You’re a data scientist working with non-technical team members. How can you get your point across?

Dive into understanding your audience—whether they're decision-makers or collaborators. Once you’ve cracked their needs and data familiarity, tailor your message accordingly. Then, captivate them with relatable examples and visuals. After understanding them, craft a story that resonates with their goals, making your data come alive in their world.

You don't want to sell a comb to a bald man. Tailor your communication precisely to your audience. Notice what terms they use. What they speak. And use it to win them. People really groove on that personal touch, and gives them them a sense of comfort. That's my ace move, works like a charm from SMB's to Enterprises.

In the corporate world, knowing your audience and what you are communicating with them is important. Data scientists often work with complex algorithms that use many statistics, math, and ML concepts. They build ML systems that process, transform, and predict something. Communication becomes the key when such a use case has to be presented to stakeholders. As a data scientist/ML engineer you need to know what they would like to know and how they would value it. So, explain the business problem well and how your use case would bring value to them. Tailor your answer in such a way that it brings in analogies, charts, sample data, etc. This is a crucial skill to build through experience. It's important to complete a data science project.

Being able to explain data science concepts to non-technical colleagues is a crucial skill in the analytics field. This skill facilitates the transfer of ideas across departments and generally improves the speed and efficiency of project execution. Some ways to improve DS communication skills are: 1. Identify the audience and speak their language. Avoid technical jargons 2. Use appropriate visuals to summarize insights and analogies to simplify technical concepts. 3. Focus on the results and clearly articulate the implications of the analysis and how they relate to the broader goals of the organization. 4. Create an interactive and engaging communication style and be ready for questions and feedback.

In the world of data and analytics it is super important to be able to tell a story. Your story should describe the motivation, steps taken to achieve this and the benefits. Through these three use simple yet effective communication that executive leaders are able to understand without using too many technical terms. If the information is not understood, it is simply ignored. Bring key facts with numbers, graphs and examples to prove the benefits and outliers that impact your data or possiblities to derive insights. Some times its not the best data scientist or data engineer who makes a difference but also a good story teller who can stitch all the ends together.

Simplifying data is like translating a complicated language into one that everyone understands. Use visuals—charts, graphs, or tables—to paint a clear picture. Highlight the important bits, cut out the noise, and keep it crystal clear. Consistent colors and labels act as signposts in the data maze. Explain the relevance of your data, making it easy for everyone to connect the dots effortlessly.

Make your data as easy to digest as Sunday brunch. It's not about dumbing it down; it's about making it crystal clear. You've got this treasure trove of data, right? But if you throw all those numbers and stats at your team, it's like a fastball they can't catch. Break it down. Use simple, everyday language. Think of it like explaining a recipe – you wouldn't start with the chemistry of baking, you'd talk about the ingredients and the steps. Same thing with data. Keep it straightforward, focus on the key points, and before you know it, everyone's on the same page, ready to cook up something great. Remember, the goal is to make your data as approachable as a friendly neighbour, not as intimidating as a rocket science textbook.

Data can tell some amazing stories. They say an image is equal to 1000 words. You have the data, by the power vested in you by the visualizing libraries, you can create a picture out of the data ??. To accompany the visualization, prepare a story that helps you connect the dots for applying "Data Driven Solutions" for the business problem. The story should contain below points, ?? what is the business problem? ?? what data do u have & from where? ?? how can ur data be used for solving the problem? ?? what insights did u got from the data? ?? what other solutions are out there and why is your solution better? What impediments can one expect? If you are using jargons make sure to give an example so that they can easily understand it.

In my opinion one of the most underrated and most frequently not so well done things I see. There is just too much information displayed. We often try to get as much in as possible. But less is definitely more and we should actively work against our inherent bias of making things too complicated because we think more is better. It should be simple and clear. No clutter. Also we should "get the point" of reports quickly. If this is not clear by briefly looking at it, I think it should be simplified.

Understanding what message you are trying to communicate is very important when working with non technical stakeholders. By understanding what you are trying to communicate, you can simplify your data to only include the parts that help get your point across. Come up with a clear and concise answer to the business question, use a simple chart to support your answer, and do away with the rest.

In my experience, data-driven storytelling is key for engaging any audience. Start by setting the context: outline the problem, your analytical process, and the solution in a few points by focusing on presenting only the essential information in a fluid narrative. Use simple language to explain challenges and solutions, avoiding technical jargon. Critically evaluate your work and seek diverse feedback to uncover new insights or improvements. I often use ChatGPT for this ;) Finally, present your findings and recommendations with clear, easy-to-understand visuals like charts and graphs, making complex data accessible and engaging.

The important part is story telling. And, we can’t start with X1, X2, X3… Y. First try to understand the problem, then data (available and appropriate), ask questions to set the right contexts and define the scope of work properly. Now create a story board, like user and data flow. Spend more time with the client or domain expert. And, finally - don’t fall in love with your own algorithm :) You may have to dump based on feedback!

Stories are the easiest way to put your point across. Start your story by sharing the estimated impact of implementing the data-driven solution you're recommending. By establishing what is in it for them is the easiest way to get their attention and keep them interested. This exercise also helps you, as a data-scientist, to view the problem from the business point-of-view and maybe even improve your positioning, approach or solution.

Some pointers that have worked well for me in terms of storytelling: 1) Do your homework, analyze your data, and obsess over the problem. 2) Understand what is essential for your stakeholders, e.g., complex numbers, beautiful visuals, or something else. 3) KISS - keep it simple stupid, i.e., don't use over-complicated language or visuals. Sometimes, a simple annotated bar chart takes you a long way. 4) Less is more. Each visual, graph, or slide should have a story point around it. If not, hide it or put it into —the appendix. 5) Include an executive summary with your key findings and recommendations. 6) Do dry runs to get feedback on whether you get all your points across or not.

Avoid technical jargon and use plain language to explain concepts. Analogies and real-world examples can make complex ideas more relatable.

In AI and Data Science, we often get our ideas from stuff we see in nature. Think about how neural networks mimic the human brain, or how the bees algorithm gets its cues from actual bees. Sometimes, it's just about plain old simple logic that anyone can get. So, your communication should draw from real life example or analogies. As a Data Scientist, you're going to be working with people from all sorts of fields. The trick is to keep things simple. Why? Because when everyone gets what you're saying, they can give better feedback, and everyone in the team, no matter their background, can chip in. A lot of tech folks love using big, fancy words. Maybe they think it makes them sound smart or efficient. But honestly? That's not the way to go

I love to do this always :) e.g. “You have a weighing machine and a measuring tape. You took the heights and weights of say 30 people. Now, the weighing machine stopped functioning. Can you predict the weight of the 31st person by just measuring the height? If yes, how?” Before I start explaining the Liner Regression! And, similar approach for all algorithms!

If you are familiar with the domain of your audience, it could be beneficial to use analogies by connecting insights to their field of expertise so they understand the findings easier.

I've discovered that using everyday analogies to explain difficult concepts works well. Pick analogies that are relatable. For example, I use currency conversion to explain the importance of data transformation and normalisation. If you've received job offers from 3 different countries, in different currencies, you need to convert them into the same currency in order to compare which job is going to pay you the most. People will usually understand that the data (or money) needs to be converted into a common scale (or currency) in order for ease of comparison.

Analogies make complex data concepts graspable. Examples: - Counterfactuals: Without a counterfactual group in pre vs. post comparisons, like another store without a marketing campaign, we might incorrectly credit all sales changes to the campaign, overlooking factors like unanticipated seasonality and trend. - Non-experimental comparisons: Comparing clients simply based on the use of a feature is as flawed as comparing gym-goers' health without accounting for lifestyle differences. Experiments overcome this. - Mean and outliers: The mean's vulnerability to extremes is like averaging the temperature with one's head in a freezer and feet in an oven; it misrepresents reality. In business, a few large transactions can skew the average.

Two-Way Communication: Encourage questions and discussions. Create an open environment where non-technical team members feel comfortable asking for clarification or expressing their thoughts. This exchange not only enhances understanding but also encourages collaboration.

Always be one step ahead. Think like a chess player. What questions will they throw at you? What might make them balk? Have your answers ready, and not just any answers – make them clear, concise, and relevant. It's like having an ace up your sleeve. When you're ready for their curveballs, you'll hit it out of the park every time.

You should always foresee the possible questions and objections that could arise from your audience and you ahould be prepared to answer them in a concise and clear matter so that there will not be any objections left.

Start educating all, AI is not magic. It’s just pattern based learning. Let the non-techies learn from us first before we teach the machines. Throw parties and have funny discussions on AI / ML / DS to clear the concepts (and sometimes misconceptions).

- Understand who your audience is and what they are looking for. - If they do not need to know the minutia of how something is achieved, leave it out. - Stick with a simple bottom line. - Understand the person’s background. Work with what they already know to build deeper understanding on a topic. - Stay relevant to them to keep them engaged. - Stay away from multiple new acronyms. Use plain English when possible. - NEVER treat a person like they are stupid for not knowing something. They are not stupid, they simply do not have the same job as you, which is a good thing. They are smart in different ways. - Be direct. - Stress only one or two main takeaways at a time.

Como analista de datos, transmitiría mis puntos de vista de manera clara y efectiva utilizando visualizaciones de datos impactantes, presentaciones concisas y argumentos respaldados por evidencia sólida. Además, me aseguraría de comunicar de manera coloquial, evitando jerga técnica innecesaria para que mi audiencia pueda entender y apreciar plenamente mis análisis y conclusiones.