How can you apply data communication principles to any industry?

To build rapport you should aim to understand as much as you can about the audience's industry before entering the room. Further, through the dialogue, tease out the nuances and challenges they face that are specific to their industry. That means understanding their key business objectives, the operating model, the jargon they use, their pain points, and their competitors (among others). Although data can be considered a universal language, to build meaningful relationships with your stakeholders you still need to delve deep into the specifics of their business.

Tailoring data communication to your audience is paramount. Their expertise, interests, and attention span should shape your approach. For tech-savvy audiences, delve into specifics with jargon and code; for laypersons, embrace visuals and metaphors. Understanding your audience ensures your message lands effectively, fostering comprehension and engagement. The right format, tone, and language stem from this knowledge, forging a connection that bridges the gap between data and understanding.

I actually think the first rule of data communication is: Don't talk about data - rather talk about the insight you have generated . The second rule is: Be clear on the real question you need to answer. Sometimes the question asked is not the one you really need to answer to make an impact. Almost always ask why the question was posed in the first place to ensure you're answering the real driving question being asked. Then know the audience you are addressing. – You need to understand your audience to tailor your message and critically understand their motivations for asking the question. Did their boss ask the question? Is there an urgent issue behind the request and how could the answer affect individual or collective reputations?

As a data person, you need to be very careful on how you communicate with the audience. AVOID jargons when :: participating in one way communication with an open group. :: early product/solution discussions. :: non-tech audience is 25% or more of total. Tech Answers are appreciated if :: documented with comments :: supported with visual aids :: in-line with customers [business owners, end users, etc]

While it's ideal to speak with the audience, the "rubber duck" technique which is generally applied to debugging, is very handy here in my experience. It's a great idea to try and explain the concept to an imaginary representative of your audience. I've found while speaking out the presentation or ideas, it's easy to identify many gaps in the explanations.

The audience needs to know what you have been looking at in the data. They aren't too keen on knowing how you're doing it, more on why are you looking at that particular data set and what have you found. It often helps if you have a hypothesis that you were trying to prove with the data. An unstructured business problem provides ample hypotheses to investigate. One aspect I have found immensely useful is to recap the insights/takeaways/next steps from the previous discussion to set the context for the current one.

The more important aspect here is, always lead a presentation with insights. In school, most of us learn to teach in pyramids where we share the foundational knowledge first, then important insights and finally suggested actions. In business (or real world), it's great to follow the inverted pyramid and lead with the insights first. If the stake holders have time for questions, you can go into further details otherwise continue. Don't get hung up on the fine-details!

Craft a story like a news reporter not an expert. – Invert what you did for your thesis or dissertation. Think about your around three 'What's': 1) What? 2) So What? 3) Now What? 1) WHAT? - What is the headline framed around the exam question, but pulling out the insight that will make your audience want to carry on listening to you. 2) SO WHAT? - Share the insight and why it is (or will be) so important to your audience and why your insights are robust. 3) NOW WHAT? – Draw your insight into a conclusion and translate it into recommendations/actions that your audience can ask their colleagues to follow-up and be clear how your recommendations help with the opportunity or challenge your insight has highlighted.

I would say that this point is closely tied to the one before. If I am a data analysis specialist, and after analyzing data and applying some formulas and making some mathematically-based prediction, and finally coming up to some conclusions, I need to expose those recommendations to the interested persons (stakeholders, managers, directors, etc.). But to be an effective communicator, it is required that I know and take into account their principal interests and concerning about the matter, for knowing what exactly I have to answer to, and what kind of solution they are looking for. This is like combining technical and personal abilities, to "translate" scientifically supported conclusions into an understandable language for others.

Selecting and organizing data is a pivotal aspect of effective data communication, and my experience as a Project Manager underscores its importance. Just as I've chosen the most relevant and reliable data sources and structured them into interactive formats, I've recognized that this step is akin to tuning an instrument before a performance. Crafting a cohesive narrative is akin to composing a symphony, ensuring that the data story flows seamlessly. Using headings, subheadings, transitions, and summaries acts like musical notes, guiding the audience through the data journey. In essence, organizing data isn't just about presenting facts; it's about orchestrating a compelling and trustworthy narrative that resonates with your audience.

Another way to test your visualisations would be: - Are the self contained? - If you remove all the text, does it tell the story? - Can the story be unrolled within 5 seconds of looking at the visuals? Ask the questions above in order and iterate until the answer is no. Most of the times, the visuals are supposed to tell the story and your presentation or commentary only adds to it. Ensure that it's a complete story without the comments.

Less is more - Use pictures not words and tell stories not data points. – A picture is worth a thousand words so focus on visualising your data. The chances are your audience will be time poor (because of diary/agenda pressures and/or disinterest in your field). So, unless you are submitting something or presenting to an academic journal/audience: - Tell your story around charts and visuals remembering the three ‘What’s’ above. - Illustrate your data with stories - examples to bring you points to life. The acid test is that your visuals and headline can tell the story on their own without any supporting text. - Limit yourself to 3-5 slides and a 15-minute presentation or 1 side of A4 paper if the audience wants a paper - e.g., Boards.

Another important (perhaps the most important for data communication across functions and different industries / domains) aspect of contextualizing is using Domain, industry or function specific jargon (eg Revenue or sales can be referred to as subscription revenue in companies like Netflix/Verizon and Product revenue in companies like Nike). Further contextualizing would refer to the organization and team's goals as mentioned earlier - use organization specific/team jargon (like referring to customers as clients/subscribers). This is very important in applied analytics industry.

Beyond improving decision-making, effective data communication fosters transparency and trust. You can explain how data was collected, analyzed, and interpreted, building confidence in your insights. In today's data-driven world, mastering data communication is essential for success, as it ensures your data doesn't remain a mere collection of numbers but becomes a powerful tool for informed choices and problem-solving.

In the last paragraph, I would suggest a correction: *(that) will help you ensure your data communication is of high quality and meets ITS purpose and audience's needs* RATIONALE: My understanding of this is that the important thing here are not "my" needs, but my audience's needs, that's why I would refer to "its purpose" (the purpose of the communication) instead. Another observation is that "data communication" is written three times in the last paragraph only.

Remember this is a life-long skill and one you will have to keep refining as you progress. The best learning is by doing and it's always a great idea to practise by applying. Kaggle Notebooks and datasets are a great place to test and sharper your presentation skills to a large audience. Happy learning!

It's important to define what Data Communication it is not. It's not only data visualization and data storytelling. It is being able to recognise terminology, main involved data roles and main data science processes in an industry. It's being able to work with the teams in the line of business and in the centre of excellence. You must know their main attributions and responsibilities. So you communicate effectively. Ideally, you know what roles typically build data pipelines. You don't ask a data engineer to provide an opinion about ethical concerns of a machine learning model. There are specific contributors for these issues. You build lasting relationships because you know the internal audience's pains and limitations.

Remember you are never presenting your work, rather you are telling a story. Telling stories is a critical skill of all data scientists but it takes lots practice. Make it your default communication style and always ask for feedback. Jesus told parables to land His Message. Albert Einstein, one of the greatest scientists of all time and probably the only physicist ever to become a household name, is also one of the most quoted scientists in history because he was also a creative written communicator. Follow these rules won’t make you an icon like Einstein but they could help transform the impact your work and you have on your audience and elevate yourself from presenter on data science to data scientist and sought after speaker.

Applying data communication principles depends upon the industry. Industry such as healthcare where data security and accuracy is important or manufacturing where automation and quick decision is important. Thus data communication principles should be implemented in various business domains by understanding the 360 view of the business.

My strong advice is to start with data ethics. Define right and wrong concepts in a transparent manner, this will build trust and true involvement. It is one thing when you say that you embrace emerging technologies to drive success and very different thing when you say the same but with reference to strong ethical fundament. It is even more important when it is about data. This approach will help you both to comply with existing legislation on data privacy (and deep understanding of the matter), and provide a responsible and controlled step towards unprecedented challenges of data driven decision-making.