You will need Power BI Desktop for the workshop exercises. Download free at powerbi.microsoft.com/get-started
By the end of this lesson you should be able to:
Communicating data effectively
Non-verbal communication
Narrative intelligence & storytelling
Power BI workshop — Olympics data
Before you communicate, ask yourself: am I still exploring, or am I explaining a finding I've already made?
| Exploratory Data | Explanatory Data |
|---|---|
| All activities performed to understand the data | Selected output curated for your audience |
| Broad: you don't yet know what's interesting | Focused: you've already found what's interesting |
| Many charts, hypotheses, dead-ends | One clear narrative with supporting evidence |
| Internal — for the analyst | External — for stakeholders |
We have generally been focused on explanatory data — you have already identified interesting observations, tested hypotheses or applied a model, and now need to communicate findings to a target audience through data visualisation.
Public speaking serves four purposes in data communication: to Persuade, Inform, Educate, and Entertain. Great presenters are not born — they are trained through deliberate practice.
"If you fail to plan, you're planning to fail." — Benjamin Franklin
Six-step preparation framework:
Research shows audiences trust their eyes over their ears — if a presenter's body language is inconsistent with their words, the audience believes the body. Six key dimensions to manage:
Face your audience. Turning away signals you are not interested in them.
Stand straight but relaxed. Open palms signal honesty. Limit repetitive fidgeting.
Lean slightly toward audience members you are addressing to signal engagement.
Six universal emotions: happiness, surprise, sadness, disgust, anger, fear. Smile genuinely.
Small groups: make contact with each person. Large groups: section the room and sweep each zone.
Pace, pitch, volume, pauses. A well-timed pause is more powerful than filler words.
Common mistake: Reading directly from slide text and turning your back to the audience. Know your material well enough to maintain eye contact.
Data can circulate as stories — from media, competitors, or consumers — that may be false or misleading. Businesses must:
Narrative Intelligence is the ability to understand the stories of others and strategically use stories for various purposes — to inspire, persuade, and educate audiences.
Why narrative intelligence matters in business:
Opening → Build-up → Problem → Solution → End
Most persuasive data presentations follow this arc. The visualisation supports each stage.
For the Olympics dataset you will analyse today, think: which year is the climax? What is the resolution?
Explain — the story behind the numbers. Narrative based on data articulates why something happened.
Enlighten — charts arising from data reveal patterns that would be invisible in a table.
Engage — combining narrated story with compelling visuals creates the most impactful data communication.
All three together: a complete data story that drives change.
The unique characteristics of your data that make your story credible and worth telling.
The central idea — one sentence that captures what the audience should walk away understanding.
The supporting details that prove or contextualise your main point. Not everything — only what's relevant.
Each slide or section builds on the last. The audience is led through a logical progression.
Tension, contrast, surprise. The element that makes the audience lean forward. Good data has inherent drama — reveal it.
Images, charts, and photos that make the data tangible and emotionally resonant.
As you build your Power BI dashboard, identify which chart serves as the visual anchor, which statistic is your main point, and where the story has drama.
| Pitfall | Why It Fails | Fix |
|---|---|---|
| Charts too complex / overloaded | Audience cannot read or interpret quickly; cognitive overload kills the story | One chart, one message. Split complex visuals into separate slides. |
| Pie chart percentages ≠ 100% | Destroys credibility immediately. An infamous Fox News chart summed to 193%. | Double-check arithmetic. Prefer bar charts for most comparisons. |
| No variety in chart types | A dashboard of identical bar charts is visually monotonous and fails to signal different data relationships | Match chart type to question: trend → line; comparison → bar; composition → stacked/donut |
| Truncated or misleading axes | Starting a Y-axis at a value other than zero inflates differences visually | Start at zero for bar charts; annotate clearly if a line chart uses a non-zero baseline |
| Inconsistent colour coding | Same colour used for different things across visuals confuses comparison | Establish and stick to a consistent colour palette across all charts in a report |
Great data stories share these qualities:
XR (Extended Reality = VR + AR + MR) combined with storytelling transports the user into the data. Benefits for analytics:
XR remains an emerging field — the storytelling principles you learn today apply regardless of delivery medium.
File: womens_participation_in_the_olympic_winter_games.csv — 22 rows · 7 columns · 1924 – 2014
| Column | Type | Description | Range |
|---|---|---|---|
year | Integer | Year of the Winter Games | 1924 – 2014 |
sports | Integer | Number of sports featuring women's events | 1 – 7 |
womens_events | Integer | Number of events in which women competed | 2 – 49 |
total_events | Integer | Total events in that year's Games | 16 – 98 |
%_of_womens_events | Decimal | Women's events as % of total events | 12.5% – 50.0% |
women_participants | Integer | Number of female athletes | 11 – 1,120 |
%_of_women_participants | Decimal | Women participants as % of total | 4.3% – 40.7% |
The data spans 90 years of history. Before analysing it, ask: what story do I expect to see? Formulating a hypothesis before visualising prevents confirmation bias in how you read the charts.
Home → Get data → Text/CSVwomens_participation_in_the_olympic_winter_games.csv and click OpenComma and the data shows 22 rows. Click Load.womens_participati…. You should see all 7 columns listed.year in the Fields pane. In the Column tools ribbon, change Summarization from Sum to Don't summarize. This prevents Power BI from aggregating years.Olympics for clarity.year to the X-axis well.%_of_womens_events to the Y-axis well. Set aggregation to Average.#C8102E).Notice the sharp plateau between 1960 and 1988 — despite early progress, women's events stagnated near 40% for nearly three decades before accelerating again.
Does offering women's events lead women's participation — or lag behind it? This chart answers that question.
%_of_women_participants and drop it into the Secondary Y-axis well (shown separately below the primary Y-axis well).55 to match the primary axis.%_of_women_participants line colour to #1e3a5f (navy) in Format → Lines.The gap narrows post-1988 but never closes. Women's events % consistently outpaces women's participation %. This gap reveals barriers beyond simply offering events.
year to the Category field well.womens_events to the Y-axis well. Set aggregation to Sum.#1a7340), Decrease to red (#C8102E), Total to navy (#1e3a5f).The largest single-Games jump was 2010 → 2014: +8 events (from 41 to 49). The second largest was 1984 → 1988: +7 (19 → 26). The waterfall makes this immediately visible.
year to the X-axis well.total_events to the Y-axis well. Then drag womens_events also to the Y-axis well — this creates a clustered pair per year.total_events to grey (#cccccc) and womens_events to red (#C8102E).In 2014, womens_events = 49 and total_events = 98. For the first time, the red bar is exactly half the height of the grey bar. This is the story's climax in Freytag's terms.
Card visuals are the most direct way to surface a headline number. Use them to anchor your story's main point at the top of the dashboard.
women_participants to the Fields well. Change aggregation to Maximum. This shows 1,120 (peak in 2014).%_of_womens_events, aggregation → Maximum. Shows 50.0. In Format → Callout value, add suffix %.sports, aggregation → Maximum. Shows 7 sports with women's events by 2014.A slicer allows the audience to explore different time periods. This is especially powerful for storytelling — you can guide the audience to focus on the "plateau era" versus the "acceleration era".
year to the Field well.When presenting, use the slicer to split your story into two eras: 1924–1988 (slow growth) and 1992–2014 (rapid acceleration). Ask: what changed between these periods?
A static chart tells one story. An interactive slicer lets the audience test their own hypotheses — this is the difference between explanatory (you present) and exploratory (audience discovers).
In a live presentation, use the slicer to create drama — filter to show the plateau, then zoom out to reveal the breakthrough. The surprise is built into the interaction.
Filter the slicer to 1960–1988. The card showing peak women's events % should drop significantly. What does this tell you about the story during this era? Are there political events (e.g., boycotts, Cold War) that might explain the patterns?
Insert → Text box. Type: "Women in the Winter Olympics: 90 Years of Progress". Font: Segoe UI, size 18, bold.Format painter in the Home ribbon, then click another chart to copy all formatting. Use this to create consistency across visuals.View → Themes, apply a theme or in Format pane (with canvas selected), add a subtle background colour (#fafafa).Now that your dashboard is built, apply the storytelling frameworks from Section 4 to evaluate and narrate it.
| Framework Element | What to identify in your dashboard | Example answer |
|---|---|---|
| Data Foundation | What is unique or surprising about this dataset? | Historical records from 1924–2014 showing a 90-year arc; rare longitudinal Olympic data |
| Main Point | One sentence: what does the data show? | "It took 90 years for women to reach equal representation in Winter Olympic events." |
| Explanatory Focus | Which 2–3 statistics best support your main point? | 1924: 12.5% → 2014: 50.0%; participant growth 11 → 1,120; plateau 1960–1988 |
| Drama | Where is the tension, surprise, or turning point? | The 28-year plateau (1960–1988) — what caused it? What broke the stagnation? |
| Visual Anchor | Which chart most powerfully communicates the story? | The dual-line chart: the narrowing gap between events % and participants % is visible and compelling |
| Freytag Climax | Which year is the peak turning point? | 2014 — the first year of gender parity in events; the goal of the entire 90-year arc |
Share your completed dashboard with another group. Evaluate their visual choices against this framework. Does their main point come through without explanation? Is the drama visible in the charts?
year field when using it on the X-axis of a line chart?