K#1. The skill you don't know you need
How to deal with Product Managers
Heyyy 👋
Welcome aboard this first edition of K’s DataLadder ✨ – the quick & digest newsletter where I help you climb up your data science ladder 🪜
Each week I bring you one story straight from my life as a Tech Data Scientist and the learnings that come with it. Here, we level up our data science game together!
Also, it’s only the beginning but we’re already 679 people in this newsletter!
It’s all thanks to readers like you, so thank you so much❤️
If you haven’t done that already, you can also:
Join the 4.6k followers of my Medium blog
Check out my latest data science piece, already 20k reads 😮
Lezgoo 👇
Today’s Agenda
This week’s story
The skill no one taught you
Why it matters so much?
5 tips to develop the skill
This week’s story
Recently, an interesting experience taught me how to better deal with product managers (PMs) and why it’s so important.
For the sake of the story, and because I don’t want Spotify coming after my butt, we’ll keep things confidential of course and we’ll simply call the PM: SpongeBob 🤓
So like I said I was working closely with SpongeBob on a project related to analyzing metrics for a super cool new feature.
When my data science peers handed me down the project, I was given the context, underlying goal, and metrics to analyze.
I often exchanged with SpongeBob on the results of my analysis. But on the day I shared with him the final deck, he wasn’t fully on board with my storyline.
SpongeBob had a specific narrative in mind that he needed the data to support.
But the way I visualized some of the metrics painted a slightly different story than the one chilling in his mind.
Sooo he requested a series of visual changes – which I did.
But despite the back and forth, and all my efforts to adapt the presentation of the data to his story (which the executive summary was already aligning with, except for that one metric), we still struggled to find common ground.
We had different views on what the data was showing us.
And that was problematic.
The skill no one taught you
We often think that being a data scientist means running ML models all day, analyzing data, and sharing the findings with key stakeholders.
While it might be true, hard skills only make up a fraction of what you spend your time doing in practice.
In the pyramid of skills that matter the most, soft skills rank higher than hard skills; they have the most impact on the fate of your work (and yours too).
I’ll tell you why very soon.
They’re also more challenging to learn – there’s just no handbook to learn those.
Also, you’ve probably heard a lot about storytelling, communication, and whatnot, but some skills that are far more important and trickier rarely get spoken about, skills like:
Dealing with non-tech stakeholders, especially product managers.
Why it matters so much?
1. You can’t escape it
In tech companies, product managers and data scientists work hand in hand together. We’re like SpongeBob & Patrick. It’s the stakeholders we collaborate the most with.
So your ability to communicate and negotiate with non-tech folks, especially PMs, directly impacts the success + direction of projects – and yours too at the same time.
If they’re not convinced, they might not push your findings forward.
If the story they have in mind doesn’t align with yours, they won’t let you be.
So you must learn to influence others and advocate for your work to show that:
Your work is meaningful
Data science is useful
and you with it.
It’s the most valuable skill a data scientist can have as he/she progresses over time. I learned that during my first year at Spotify.
2. You’re responsible for what happens to the data
The data part is your responsibility – from the moment you pull it out to the moment it ends up on a roadmap driving future development.
It becomes your duty to make sure that non-tech folks don’t distort the data.
At times, they might misinterpret your findings, or use the wrong wording when sharing those because they might be under the influence of whatever story lives in their mind – story which you might have contributed to with your poor viz or communication skills.
It also becomes your job to make sure the interpretation of the data remains true to its actual insights, despite narrative and time pressures – which happens a lot.
The numbers are absolute but the interpretations are infinite. It’s your responsibility to make sure the right narrative – the interpretation that is the most faithful and representative of the data – is the one that gets pushed at the end.
3. Your professional development depends on it
Learning to deal with stakeholders effectively is a key skill for career advancement.
You have to show that your work contributes to the business and you have to scream that loud enough too.
Something I learned quite late is that if you want to move up in your career within the company you need to get noticed, especially if like me you’re team work-from-home. Something that makes total sense, but isn’t always quite obvious.
Getting your work highlighted by PMs to the decision-makers is crucial. It provides solid proof of your achievements for when you're looking to push your promotion case.
5 tips to develop the skill
Tip #1: Understand the PM's goals
I made a mistake when I took the goal from my peers without clarifying it with the PM again. Information can get lost when it gets moved from one person to another.
So make sure in the start to have an deep discussion with the PM to understand the objectives, expected outcomes, and the narrative they wish to explore.
This early alignment ensures your analysis matches the PM's vision because even the smallest misunderstanding can significantly impact the direction of your analysis.
Tip #2: Collaborate & communicate non-stop
Keep the PM updated on your findings as you progress. This allows to adjust expectations and explore different narratives that the data might support.
Don’t wait until the end, do it continuously to avoid editing big stuff last minute.
Tip #3: Acknowledge their suggestions
When SpongeBob suggested a viz approach that didn't mathematically make sense, it was still important to acknowledge his idea, even though I knew it was wrong.
So I took the time to educate him on why the approach was incorrect and the importance of balancing data integrity with personal narratives.
This builds a foundation of trust and understanding.
Tip #4: Consult your peers or mentors
Validate your analysis by consulting with your data science peers or mentors.
This step helps you make sure that your interpretation is accurate and unbiased – even when others might question it.
When the PM questioned my viz, I shared the issue with my peers to get an external opinion – hell I even asked ChatGPT how to deal with the issue. Whatever helps.
Tip #5: Framing is everything
When the PM still disagreed with my edits, one of my peers advised me to reframe some of my insights in a more positive light.
Sometimes, it’s not a matter of changing the data, but reframing your interpretation in a way that is both true to the data and potentially supportive of the PM's narrative – where appropriate.
Remember when I said “numbers are absolute but the interpretations are infinite” – I learned it from here.
It’s like rearranging furniture in a room. The same pieces, but a different layout can make the room much nicer.
In Brief
Continuously discuss and shape findings to fit the end goal both you and the PM are trying to reach.
The numbers are what they are. PMs usually come with an idea in their mind they want to prove but it’s part of your job to make sure data integrity is respected.
Ultimately you make your own role, you just need to make sure that your role is not reduced to just following others’ directions. Stand your ground.
Awesome content and value, on top of great writing!
" SpongeBob suggested a viz approach that didn't mathematically make sense, it was still important to acknowledge his idea, even though I knew it was wrong. "
sound more like patrick to me ;))
Thanks for the tips tho :3