Effective Communication: A Key to Success for Data Scientists
Written on
When I began my journey in data science, I was intensely focused on mastering Python, Machine Learning, and Statistics. While these technical skills are crucial, there is another vital ability that many professionals neglect—communication. In this article, I will explain the necessity for data scientists to excel in this area and offer seven helpful strategies that have aided my development.
During my graduate studies, the Physics department hosted weekly colloquia, featuring guest speakers presenting their research. It was common to see the audience grasping the first slide (the title) and perhaps the second (the agenda), but they often became confused afterward.
This phenomenon is mirrored in data science, where non-technical stakeholders often struggle to follow presentations given by data scientists. They might say, “I understood it until you mentioned train-test splits and AUCs.”
While this may seem like an unavoidable aspect of data science, I have discovered that enhancing the clarity of these discussions is not only achievable but crucial for career advancement in this field.
In this piece, I will share the essential communication strategies that have helped me secure promotions, attract clients, and explain artificial intelligence concepts to over 100,000 individuals.
The Significance of Communication
The significance of communication may surprise some and might even be met with skepticism. Allow me to elaborate.
Data scientists typically do not resolve their own issues; instead, they tackle the challenges faced by others (i.e., stakeholders). This is the source of value generated by data scientists in a business environment.
Thus, the level of value a data scientist contributes is directly related to their ability to effectively collaborate with non-technical stakeholders. To put it simply, if stakeholders do not comprehend and embrace your solution, it adds no value.
Communication is a Learnable Skill
Some may believe that communication is an innate skill that one either possesses or lacks. However, this is a misconception. Communication (like any other skill) can and must be cultivated through practice.
For example, I began my journey as an overly technical physics graduate student, but after five years of delivering presentations, writing articles, creating YouTube content, hosting events, interviewing entrepreneurs, and offering technical consultations, I now receive recognition (and compensation) for my communication abilities. If I can achieve this, so can you.
7 Essential Communication Strategies for Data Scientists
The following communication strategies are the ones I utilize most frequently. While I am focusing on technical presentations, these tips can be applied to discussions, writing, and other forms of communication.
One advantage of enhancing this skill as a data scientist is that the standard is relatively low, meaning that even becoming a competent communicator can set you apart from many of your colleagues (I can personally attest to this).
Tip 1: Utilize Storytelling
Storytelling is one of the most impactful ways to convey information. Our brains are naturally inclined towards narratives. The more you can incorporate stories, the better.
When I refer to "story," you might think of its conventional definition, which is an account of imaginary or real events and characters. However, I mean it in a broader context, as outlined in the book The Storyteller's Secret.
The author defines a story as any three-part narrative. Examples include:
- Status quo. Problem. Solution.
- What? Why? How?
- What? So what? What now?
Here's how the first example looks in practice:
> AI has revolutionized the business landscape (status quo). While its potential is evident, translating it into specific, value-generating business applications remains challenging (problem). In this article, I will outline five AI success stories to guide its implementation in your business (solution).
Tip 2: Provide Examples
Data science is rife with abstract concepts that often seem disconnected from everyday life (e.g., features, overfitting, curse of dimensionality). A powerful method to make these abstract notions relatable is through concrete examples.
For instance, if a stakeholder asks, “What’s a feature?” your instinct might be to respond with a definition: “Features are what we use to make predictions.” However, this is a rather vague explanation.
A clearer approach would be to follow up with a specific example: “For example, the features in our customer churn model are Account Age and Number of Logins in the Last 90 Days.”
Tip 3: Employ Analogies
While examples are effective, sometimes they might not suffice. This is where analogies come into play. Analogies are valuable because they connect the familiar to the unfamiliar.
For instance, when I had to explain Mechanistic Interpretability to a non-technical client, I encountered a daunting term (even for data scientists). Here’s how I framed it:
> Modern AI systems like ChatGPT are incredibly powerful, but we don't fully understand their inner workings. The concept of Mechanistic Interpretability involves examining these systems to understand what different components do.
By likening a large language model (unfamiliar) to a car engine (familiar), the abstract idea becomes much easier to grasp.
Tip 4: Use Numbered Lists
In a multitude of ideas and words, numbers often stand out. This makes them an effective means of presenting information.
For example, I’m utilizing this strategy to structure the seven communication tips in this article. This technique goes beyond the typical list you might encounter online.
Another way to employ numbered lists is when making multiple points in your communication. For instance, I want to convey two points here: 1) numbers attract attention, and 2) they provide a clear framework for organizing information.
This approach works well because numbers like 1, 2, 3, etc., are simple and familiar concepts that require minimal cognitive effort to process.
Tip 5: Embrace Brevity
> “I didn’t have time to write a short letter, so I wrote a long one instead.” — Mark Twain
This principle embodies the essence of effective communication. Your audience has a limited amount of attention to offer. Therefore, as communicators, we must be judicious in how we use that attention.
While it may seem that using fewer words would mean consuming less time, the opposite is often true. Distilling ideas into their most essential forms requires numerous iterations.
This could involve reducing the number of slides in a presentation, minimizing the elements on each slide, or even cutting down the character count in the title.
Here are some guidelines I follow in a business context:
- Keep presentations to 20 minutes or less (approximately 10 slides or fewer).
- Limit each slide to 3–5 elements.
- Make bullet points as concise as possible.
Tip 6: Visualize Instead of Verbalize
An extension of the principle of brevity is favoring images over text. Processing text demands more cognitive resources than interpreting visuals, making images an exceptionally effective means of conveying ideas while preserving the audience's limited attention.
Here’s the fine-tuning analogy from Tip 3, contrasted with a visual representation of the same concept.
This underscores the significance of data visualizations. Although this topic warrants its own article, it shares the foundational principle of brevity.
Tip 7: Take Your Time
This final tip was transformative for me. In the past, I often rushed through my presentations, likely due to nerves and a desire to finish quickly. Eventually, I realized that my nerves would diminish if I slowed my pace and adopted a calmer tone.
Slowing down not only enhances the audience's experience but also creates a more enjoyable presentation atmosphere. A rushed talk can feel overwhelming, while a well-paced one resembles a gentle stream. Thus, a brief, hurried presentation can be more taxing than a longer, well-timed one.
Bonus: Understand Your Audience
While the strategies mentioned can lead to immediate enhancements in communication, their effectiveness will be limited if the message is not tailored to the audience. This highlights the importance of empathy.
Empathy involves understanding things from another person's viewpoint. It is vital for effective communication as it provides the context needed to frame all aspects of your presentation.
The more you can empathize with your audience, the more effectively you can address their interests and comprehension.
Conclusion
For many data scientists, the primary constraint is not their technical expertise but rather their capacity to communicate effectively. Developing this skill is one of the most beneficial steps a data science professional can take to further their career and make a more significant impact.
Here, I shared seven strategies that have greatly assisted me in refining my communication. If you have additional tips that have worked for you, feel free to share them in the comments!
Resources
Connect: My website | Book a call Socials: YouTube | LinkedIn | Twitter Support: Buy me a coffee
The Data Entrepreneurs
A community for entrepreneurs in data & AI. Join the Discord! medium.com
[1] The Storyteller’s Secret: From TED Speakers to Business Legends, Why Some Ideas Catch On and Others Don’t by Carmine Gallo [2] Oxford Languages. (2024). Story. Retrieved June 11, 2024, from https://www.google.com/