Good morning, friends!
Today, we have something special for this newsletter. Recently, I have been speaking to
, who was a lead data scientist and ex-Meta, about her transition from corporate to solopreneur life.Mandy's wealth of knowledge and experience in data science is not just valuable; it's directly relevant to every data professional.
That's why this edition of Dishing The Data is a guest post by Mandy! Let's dive in!
Mandy has a free, weekly newsletter called The Curious Techie for Tech friends seeking to advance their career, become more productive, and build growth and income outside their 9–5s.
Becoming a data scientist isn’t just about crunching numbers — it’s about navigating a minefield of potential missteps.
On my journey, I made some blunders that were not only painful but also incredibly enlightening.
Here are my top three mistakes, so you don’t have to learn the hard way.
1. Concealing weaknesses
I used to believe I had to excel at everything to be the perfect data scientist — the one who knew every detail about A/B testing and the latest ML models.
But it wasn’t possible.
The data science field is too vast to master it all. Clinging to the idea of perfection, I hid my weaknesses and never admitted my gaps in knowledge to my peers.
Then I met Jason, a senior data scientist at Meta.
Jason was incredibly knowledgeable and supportive. One day, as we prepared to launch an A/B test for a feature, he approached me and said,
“My strength is in ML models. My last team didn’t need experimentation, so my Meta-specific A/B testing knowledge is lacking. Can you lead this effort and let me learn from you?”
Wow. He is an actual human being with weaknesses, I thought.
He delivered it so gracefully, and it had a profound effect on me:
I felt a confidence boost knowing he trusted me and relied on my expertise in experimentation.
I was impressed by his self-awareness and his choice to lean in on his strengths while prioritizing learning.
I realized that admitting weaknesses made him more trustworthy and reliable.
This taught me the importance of authenticity.
Later, I applied Jason’s lesson to my own work.
I admitted to a colleague that I was rusty in building ML models and asked for a weekly catch-up for his advice on building a propensity score matching model. During peer reviews, he told me he was impressed by my self-awareness, my willingness to admit weaknesses, and my proactive approach to seeking help.
Concealing my weaknesses was a major mistake. Once I understood that “having weaknesses” isn’t a weakness itself, I felt freer from imposter syndrome and adopted a growth mindset. It enabled me to demonstrate self-awareness, build trust, and cultivate allies.
In today’s fast-evolving tech landscape, we need authenticity more than ever.
2. Failing to fully own your projects
When starting our careers, we’re often told what to do rather than defining our own work and ensuring it gets the visibility needed to progress.
That approach quickly becomes insufficient as we advance.
For any project we initiate or that falls into our jurisdiction, understanding how the project goal aligns with the wider team’s objectives is crucial. If we don’t actively seek alignment from partners, we risk unnecessary friction.
Once, I took over a project from a senior colleague. She had already started the analysis, so I assumed she had communicated its purpose to the broader team. I thought everyone knew what was coming.
But when I presented my analysis and recommendations, the engineering manager’s feedback stunned me:
“We like the analysis and insights, but we wish Mandy had brought us on board from the beginning, so the process could be more iterative and the Engineering team could prioritize her recommendations and act straight away. Right now, we have a jam-packed roadmap - it’ll be hard to revise.”
I was shocked.
I hadn’t anticipated that the team might not be fully aware of the analytics work. Since the project was passed on to me, I assumed my predecessor had done all the initial alignment.
I assumed. I assumed. I assumed.
Turns out, assuming is a mistake. Once a project lands on your desk, it’s fully yours. Not your predecessor’s. Not your manager’s.
From that moment on, I realized I needed to own my work completely. I had to challenge every assumption, align goals and priorities with every involved partner.
Eventually, the engineering team implemented my recommended changes, but with a three-month delay.
The lesson? Question everything until you achieve abject clarity.
Otherwise, your analyses will end up gathering dust.
3. Presenting graphs without proper context
This mistake is a personal pet peeve of mine, and it’s surprisingly common.
Picture this: you’re in a presentation, and the presenter lands on a slide with a beautiful, complex graph. They say,
As you can clearly see…
And then rush to conclusions (or worse, offer none).
All I want to do is shout:
No, I cannot see clearly at all!
They waste a detailed, information-packed visualization by not giving the audience time to absorb it.
So, here’s what to do instead:
Read the title of the graph or summarize what the audience is seeing in one sentence.
Call out what’s on the horizontal and vertical axes (avoid using “x” and “y” axes for clarity)
Share one or two key insights from the graph.
Clearly state your recommendations.
For example, if you are presenting a marketing cost comparison:
You can say:
The graph on the left shows cost per acquisition among Facebook, Google, and TikTok, while the graph on the right shows cost per retained acquisition after 30 days.
The horizontal axis displays the three channels, and the vertical axis shows the cost in USD.
You can see from the left graph that the cost per acquisition is highest on Facebook at $4.50, but from the right graph, the cost per retained acquisition is lowest on Facebook at $7.50.
I recommend we primarily market on Facebook to acquire high-quality users.
Now you’re doing justice to that painstakingly crafted seaborn graph with a color palette that you spent half an hour googling and perfecting. 😉
Take-aways
The three big mistakes that stand out on my journey as a Data Scientist:
Concealing weaknesses: Hiding what you don’t know only holds you back
Failing to fully own your projects: Every project you touch is yours to drive, align, and deliver.
Presenting graphs without proper context: Great visuals need clear explanations to be impactful.
Embrace your vulnerabilities, own your work, and communicate effectively.
You might also enjoy this: 3 Years at Meta: Transformative Data Science Lessons
Weekly Favourites ❤️
📚 Book — Million Dollar Weekend by Noah Kagan. I started reading this today after seeing several recommendations online. I was skeptical, as the title is quite "clickbaity," but after reading a few chapters, I realised it's far from it, and all the advice is really tangible. You are not actually going to get a million dollars at the weekend (obviously), but you will gain ideas and processes that will help you achieve it over the long term.
🎙️ Podcast — How To Start A Business In 48 Hours—Noah Kagan. Ali Abdaal interviewed Noah Kagan on one of his podcasts, so I had to watch it. The conversation also revolves around the Million Dollar Weekend book but delves into other things, like growing a YouTube channel.
📩 Newsletter — To Be a Data Scientist. A great newsletter written by
who has been a data scientist for over 6 years. If you want top-tier data science advice, then make sure you check it out!
(PS: Some links are affiliate links that I get a kickback from with no extra cost to you 😎)
My Latest Content 🎬
You can reach me on:
LinkedIn, X (Twitter), or Instagram.
My YouTube Channel and Medium Blog to learn technical data science and machine learning concepts!
💡 If you are interested in sponsoring this newsletter see here.
Telling a story through visualization is such an important skill and so underrated. Initially in my career not having the right information to tell through the graphs led to lot of unnecessary confusion in the project presentations. Thanks for sharing this insights.
These insights are great! “Assuming” was also a problem of mine early in my career. Not falling into that trap helped me accelerate my career development.