Introduction and Background
I attended college in the 90’s and mentorship is not something that was discussed at the time. In recent years, I have heard the conversation of “find a mentor, you need a mentor to advance your career.”
In early 2018, when I first began writing and when my fastai_deep_learning GitHub repository of documentation received increased traffic, I received numerous requests from readers around the world to be their mentor. These requests encouraged me to think more closely about mentorship and the general expectations in the data science community.
Survey on Mentoring
I held certain beliefs about asking for mentorship, but thought I would ask around and see what others thought prior to writing a blog. In early 2019, I investigated this further by interviewing colleagues and conducting a survey on mentoring. The survey was distributed on Slack, Twitter and LinkedIn and 50 people responded. 70% of respondents did not have a mentor, 55% had never been a mentor and 90% wanted to obtain a mentor. The demand for mentors is high; the supply is low.
This research elucidated that there are various understandings of what a “mentor” is and the roles can be categorized primarily in five ways which are Mentor, Sponsor, Tutor, Career Coach and Therapist.
- Provides advice, support and guidance
- Provides guidance on navigating work situations
- Offers feedback; provide constructive criticism
- Shares technical resources and expertise
- Helps to find solutions
- Shares different perspectives
- Generally someone more senior and established in the field
- Opens up new opportunities for people
- Provides a connection to a valuable network of people and resources; provides introductions
- Advocates for them at work or other professional situations
- Provides opportunity for advancement to the next level
- specific, technical expertise (programming languages, algorithms, software engineering)
Job / Career Coach
- Develops skills related to public speaking and networking effectively
- Provides guidance on emotional intelligence skills
- Advises on navigating political situations
- Managing stress
Therapist / Psychologist
- Helps patients/clients make decisions and clarify feelings in order to solve problems
- Provides support and guidance
The good news is that depending on the support needed, the last few roles, tutors, career coaches and therapists, are people who can be hired and are accessible.
CodementorX provides technical tutoring on a variety of programming languages. Even though the word “mentor” is in its name, it is more accurately paid technical tutoring.
Avenues to Obtain Mentorship
These seem to be a few main avenues for obtaining mentorship:
- Work, formal programs
- Work, informal connection
- Outside of work; community
Large companies have structured, established programs. For people who are seeking mentors, it is a good idea to work at a larger company.
Also, when choosing a job, choose a boss over the company because your direct supervisor and your relationship with your manager has the greatest influence over the direction of your career.
What is the best way to find a mentor?
This is advice from others:
- Ask someone you know
- Ask someone you know to recommend someone
Understanding the Demand for Mentorship
Looking back at the numerous companies where I worked, I was fortunate that at times, particularly earlier in my career, I had inspiring managers or colleagues who were knowledgeable, patient and generous with their time. But, for the majority of organizations, my managers were generally unavailable, uncommunicative and indifferent about offering opportunities or encouraging career growth. I think it is almost expected that one should leave the company for better opportunities. I think that could be due in large part to being in the field of statistics.
I was continuously searching for a supportive manager, just as the elusive, non-existent “Fountain of Youth”, there was an endless search for “Involved and Supportive Manager.” The desire to progress in one’s career is perfectly natural, even a positive signal. And yet, so many workplaces do not have that process built into their management structure.
Traditionally, corporate America assigns managerial and leadership positions to those who have advanced degrees (preferably Ph.D.’s) and/or have attended top tier schools. Rarely are they assessed on their management, leadership or communication expertise. And yet, it is these skills that make a formidable and impactful leader. Companies do not consider allocating these positions to managers who have the requisite “people skills.” There are a few ways that companies can address this paucity of skills:
- Train managers –> not easy
- Hire technical managers with good people managing and support skillset –> not easy
- Create a position for managers to actually manage, that is non-technical. Have 2 managers that can fulfill needs for technical and career support and advising.
Given the high cost of turnover, consider that this may actually be a cost-benefit to the company in retaining employees.
Actionable Advice for Those Seeking Mentors
There are a plethora of articles out there about why mentorship is important, but a dearth of resources that delve deeper into successfully obtaining a mentor. Until you find an official mentor, here is some advice.
Look broadly for a mentor
It may seems like someone who is highly visible is a good person to ask, but imagine how busy they are. Ask someone who is more available.
Test the waters
Some people are interested in mentoring, and others are on the verge of burnout. Find the right person.
Time is a Commodity
Time is an irreplaceable commodity. Valuing other people’s time is an effective way to give and earn respect.
Make SMART requests
SMART requests are:
- Meaningful (why you need it)
- Action-oriented (ask for something to be done)
- Real (authentic, not made up)
- Time-bound (when you need it)
Do not ask questions that are easily found via a Google search.
A potential and effective “mentor” will evaluate a request with the following criteria (from Adam Grant):
Can I offer something of unique value to this person that will take me 5 minutes or less?” [low cost / high benefit —> win / win]
Be open to feedback and advice
Do not ask for assistance and then ignore the recommendations that are offered.
Just like any relationship, the ones that work best are two-way. Find out what the need is on the other side and offer that.
Remember that reciprocity is a two-way street. Giving and taking are essential for individual success and positive cultures.
Offer something of value in return. Ask what you can do first before making a request.
Manage your expectations
You will still have to do the hard work. Give up the idea of a “Prince Charming” who will step in, solve all problems and offer the perfect career opportunity. It is a fantasy, not reality.
Role models seem out of reach if you only see them at their peak. To make them relatable, rewind the clock. What looked like natural talent is revealed as hard-earned skill. Study the trials and tribulations that tested—and shaped—their character.
My Story on Mentorship
To share, I have never asked anyone to be my mentor. I have no official mentor. And yet, I feel abundandance with the resources available to me.
This is my advice:
Think Differently About Who a Mentor Could Be
You do not have to have a formal relationship with someone for them to be a mentor. They can be people you see (on tv, at conferences, YouTube) or people whose work you read.
I have not yet met or spoken to Jeremy Howard. Yet, I observed how he involved the community in contributing to fastai open source by his video lectures. Last fall, when I was feeling overwhelmed with the WiMLDS meetup group as all 3 of my co-organizers were on maternity leave, it occurred to me that I could reach out to the WiMLDS community to get assistance for the events. I learned from Jeremy without ever meeting him or speaking with him.
Something to be Learned from Negative Role Models
If you see behavior by someone which is not appealing, that is also educational. We cannot control the behavior of others. But, we can observe it and learn not to behave that way.
One of my previous supervisors exhibited these traits:
- Lack of communication
- Treating people unfairly, playing favorites
- Being unavailable. On the rare occasions I saw her at work, she began every conversation with “I am so busy, I am so overwhelmed at work….” That does not invite any fruitful discussion.
With my meetup co-organziers, whom I have been mentoring for the past year, I have adopted the following behavior:
- Be available and invite questions and discussion
- Communicate regularly
- Treat all individuals on the team fairly; when a team functions well, it is because the leadership creates the environment where everyone feels supported and respected, rather than playing favorites and politics.
- Check in regularly and ask:
How are things going? What is working well, what is not, and what can I help with?”
- Offer constructive feedback
- Explain decisions, the rationale, possible outcomes
- Communicate when I am away and unavailable by either emailing or blocking out my time away on a shared calendar
Think Outside of Data Science: “Transfer Mentorship”
A mentor does not have to be in data science. One of my key confidants is a college friend who works in government, hundreds of miles away from me, doing non-quantitative work. I often share workplace dilemnas with her via phone calls and her perspectives are insightful and useful.
Just as in transfer learning where we transfer knowledge from one machine learning task to another application, we can do the same with mentoring expertise.
Think All Genders
Men, women and gender minorities all offer uniquely valuable perspectives and a diversity of opinions.
Think All Interaction
Anytime I interact with someone and I learn something, anything, from them, that is mentorship. And the reverse holds true as well.
Read and Read More
These are all books that I have read and re-read. I have physical copies of these books which are on my bookshelf, and which I often reference:
- Power, by Jeffrey Pfeffer
- Crucial Conversations; Tools for Talking When the Stakes are - High, Patterson, Grenny et al
- Micro Messaging, Stephen Young
- Getting Things Done, David Allen
- Your Job Survival Guide: A Manual for Thriving in Change, Gregory Shea
- How Google Works, Eric Schmidt & Jonathan Rosenberg
- Lean In, Sheryl Sandberg
You can learn quite a bit from observation. When you observe someone, whether on stage, in conversation, on video, in writing, observe what is appealing about that person and emulate it. Look at how they conduct themselves. Ask yourself:
- What do I admire about them?
- What do they do well?
- How do they interact with others? In person, in writing, on social media?
That is available to all of us without requesting a coffee meeting.
Embrace (Negative) Feedback
While it can be instinctive to be defensive with negative feedback, recognize that it can also be a powerful form of mentorship. We do not have to agree with all feedback, but we can consider it and give it some space.
Put Some Effort Into It
People reach out via email, LinkedIn, Twitter and Slack because it is so easy:
- low effort
- low risk
- thus, low return
Research what areas the person works in and see if there is something in common where a reciprocal relationship can be established.
Open Sourcing Mentorship?
There has been discussion in data science community organizations about offering mentorship programs. However, all of these organizations are operated primarily by volunteers who are already overworked and in high demand. They include meetup organizers, conference organizers, open source contributors, educators and more. It is recognized that there is a demand in the data science community for mentorship. But, the problem with open source is that it is not really free. They are products and services that are primarily created and maintained by volunteers. There is a challenge in obtaining funding by corporations and donations by users. So, if a Mentorship Program is offered, for free, someone is dedicating hours of their personal time for that.
Conferences (such as PyData, PyCon, NeurIPS, ICML, SciPy and others) are primarily organized by volunteers. If they were organized by a commercial organization, the registration prices would be exhorbitant and inaccessible to the majority of attendees.
Personally, I think that offering mentoring services by the data science community would be unsustainable.
Get Involved in the Community / Volunteer
People can interact with experienced people by becoming more involved in the data science communities:
- help at meetups
- review conference proposals
- volunteer at conferences
- ask busy people what they need help with and offer to do those tasks
- contribute constructively to social media discussions in data science
As an organizer for two meetup groups in the bustling city of New York, I can use all the help I can get. For the past year, I have mentored 4 women:
- Melissa Ferrari, NYC PyLadies
- Noemi Derzsy, NYC WiMLDS
- Prithvi Gandhi, NYC WiMLDS & NYC PyLadies
- Antonia Blair, NYC PyLadies
The bonus for me is that I am able to obtain desperately needed help to effectively run these meetup groups. For the 4 women, they receive mentorship from me on communicating with companies, organizing events, dealing with last minute crises, strategically planning for the future and more.
I have asked all of them to add new volunteers to our committees and mentor them through the process so that there is a good pipeline of organizers so the community can benefit from regular meetup events. It has been a mutually beneficial relationship.
I encourage people to email me with questions: rs2715 at stern dot nyu dot edu. While I do not promise a response to every email, I do read them all and write blogs for the questions I receive over time.
Jonathan Rosenberg quotes Bill Campbell:— Reshama Shaikh (@reshamas) April 16, 2019
A coach can
- see the things you can’t
- hear the things you can’t
-SO you can be the person you want to be. #LeadershipDevelopment
Vicki Boykis, Data Science is Different Now
After fumbling on my own for a very long time, I’ve now been established in “data science” for the past 6 years, and, to serve as the mentor that I didn’t have, I’ve been answering emails and having coffee meetings with people looking for advice to get into data science.
Lana Garmire, Call for Mentors
Tweeps, I am looking for female academic mentors, who are: mothers + exceptionally successful senior faculty/scientists + heavy admin duties + advocate for women in STEM . Could you nominate someone? :)— Lana Garmire (@GarmireGroup) November 8, 2018
On the Topic of Sponsorship
Emily Robinson, The Importance of Sponsorship
I’ve previously written about the importance of sponsorship, which is talked about a lot less often than mentorship, on my blog: https://t.co/pdc4FItg3Q.— Emily Robinson (@robinson_es) August 1, 2018
And Lara has a great post on how to be a sponsor even when you’re not a manager: https://t.co/kcmU6543xS
Great example of sponsorship: @lara_hogan wrote her first book after meeting an o’reilly editor at a conf. Editor didn’t just mentor her through the proposal process, she went to bat to help Lara get the contract. And now Lara’s written three books 🙌https://t.co/OzRPY6t2Z1— Emily Robinson (@robinson_es) August 1, 2018
Perspectives of Busy People
Jeremy Howard, in fastai Lesson 10
You’ve got to be a bit careful, right? Because sometimes I get messages from random people saying “I’ve got lots of good ideas, can we have coffee?” I don’t want to… you know, I can have coffee in my office anytime, thank you. But it’s very different to say, hey, I took your ideas and I wrote a paper and I did a bunch of experiments and I figured out how your code works. They added documentation to it. Should we submit this to a conference? You see what I mean? There is nothing to stop you doing amazing work and if you do amazing work that helps someone else, like in this case, ok, I’m happy that we have a paper. I don’t particularly care about papers. But, I think it’s cool that these ideas now have this rigorous study. Let me show you what he did.
If you’re thinking about asking a stranger to mentor you, it’s a bit like asking a stranger to marry you.
Lisa Martinez of Google
If you’re in the room, add value.
Like any meaningful relationship, mentorships are long-term, and loaded with mutual respect and benefits on both sides. They take time to cultivate and nurture.
Tough Love from Becky Davis
Becky Davis gives some blunt advice on mentorship in her article, Unable To Find The Right Mentor? Maybe You’re The Problem
- Give before you get
- Build a relationship
- Commit to the process
Don’t do these
- Don’t use the word “mentor”
- Don’t ask a stranger
- Don’t ask to pick their brain over coffee
- Don’t ask for their blueprint, roadmap or strategies to success
- Don’t give the victim story
- Don’t quit
Barron Hall on Mentorship
Mentoring in a structured program is tough if the mentees don’t take an active role by showing initiative and striving to grow.
Finding time and commitment to the time is key to sustaining the mentorship relationship
It’s painful to the mentor when they do listen and can’t perform, or when they easily have potential but won’t listen. It’s painful when people want to mentor you for everything except the thing you actually need.
I had another mentor before, but both of us did not really know what to do. The first session was great but then we were not able to maintain a contact, because I did not have specific aims for that particular interaction. Also, many people are ok with being mentors, but have no clue how to do it.
Sheryl Sandberg Lean In | Chapter 5: “Are You My Mentor?”
I realized that searching for a mentor has become the professional equivalent of waiting for Prince Charming
The strongest relationships spring out of a real and often earned connection felt by both sides.
Studies show that mentors select proteges based on performance and potential. Intuitively, people invest in those who stand out for their talent or who can really benefit from help.
Mentors continue to invest when mentees use their time well and are truly open to feedback.
Peers can also mentor and sponsor one another.
Laurie Winkless on “free consulting services”
Hi salaried people. Please don't get annoyed when a self-employed person sends you their rates in response to your invite to a 3.5 hour meeting where you want to 'share ideas'. That's called consulting, and if I do it for free, I can't pay my rent. #freelancelife— Laurie Winkless (@laurie_winkless) July 29, 2018
Reshama Shaikh on “Mentoring Time Investment”
We need to move away from this idea of “free mentoring”. It’s valuable time, especially for women who are already active in the community.— Reshama Shaikh (@reshamas) November 9, 2018
If people want mentoring, aka coaching, we should begin a conversation on compensation. https://t.co/ny8Q72CLPc
Frankly, your description sounds like someone who wouldn’t or shouldn’t want to also take on the community service of mentoring a stranger. Sounds like a full plate to me.— Mollie Manier (@maniermk) November 8, 2018
Mollie Maniere on Mentoring Overload
I agree, I already do approximately a hundred times more mentoring than I received from senior women.— Anne Carpenter (@DrAnneCarpenter) November 9, 2018
What Does Sponsorship Look Like?, Lara Hogan
Want a mentor? Stop asking for one., Bethany Crystal
The Importance of Sponsorship, Emily Robinson
Give and Take Adam Grant on YouTube
Lean In, Chapter 6: Will you be my mentor?, Sheryl Sandberg
Data Science is Different Now, Vicki Boykis