How to Talk About Data and Analysis Simply

Not long ago, I stepped into my backyard after months of neglect, and it was a complete disaster. Broken chairs, scattered tools, forgotten toys, and piles of leaves cluttered the space – junk everywhere I looked. It felt overwhelming, and I wasn't sure where to begin. I stood there thinking it would take an entire weekend, maybe even professional help, to get it sorted. But then I grabbed a rake and started in a tiny corner. Little by little, things began to fall into place. After an hour, the chaos didn't seem so daunting. That moment reminded me how easy it is to overthink a problem when, in reality, all it takes is starting small and tackling one piece at a time.

It's the same with data & analysis: it can seem like an overwhelming mess, but it becomes manageable once you break it down and address it step by step.

What is in this post?

Simplifying Communication has always been a challenge for me. I focus on details and nuances, often losing sight of the bigger picture. While this can be useful when discussing with fellow data experts, it becomes a bias, especially when communicating with high-level stakeholders.

In this post, I'll share my five-step strategy to simplify communication and quickly improve data-driven communication.

Isn't storytelling the answer?

I recently read a book titled "Storytelling or Simple Communication" [1]. According to the author, there are two types of professional communication. One is storytelling. Its ** data-driven variant is very popular nowadays. It involves using stories to convey information and influence listeners. The second type, simple communicatio**n, is direct, plain, and focuses on clear facts. In this framework, these two types are separate concepts. Yet, it's possible to combine both of them. For instance, when presenting insights, one can start with an anecdote (storytelling), followed by facts (simple communication), then show some case study stories (storytelling), followed by facts (simple communication), and so on.

I find this approach too complex. As data professionals, our goal should be to present insights clearly, whether we use a story or not. Also, employing storytelling does not automatically mean you become more descriptive and less factual. I believe you can still use storytelling yet keep the communication simple and understandable.

How much one can take…

… as far as information absorption is concerned? Not that much.

According to Miller's Law of Cognition, short-term memory typically holds around seven pieces of information, give or take two [2]. This suggests that when we attempt to present more than seven ideas at once, we risk poor retention or misunderstanding.

For data experts, the challenge lies in selecting which ideas to prioritize, especially since everything may seem necessary. The real task is determining what is essential and how to order and highlight those more critical pieces.

Based on my personal (and painful) experience, we often lack the luxury to craft complex Presentations that intertwine anecdotes and stories with facts. Building suspense, using sublime, technical words and multi-line sentences are neither options. We need to get straight to the point, and we don't have much time to do so.

My strategy to simplify communication consist of 5 steps:

• Step 1: Decide on the key points
• Step 2: Structure communication
• Step 3: Design
• Step 4: Simplify
• Step 5: Test & cut

Step 1. Decide on the key points

We now know the rule: the 7-point limit. But what if we feel that this number should be bigger? How do we select the points that matter and help us proceed with our project?

Learn your audience and what they need to know.

Last year, I wrote about the crucial role of context in storytelling. A key element is knowing your audience – who they are and what matters most to them.

Power of Context in Data-Driven Storytelling

This means prioritizing the most relevant information to your audience's needs. The points that are important to them should rise to the top of your key messages list. In a corporate setting, these priorities often revolve around revenue, profit, sales figures, cash flow, number of faults in the production process, etc.

Prioritize based on importance to the business.

But what if all our insights revolve around a single metric, such as sales or revenue? In that case, we should prioritize them based on business impact. For example, insights concerning the most profitable or best-selling items should go up. Factors like growth acceleration (or deceleration) should also be weighed when ranking importance. Prioritizing insights in this way ensures we focus on what will drive the greatest value for the business.

An effective approach would be to utilize strategic analysis tools such as the BCG Matrix [3], which classifies a company's products along two key dimensions: growth rate and market share. For instance, products identified as "Stars" (high growth, high market share) represent future profitability, while "Cash Cows" (low growth, high market share) contribute significantly to current profits. These categories should be prioritized, as they have the greatest impact on the company's long-term success. Conversely, updates related to "Dogs" (low growth, low market share), which generate minimal profits or even losses, can be relegated to the appendix. This ensures that attention is focused on the products that deliver the most value to the company.

Decide what's the minimum you want to achieve

Review your material and determine the core message or outcome you want from presenting your data or analysis. The key points that align with this goal should be presented first. This leverages the "serial position effect" [4], which suggests that people remember the first and last items in a sequence best, while the middle content is often forgotten.

However, based on my experience, leaving critical information for the end is unwise. Instead, use the "last" position to reinforce key conclusions or insights you've already highlighted, ensuring your audience leaves the most important takeaways fresh in their minds.

Use time as a limiting factor.

The topics you cover should be closely aligned with the available time. In my post, "How to talk about data and analysis to non-data people?" I introduced a formula for determining the optimal number of slides for a presentation. For example, this formula suggests a maximum of seven slides for a 20-minute time slot, including technical slides like the title and agenda. Add a summary or a takeaways slide and you are left with a maximum of four slides for your key points [5].

This timeframe could shrink further when accounting for potential delays, such as technical issues or waiting for attendees from other meetings. If the topic is likely to spark controversy, allocating extra time for discussion and questions is important. Anticipating these scenarios helps you stay within the allotted time while ensuring you deliver a clear, focused message.

Step 2. Structure your communication

Is a presentation/deck the only option?

Not necessarily. An effective alternative is the so-called Amazon 6-pager [6]. This approach changes how meetings are conducted using a collaboratively prepared document, for instance, one created in Google Docs. Meetings are divided into two parts: reading the document and discussing it. The reading period, one-third to half of the meeting, ensures everyone is familiar with the content. Collaborative features allow for questions and quick responses directly in the document. The discussion and decision-making follow. This method is especially useful for complex or controversial topics. A helpful tip is to add a timer in Google Docs to manage reading time.

Order insights/points of your story according to their importance

When organizing the order of your points, arguments, or insights, you should begin with the most important or the strongest one. While there's a natural tendency to tell stories chronologically, it's essential to develop the habit of prioritizing the most critical information first, followed by less significant details. This method, known as the "reverse pyramid," is commonly used in journalism [7].

Though some storytellers excel at building suspense, my experience suggests it may not work. In business, clarity and impact are more valuable than creating tension.

Tell your audience what you need from them right at the beginning

A strong approach is to begin your communication by clearly stating the goal of the document or meeting, particularly if the audience is expected to make a decision. This upfront clarity sets expectations and helps direct their attention to the information supporting or challenging that decision.

Additionally, providing a high-level "executive" summary at the beginning can be highly effective. This overview will be enough for some, while it will help others focus on the areas that matter most to them. It will help you structure communication logically. It will also allow listeners to reorient if they get distracted during the presentation.

One thought – one slide (or paragraph)

A key rule I recommend is limiting the number of distinct points on each slide or paragraph. The more you adhere to this principle, the more likely your communication will be easy to remember. If you plan to use the material for public speaking, this rule becomes even more critical to follow. However, if the material is meant for standalone reading, such as a so-called slideument, you can be more flexible with this guideline. So, how can you apply it? Here's an example of how to structure your presentation:

• Title slide
• Summary slide or action slide with 3–4 key points to be addressed next
• 3–4 slides, each covering one main point previously signaled on the summary slide
• Conclusion or recap slide.

Step 3. Design

Use titles

Titles are often underestimated in communication, frequently limited to technical phrases like "Research Results" or "Sales Forecast for Next Year." Instead, titles should convey key messages, such as "Research Reveals Opportunities to Cut Costs in Production" or "20% Sales Decline Expected Without UX Improvements."

This approach offers several benefits:

1. It allows highlighting critical information in the most visible section of the slide or document.
2. It sets clear expectations.
3. It optimizes space by delivering valuable insights upfront.

This principle applies to slide headings and visuals like charts and tables, where titles should reflect key insights drawn from the data, even guide interpretation or explain potential doubts.

Use two dimensions

Please compare these two slides:

I think the key points from the second slide are much easier to remember. This is because the main conclusions are highlighted and separated from the detailed explanations: moved to 2nd dimension. By making this adjustment, the audience can retain the important information while you still have room for more in-depth descriptions. It's a win-win situation!

Bring structure to your slides

Another quick design improvement involves creating a consistent structure for how you write comments or conclusions on your slides. Let me demonstrate this with another example. Here, the elements the slide author considers important (those in bold font) are scattered throughout the sentences inconsistently. By bringing more order and uniformity to the placement of key points, you can enhance clarity and improve the overall flow of the slide.

Here's the improved version. The key information is now positioned at the beginning of each point and separated from the rest of the text. This approach makes the content easier to read and helps the audience retain the most important details more effectively.

Align & distribute

A quick way to simplify and improve your presentation is to get familiarized (and become friends) with the "align" and "distribute" functionalities of your presentation software.

When somebody asks me to improve their presentations or documents, bringing order to the material is one of the first things I do. Taking care that all the elements are proportionally distributed on the slide and aligned with each other can bring much difference.

Take a look at the slide below. The possible excuse is that it was prepared in a hurry. But using it can cause more harm than good. First of all, it does not look professional. Second, even this little information, when shown unaligned, can cause headaches and prevent us from communicating important things.

People like order, proportion, and alignment. Even simple slides, when proportionately designed and with elements well-aligned, look professional and are more convincing. A helpful design rule could be the "golden rectangle", which aids in designing more pleasing images [8].

So, without further ado, let's improve the slide from above. We should not refrain from using different aids offered by our software: rulers, guides, or automated arrows.

The finalized product could look like the one below. It's proportional and aligned. Even though not much is happening in the slide, it's far easier to read and remember.

Visualize

While it might sound cliché, the effective use of visuals is crucial. Prefer clear and straightforward visualizations over dense tables or complex descriptions. The primary risk, reduced by the use of visualizations, is that people may not engage with or remember the information, thereby not act on critical insights. However, you should avoid overcomplicating visualizations with excessive metrics and dimensions crammed into a single chart. Instead, consider distributing the key messages across three or four visuals to enhance clarity and impact.

Step 4. Simplify

In this step we should review how we construct our messages in slides or documents. Emphasizing simplicity from the beginning of our projects is crucial. What best practices can facilitate clearer and more straightforward communication?

Mind the font

There are two types of fonts:

• Fonts suitable for presentations
• Fonts ideal for reading.

Take a look at the article font used for posts on Medium.com. It's a serif font. Serif fonts have small lines or strokes regularly attached to the end of larger strokes in a letter or symbol within a typeface. These little lines are called "serifs." Popular serif fonts include Times New Roman, Georgia, or Garamond. They are often used in print and for body text in documents because they are considered easier to read in long passages compared to sans-serif fonts.

Now, please take a look at the slide examples I showed earlier. I used various non-serif fonts. A "non-serif" font, also known as a "sans-serif" font, features clean and simple lines. Sans-serif fonts tend to have a more modern and streamlined appearance than serif fonts. Popular sans-serif fonts include Arial, Helvetica, or Verdana. These fonts are often used for digital display, such as on websites and in apps because they display clearly on screens and have a clean, readable style.

So the rule of thumb is as follows: if you write a document, use a serif font. If you prepare a presentation, use a non-serif font.

There are also some other minor rules you could follow. For instance, in many languages there are a lot of diacritics: little elements added to letters. They are common in many languages, including accents (like acute, grave, and circumflex), cedillas, tildes, and umlauts. In Polish, for example, you'll see letters with diacritics such as "Ś" and "Ę", each indicating a specific pronunciation. It's very important to select a font with such diacritics to avoid an effect like the one displayed below.

Lastly, it makes perfect sense to select a visually appealing font. It may make a huge difference to the reception of your presentation. My personal choices are Inter or Lato fonts. They are both elegant and readable. It's a matter of taste, yet I would refrain from using hackneyed fonts like Arial or Calibri.

Be consistent

A crucial principle is consistency in writing, listing, and emphasizing key points. If you typically use short sentences in your slides, continue this style unless a change is necessary. Similarly, if you employ bulleted lists, consistently use the same bullet symbol throughout the document. For instance, if you use hyphens for all lists but suddenly switch to stars, it might lead one to question the reason – does it signify something important, or was it a mere oversight in formatting? Furthermore, once you choose a method to differentiate between more and less important information, stick to it. Avoid unnecessary complications or refinements.

Be precise

Version 1. Take a look at the Medium font.

Version 2. Take a look at the article font used for posts on Medium.com.

Now, imagine you have never heard of Medium as a blogging platform. Which of the sentences may raise questions or mislead?

Write simply

Now, things become tricky. From my experience, unless you are well trained in this area or have simple communication "in your blood", there is much work ahead of you. In my culture, I was taught how to write elaborate, long sentences, avoid repetitive words, etc. Now, retraining myself to direct, plain communication is difficult. The challenge here is to talk about complicated stuff as simply as possible, yet not oversimplify. Let me share some of my ideas on how to address that.

First, let's look at the three examples of overly complicated sentences. I generated them with ChatGPT.

1. Leveraging big data and analytics has enabled corporations to maximize the utilization of actionable insights derived from the intricate assessment of voluminous datasets to drive strategic decision-making processes, thereby augmenting operational efficiencies and fostering a competitive advantage in their respective markets.
2. By integrating advanced predictive analytics into the core operational strategies, businesses can scrutinize vast amounts of unstructured data to unearth patterns that preemptively signal shifts in consumer preferences and market dynamics, thus significantly mitigating potential risks associated with market volatility.
3. Utilizing multifaceted machine learning algorithms to dissect and interpret complex data structures enhances the accuracy of forecasted outcomes. It substantially optimizes resource allocation and maximizes ROI across various business segments.

Okay. Now, let's try to simplify them.

1. Using big data helps companies make better decisions, improve operations, and stay ahead of competitors by understanding large amounts of information.
2. By using advanced analytics, businesses can analyze large, complex data to spot trends and predict changes in customer behavior and market conditions, reducing risks.
3. Machine learning lets businesses better predict outcomes and improve how they use resources and invest money, benefiting different parts of the company.

How do we achieve such an effect?

The fewer words, the better. Version 1: If you can say or write something with fewer words than you originally intended, simply do it.

Version 2: Select the shortest possible way of saying or writing something.

Version 3: Say or write with the least number of words.

Version 4. Use fewer words.

Tags: Communication Data Storytelling Deep Dives Presentations Visualization

Comment