone to have a certain amount of industry experience. That is where I would say I understand it.

Kirill Eremenko: Yeah. Yeah. Gotcha.

Matt Corey: If it was a permanent role, I would say no. I don't think it requires in this case. And also depending on the role in general. But I think the more ... The less time you have, I think it's justifiable to say, it is all right ... Again, depending on how important the role is with respect to having some industry experience.

Kirill Eremenko: Mm-hmm (affirmative). Yeah. Okay. Makes total sense. And can you tell us a bit how often do you recruit for permanent roles, versus temporary roles like you just mentioned, six, 12 month projects? What is kind of the slit that companies are looking for?

Matt Corey: It's primarily in my case here in London, it's primarily ... Or in the U.K. I would say it's more so on the permanent side than the temporary. I've also worked more in the permanent market. But I would say so far for me, it's been more on the perm side.

Kirill Eremenko: Mm-hmm (affirmative). Why would you say that is? Is that because companies wanna build out their internal data science divisions more than they just wanna get a project done?

Matt Corey: I think it's also there's a cost element as well. Because when you hire someone on a permanent basis, it is more cost effective as well. When you hire someone, in this country at least, on a temporary basis, you're hiring them as a contractor. You're paying them more, much more than you would be paying them on a permanent contract. Or at least in this country again, we also have a term called fixed term contract, which is for a year or two years.

Kirill Eremenko: Mm-hmm (affirmative). Gotcha.

Matt Corey: Which can be ... So if I gave you a salary in terms of U.K. Pounds. So if I said to you that someone's earning £70 000, U.K. Sterling, versus someone who's earning then ... What can I say? A salary from 70 000, then they would be earning something like ... I don't know if they were earning 600, 700 a day, 800 a day, 900, a 1000 for example, a day. That is a very very different sort of model in terms of hiring someone on that basis. And it's quite costly. And in this case also, in this country at the moment, the public sector which is government, doesn't normally hire at that rate as much. It's been ... Things have changed here. So it comes down to also, are we talking about the private sector or the public sector? So we know there's private sector of course, private companies. Or public sector meaning government. And obviously if we look at this as a global podcast in every country, it's different.

Kirill Eremenko: Yeah.

Matt Corey: If it's the U.S. If it's in India, if it's Australia, if it's the U.K. If it's Germany. It's different every single market. I mean, now we're talking about sorta local differences.

Kirill Eremenko: Yeah, okay. Fair enough. Gotcha. Okay. That was quite insightful into the world of recruiting. Thank you for that little discussion. And now I wanted to move on to something a bit different. And that is your book. Congratulations, your book just got published. It's very exciting to see it on Amazon. And-

Matt Corey: Thank you.

Kirill Eremenko: ... you showed me the hard copy when we were talking on video. So how are you feeling about that? Must've been quite a lot of work that went into it.

Matt Corey: Yes. I mean, it was quite a bit of work. Surprisingly I wrote it I think within a few months. And I think it's been an amazing, amazing learning curve in terms of writing a book. I think people say, "Oh, wow. You wrote a book." Or you know, "That must be amazing. I would've never thought of writing a book." And I had thought about writing books, but not necessarily ... I never thought I'd write a book so quickly.

Kirill Eremenko: Yeah.

Matt Corey: And I never ... I think in your book, Cognitive Data Skills, as you mentioned one doesn't sort of grow up and think that they wanna become a data science necessarily when they're growing up. But I never thought growing up that I'm going to be writing a data science book of quotes.

Kirill Eremenko: Yeah. Oh, yeah. I'm sorry. For the listeners, I forgot to mention the name of the book is Data Scientist's Book of Quotes. Please continue, Matt.

Matt Corey: Thank you. Thanks so much. Yes, the book is available at the moment on Amazon as a Kindle book. And the paperback will be available hopefully in about from let's just say on the safe side, will be about maybe 10 days to two weeks.

Kirill Eremenko: Yeah. Well, by the time this goes live, it'll probably be available. We have-

Matt Corey: Okay.

Kirill Eremenko: It'll live in a few weeks anyway. And I wanted to say that I had a look at some of the quotes. I don't have it yet, but I'm definitely gonna order it as soon the hard copy's there. And I had a look at some of the quotes examples on Amazon. You can do a quick preview of a book, it'll show you a few pages. And so basically it's broken down into different chapters where you can ... You get quotes from different people in that space. For instance here's one I like, "Without a grounding in statistics, a data scientist is a data lab assistant." That's Martin Jones, Managing Director in Cambrian Energy.

Here's another one, "Data scientist are kind of like the new Renaissance folk, because data science is inherently multi-disciplinary." John Foreman, Vice-President and Product Management of Mail Chimp. So some very interesting ones that make you pause and think. And it reminds me of the book I'm reading now. What is it called? The Art of Life. It's about stoic philosophy, but explained in simple terms. And it's got a lot of these, not quotes, but kind of like little passages. And there's no way you can just sit down and read cover to cover in one day. Because even though it's a small book, simply because it provokes so much thinking. And what-

Matt Corey: Yes.

Kirill Eremenko: ... I like about like a book like yours, like with quotes, whereas you open up a page and you read a couple of quotes, and then you sit down and you think about them. And it provokes some new ideas in you. And on top of that, what I found useful, or I'm looking forward to finding useful when I read your book is that you broke it down into chapters by grouping the quotes together by their different style ... Or not style. More topic.

So for chapter one is like, "What is a data scientist?" Chapter two, "Power and potential of data and data science. Data's value." Then you go all the way onto ... Let's go through them, "Treatment of data." Chapter four, "Not having the right data. Potential risks of data. Challenges with data. Machine learning. Deep learning. Artificial intelligence. Data ethics. And data privacy. Future of data." So if I'm gonna be like, "I want to learn about ... "

I wonder, "I have a problem on data ethics," that I have a discussion with someone I need to have soon, I'm gonna open up chapter 11. And I'll read a couple of quotes on data ethics and privacy. And [inaudible 00:47:57]. And I again, I haven't read it. But it sounds like a book good to have, nice to have in your library for the time when you're gonna need to pull out when you have some free time, or you need to learn a bit about it. So really cool idea. How did you come up with the idea for the book?

Matt Corey: Well, I thought I definitely want to be ... I want to write a book. And I thought, I'm not at a point where I'm that knowledgeable yet to write an entire book. I mean, I'm fascinated with how to create a data driven organization, how to have a data driven culture. I'm fascinated of course with the role of the data scientist. But I thought, "Do I have enough knowledge yet to write a book today?" I mean, I don't mean within a few months. And then suddenly I just, I saw some other books on the market, different subjects. And I thought, "Wow, you know what? I can actually write."

And I checked it up and thought, "Well, I didn't see any book like that out there." And I thought, "You know what? I can actually write a book of quotes," because I know there's obviously books from literature, et cetera where they have quotes from people. Sorry. And I also think that because I'm also, I literally write quotes in, I have these books, these journals. So I think we talked a little bit about before where I'm a huge Tony Robbins fan.

Kirill Eremenko: Yeah.

Matt Corey: And I have a few books of his with quotes. I have a journal of quotes by him. And he also quotes people in the past when he first started his career. And he literally has a book of quotes from people that he admires. And I remember writing a lot of these quotes in my own journal. So I have a kind of predisposition to writing quotes. Because I think that this is where people provide these nuggets of knowledge and also life experiences. And it makes you really wonder.

Because I am a believer that life is very short. And life can be very full. And I do view life as being half full and half empty. And it's what you make of it. And it really is about making the most of it and doing your absolute best every day. And how you think and what you believe in, and if you believe the worst, then the worst will happen. If you believe in the best, the best will happen. You stumble along the way in life, but you need to pick yourself up, dust yourself off and keep going. And I'm a firm believer of that. And there is a book that I read many, many, many years ago. And it absolutely changed my whole life.

And that book today, I mean it still is out there. And it's called, The Power of Positive Thinking by Doctor Norman Vincent Peele. So that was the book that for me I'm gonna say ... Oprah says that books are her friends. That book not only was my friend when I was 17, but it also in a way saved my life in a sense, because I didn't do well in school on a certain course. And I remember literally failing that course. Here I am publicly saying that. And what happened was that I read that book during that Summer. And I enrolled in that course again. And I went from a failing mark to passing it with 89.

Kirill Eremenko: Mm-hmm (affirmative). Wow.

Matt Corey: And did I become a genius in that course? No. I simply believed enough, and I studied enough, and gave my all to pass it. And I came in second in the class. And I was able to continue my education as a result of it. Because I wouldn't have been able to go to university if I didn't.

Kirill Eremenko: Mm-hmm (affirmative). Gotcha. Okay. Wow, okay. That's a little interesting that you got the idea for this book. But yeah, I think it's gonna be a great success, and great help to many people in field. I guess we're talking about data science being a community. And I think we needed some kind of resource like this to be able to reference different people. My question too [inaudible 00:52:34], what's your favorite from your book? I think you have like 320 quotes in there or something. What's-

Matt Corey: Yeah, that's right.

Kirill Eremenko: What's your favorite one?

Matt Corey: Wow. Oh, that's a question.

Kirill Eremenko: Weren't ready for that, were you?

Matt Corey: No, I wasn't. I wasn't. I'm just thinking, "Oh, what do you say here?" You know what? The thing is, I also have many quotes which in the book, I know that you maybe can't see it at the moment because of the fact that you have the sample.

Kirill Eremenko: Yeah, yeah.

Matt Corey: But I will ... There's a lot of people in there who, like Warren Buffet, and Tony Robbins, and Bill ... Okay, Bill Gates, that I have quotes from. And something else which you, just to mention that after each chapter there are exercises with questions.

Kirill Eremenko: Oh, wow.

Matt Corey: And there's some notes. So people can actually answer the questions for themselves and for their organization.
Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: So I'm gonna say one quote which really does stand out of me. But it's not necessarily data science quote. So is it okay if I mention this one?

Kirill Eremenko: Yeah, yeah. Yeah, of course.

Matt Corey: Okay. "Your work is going to fill a large part of your life. And the only way to be truly satisfied is to do what you believe is great work. And the only way to do great work is to love what you do. If you haven't found it yet, keep looking. Don't settle. As with all matters of the heart, you'll know when you find it." That's by Steve Jobs.

Kirill Eremenko: Wow. That's a really cool quote. And very also right in time for this podcast, right? 'Cause we were talking about recruiting and head hunting, and how to find a job. Really cool. I really appreciate you sharing that. Great. Hopefully that will-

Matt Corey: Thank you.

Kirill Eremenko: ... get people thinking, is your heart in what you're doing? Or is it not? Matt, is your heart in what you're doing? You've been doing it for a year. How are you feeling?

Matt Corey: I love it. It's what I mentioned to you earlier that I'm excited by it for many reasons. And thank you for your question, because I'm excited with the fact that it's fresh. It's really in demand. It's much needed. You really can work in a much more efficient manner. And when I talk about sustainability, this is what I'm talking about really. When I heard a statistic a while back that we only use ... And I think, I'm gonna say we only use about ... And this is even the max. And I think it's actually 1%. But I'm gonna say 5%. I'm gonna be even more ... I'm gonna raise it up a bit more and say that a company or an organization only uses 5% of its entire data.

Kirill Eremenko: Wow.

Matt Corey: That is shocking.

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: Shocking that they don't fully utilize their data. And a data science and others in the data science arena can fully utilize all of the data. And I think that's where, to be on that sort of cutting-edge of a profession that is so much needed. And I'm gonna say something else that has to do with life. It has to do with companies who are out there, small and medium companies that don't have a lot of resources, that don't have a lot of time and money. But they're able then to fully utilize their data. It saves them time. It saves them money. It saves them hardship. It saves them ... I can tell you from my own personal experience. And if I had known that I can maximize my data in my own past, I would say that I would be in a different place today.

But I say that in a very honest and very open manner that by utilizing a data scientist for one's own business, either as a consultant or as an employee, you are working in such a more efficient and effective manner. So yeah, I am very passionate about it. And it is something which I do love. And I also love the fact that this community is such that it's a very open, very giving, very new, very helpful, and you used the word, "Sharing."

Kirill Eremenko: Mm-hmm (affirmative).

Matt Corey: I think it's something which the community itself is very helpful, very giving, and willing to help each other. Very very much so. And in terms of resources. And LinkedIn is a primary example. I mean, we wouldn't have been talking today if it wasn't for LinkedIn. And LinkedIn is, you see so many books being offered. So many resources being offered. Algorithms, et cetera, "Use this." And, "I'm learning this. And this is how I got my job. And this is how I ... This is what I did." And there's a lot of sharing.

Kirill Eremenko: Mm-hmm (affirmative). Yeah. Wonderful. Thank you so much Matt, for those insights. I totally, totally appreciate your comments. And it's exciting to be a part of this community, exciting to be a part of this broader group of people who are all passionate about the same one thing, which is data science. So thank you so much. I think we'll wrap up the podcast on that. I really appreciate you-

Matt Corey: Okay.

Kirill Eremenko: ... coming on the show today sharing your insights.

Matt Corey: Thank you for inviting me.

Kirill Eremenko: Where would you say is the best place for our listeners to find you, contact you, get in touch, or follow your career? Or maybe some people are looking for jobs and would like to get in touch to a recruiter. There might be companies that are looking for a recruiter to help them out. Where is the best place to do that?

Matt Corey: Well, the website is ... So for the business, the Recruitment Change Force is the business. And as I mentioned, it is an exclusively data scientist recruitment practice. So that's on changeforceinc.com. So go Change Force INC.com. And my details are there. So in terms of phone number and the company sort of details. I'm on LinkedIn. So it's Matt Corey. It's M-A-T-T and then C-O-R-E-Y. So I'm on LinkedIn if someone wants to ask me a question. So there's that, and the business details will be on the website. The book as yeah mentioned, [inaudible 00:59:28] thank you for that again, is on Amazon. I think that's pretty much it. I mean, I'm the kind of person who either myself or my staff are very, we do our best to help people, and to find them roles, relevant roles for them. And yes, it is about data science, but we're always open to hear, to help people in general.

Kirill Eremenko: Gotcha, gotcha. And just-

Matt Corey: [crosstalk 00:59:53] data science professionals. Yeah.

Kirill Eremenko: Yeah. Just to reiterate, the book's called Data Scientist's Book of Quotes. All right. Well, we'll have all those links on the show notes for this episode. And-

Matt Corey: Thank you.

Kirill Eremenko: ... on that note, thank you very much again, Matt, for coming on the show and sharing all your wonderful insights and knowledge with us, with the listeners of the podcast.

Matt Corey: Thank you, Kirill. I appreciate. Thank you so much for your invitation again.

Kirill Eremenko: So there you have it. That was Matt Corey, a data science recruiter, and author. I hope you enjoyed today's episode. And I hope you will pick up a copy of Matt's book, the Data Scientist's Book of Quotes. As I mentioned on the podcast, I think it's a very necessary tool for people, especially data scientists to have to just take time to ponder philosophically about our industry and about the work that we're doing, and maybe come up with some new ideas based off or inspired by other people's quotes. People who are leading this space.

And I'm curious to find out what your favorite part of the episode was. My favorite part was probably when we talked about the intricate role of data science recruiter, a good recruiter. Not somebody who just tries to match the job description and find the right people who exactly match the specifics, but somebody who can talk to the clients about managing their expectations and maybe adapting them to who's available in the market, and what kind of skills are there. And understanding their actual needs, because sometimes companies create these job descriptions, and they ... Even though they describe what they think they want, it's not actually what they want.

And on the other hand, a good recruiter should also work with the candidates to help bring out the true nature of their experience. The true value that they can bring to the company, and help them see more about themselves than they actually think. So see those hidden maybe gems in their experience and their expertise, and their background that might be valuable to different job roles in different companies.

So all in all, it was fun episode. And I hope you learned a lot. You can and probably you should connect with Matt, because it's always good to have a recruiter in your network on LinkedIn. We'll include Matt's URL in the show notes, which you can find at www.superdatascience.com/179. There you'll also find all of the links to the materials we mentioned in this episode, plus the transcript for today's show. And on that note, I hope you enjoyed the episode. Can't wait to see you back here next time. And until then, happy analyzing.

Kirill Eremenko
Kirill Eremenko

I’m a Data Scientist and Entrepreneur. I also teach Data Science Online and host the SDS podcast where I interview some of the most inspiring Data Scientists from all around the world. I am passionate about bringing Data Science and Analytics to the world!

What are you waiting for?

EMPOWER YOUR CAREER WITH SUPERDATASCIENCE

CLAIM YOUR TRIAL MEMBERSHIP NOW
as seen on: