The secrets of story telling
There is nothing worse than a presenter who throws up a PowerPoint and begins to read off the screen. Despite how good the presenters data and analysis may be, the presentation comes off as dull and boring and leaves the audience with little recollection of the information. There’s an art to telling a good story, and we all know a good story when we hear one. But there’s also a science behind the art of storytelling with data.
The more a speaker conveys information in story form, the closer the listener’s experience and understanding will be to what the speaker actually intended. Storytelling is a universal human trait, spanning across cultures all throughout history. Anthropologists have discovered folktales in ancient cultures worldwide, written in languages such as Sanskrit, Latin, Greek, Chinese, Egyptian, and Sumerian.
In 2008, Jeremy Hsu wrote an article for Scientific American discussing the secrets of storytelling. In his article, he states that the story structure, known as narrative, engages its audience through psychological realism. By portraying recognizable emotions and believable interactions among characters, authors can captivate their audience creating an immersive and memorable experience.
Furthermore, story telling is an important tool for learning and for developing relationships in social groups. Stories promote social cohesion among groups, and serve as valuable methods for passing on knowledge to future generations.
What is financial data storytelling?
Data storytelling couples data visualization with compelling narratives that help audiences better comprehend and take action based on data analysis. While effective data visualization helps people grasp and remember key takeaways, data storytelling is essential for helping them understand why those takeaways matter.
Lydia Hooper, Data Visualization Expert and Information Design Writer
Lydia Hooper highlights that data storytelling is a comprehensive method of sharing information. A key feature of this approach is data visualization, which helps craft an engaging narrative. To effectively impact the audience and convey your message, it’s important to integrate data, narrative, and visuals.
Click here to read more about effective data visualization techniques!
Plain and simply, data storytelling is the process of translating data into clear terms and visualizations to help convey your message and influence business decisions. In the financial world, data storytelling can help paint a clear picture of the financial or economic situation, leading to improved outcomes and a better understanding by the audience.
In finance, data storytelling is particularly valuable as it transforms complex financial metrics into engaging narratives that highlight their significance within a broader business context. By integrating data, narrative, and visuals financial professionals can effectively communicate trends, risks, and opportunities to stakeholders. This approach not only aids in decision making, but also fosters a deeper understanding of financial health and performance.
For instance, a financial analyst can illustrate the impact of market trends on a company’s revenue, helping executives make informed strategic decisions. Similarly, for a risk analyst, data storytelling is an essential tool to better explain the narrative around potential risks and their implications.
How to tell stories with data?
Now we know that data storytelling can help turn data insights into action, and without effective communication our insights can be forgotten by the audience. As Harvard Business School Professor Jan Hammond points out, data storytelling requires both hard and soft skills to leverage data to its fullest potential and captivate your audience.
Click here to read more about effective data visualization techniques!
Let’s break it down
Three key components to data storytelling
- Data: A deep understanding of the dataset through descriptive, diagnostic, predictive, and prescriptive analysis enables you to grasp the full picture.
- Narrative: The essence of the story, also called a storyline, is used to communicate the insights gained from the data analysis. It helps to form the context around the data and actions you recommend to captivate your audience.
- Visualization: Visualizing your data helps engage your audience in the narrative and communicate the data more clearly and informatively. This can be in the form of charts, graphs, infographs, diagrams or media.
Data storytelling can be utilized internally to communicate financial insights, such as highlighting the need for budget reallocation based on performance metrics. Externally, it can be employed to present a compelling financial case to investors or stakeholders, showcasing the potential return on investment and financial health of the company.
A risk analyst might use a combination of bar charts, infographics, and narrative explanations to show how a particular risk has been mitigated over time, the financial impact of these mitigations, and the remaining exposure. This comprehensive approach ensures that the audience not only understands the data but also the story behind it, leading to more informed and effective decision-making.
The key steps in storytelling with financial data
Financial storytelling works by combining data with narrative and visual elements to create a cohesive story. The process often involves using dashboards created with Plotly or Shiny, as well as custom infographics, reports, and dashboards from Data Sense to enhance collaboration and storytelling. These tools help in visualizing the data, while the narrative explains the “why” behind the numbers, making the information more relatable and easier to understand.
Click here to get a quote on custom infographics, reports and dashboards!
The key steps of storytelling with financial data include collecting the financial data, performing analysis to provide context to the data, developing a compelling story that ties the data to the context, adding visuals to illustrate key ideas and make the data engaging, and presenting the story in a clear and concise manner.
Collecting Data
Wrangling Private and Public Data
- Private Data:
- Internal Financial Records: This includes data from accounting systems, financial statements, and internal reports. These records provide detailed insights into financial health and performance.
- Customer Data: Information from CRM systems, sales data, and customer feedback can offer valuable insights into revenue streams and customer behavior.
- Operational Data: Data from various departments such as HR, production, and logistics can help in understanding cost structures and operational efficiencies.
- Public Data:
- Market Data: This includes stock prices, market indices, and economic indicators. Sources like financial news websites, stock exchanges, and government publications are valuable.
- Industry Reports: Reports from industry analysts and market research firms provide context and benchmarks for comparing the company’s performance.
- Regulatory Filings: Publicly traded companies are required to file financial statements with regulatory bodies like the SEC, which can be a rich source of data.
Scraping Data from the Web
Web scraping involves extracting data from websites using automated tools. This can be particularly useful for gathering large volumes of data quickly. Here are some common applications:
- Financial News:
- Market Trends: Scraping news articles about market trends helps analysts identify emerging patterns and shifts in the financial landscape.
- Company Performance: Gathering news on company earnings reports, mergers, and acquisitions provides insights into individual company performance.
- Economic Indicators: Collecting data on economic indicators such as inflation rates, employment statistics, and GDP growth helps in understanding the broader economic environment.
- Social Media:
- Public Sentiment: Analyzing social media posts to gauge public opinion on financial products, services, or market events.
- Brand Reputation: Monitoring mentions and discussions about a company to assess its reputation and identify potential PR issues.
- Competitor Activities: Tracking competitors’ social media activity to stay informed about their marketing strategies and customer engagement.
- Competitor Analysis:
- Product Pricing: Collecting data on competitors’ pricing strategies to adjust your own pricing and remain competitive.
- Customer Reviews: Analyzing customer feedback on competitors’ products to identify strengths and weaknesses.
- Financial Disclosures: Scraping financial reports and disclosures from competitors to benchmark performance and identify strategic opportunities.
Data Cleaning and Preparation
Once the data is collected, it needs to be cleaned and prepared for analysis, this involves:
- Removing Duplicates:
- Identify Duplicate Entries: Use algorithms or software tools to detect and flag duplicate records in the dataset.
- Merge or Delete: Decide whether to merge duplicate entries (if they contain complementary information) or delete them to maintain data integrity.
- Automate Detection: Implement automated processes to regularly check for and remove duplicates, ensuring ongoing data quality.
- Handling Missing Values:
- Imputation: Replace missing values with estimated ones based on other available data, such as using the mean, median, or mode.
- Deletion: Remove records with missing values if they are not critical to the analysis, ensuring that the remaining data is robust.
- Flagging: Mark missing values to keep track of them and understand their potential impact on the analysis.
- Standardizing Formats:
- Consistent Date Formats: Ensure all date entries follow a uniform format (e.g., YYYY-MM-DD) to avoid confusion and errors in analysis.
- Uniform Units: Convert all measurements to a standard unit (e.g., all financial figures in USD) to facilitate comparison and aggregation.
- Text Standardization: Ensure consistency in text data, such as using standardized abbreviations and capitalization.
- Validating Data:
- Accuracy Checks: Cross-verify data entries with original sources or other reliable datasets to ensure accuracy.
- Range Validation: Ensure numerical data falls within expected ranges (e.g., financial figures should not be negative unless specified).
- Consistency Checks: Verify that related data points are consistent with each other (e.g., total sales should match the sum of individual sales entries).
Data Analysis
Descriptive Analysis
Assess the data and provide a solid foundation by summarizing the information with statistics and visualizations.
- Summarizing Data:
- Central Tendency: Calculate mean, median, and mode to understand the average performance.
- Dispersion: Use standard deviation and variance to measure the spread of data points.
- Frequency Distribution: Create histograms to visualize the distribution of data across different categories.
- Identifying Trends:
- Time-Series Analysis: Plot data over time to identify upward or downward trends.
- Seasonal Patterns: Detect recurring patterns or cycles in the data, such as quarterly sales peaks.
- Moving Averages: Smooth out short-term fluctuations to highlight long-term trends.
- Highlighting Patterns:
- Heat Maps: Use color-coded matrices to show the intensity of data points across different dimensions.
- Scatter Plots: Visualize relationships between two variables to identify correlations.
- Cluster Analysis: Group similar data points together to identify patterns and segments.
Diagnostic Analysis
Adds depth to the descriptive analysis by explaining the reasons behind observed trends.
- Root Cause Analysis:
- Drill-Down Analysis: Break down aggregate data into finer details to pinpoint specific causes.
- Fishbone Diagrams: Use cause-and-effect diagrams to systematically explore potential reasons for an issue.
- 5 Whys Technique: Ask “why” multiple times to peel back layers and uncover the root cause.
- So What?: After identifying the underlying root cause, it is important to understand what that means, the so-what.
- Correlation Analysis:
- Pearson Correlation: Measure the strength and direction of the linear relationship between two variables.
- Spearman Rank Correlation: Assess the relationship between ranked variables.
- Cross-Tabulation: Create contingency tables to explore the relationship between categorical variables.
- Variance Analysis:
- Budget vs. Actual: Compare actual performance against budgeted figures to identify variances.
- Trend Analysis: Examine how variances change over time to spot recurring issues.
- Benchmarking: Compare performance against industry standards or best practices to identify gaps.
Predictive Analysis
Offers foresight into future possibilities, helping businesses anticipate changes and plan ahead.
- Forecasting:
- Time-Series Forecasting: Use models like ARIMA to predict future values based on past data.
- Regression Analysis: Identify relationships between variables to predict future outcomes.
- Exponential Smoothing: Apply smoothing techniques to forecast short-term trends.
- Risk Assessment:
- Probability Models: Estimate the likelihood of different risk events occurring.
- Monte Carlo Simulation: Use random sampling to model the probability of various outcomes.
- Scenario Analysis: Evaluate the impact of different risk scenarios on business performance.
- Customer Behavior Prediction:
- Churn Prediction: Use machine learning models to predict which customers are likely to leave.
- Lifetime Value Prediction: Estimate the future value of customers based on their past behavior.
- Purchase Likelihood: Predict the probability of customers making a purchase based on their interactions.
- Forecasting:
- Time-Series Forecasting: Use models like ARIMA to predict future values based on past data.
- Regression Analysis: Identify relationships between variables to predict future outcomes.
- Exponential Smoothing: Apply smoothing techniques to forecast short-term trends.
- Risk Assessment:
- Probability Models: Estimate the likelihood of different risk events occurring.
- Monte Carlo Simulation: Use random sampling to model the probability of various outcomes.
- Scenario Analysis: Evaluate the impact of different risk scenarios on business performance.
- Customer Behavior Prediction:
- Churn Prediction: Use machine learning models to predict which customers are likely to leave.
- Lifetime Value Prediction: Estimate the future value of customers based on their past behavior.
- Purchase Likelihood: Predict the probability of customers making a purchase based on their interactions.
Prescriptive Analysis
Delivers actionable recommendations, guiding businesses toward optimal decisions.
- Optimization Models:
- Linear Programming: Optimize resource allocation to achieve the best possible outcome.
- Integer Programming: Solve optimization problems where some variables must be integers.
- Network Optimization: Optimize logistics and supply chain networks to minimize costs and maximize efficiency.
- Scenario Analysis:
- What-If Analysis: Explore the impact of different assumptions on outcomes.
- Sensitivity Analysis: Assess how changes in input variables affect the results.
- Decision Trees: Map out possible actions and their potential impacts to guide decision-making.
- Decision Trees:
- Actionable Insights: Provide clear recommendations based on the analysis.
- Cost-Benefit Analysis: Evaluate the financial implications of different decisions.
- Implementation Plans: Develop step-by-step plans to execute recommended actions.
Putting it all together
Effective storytelling with data combines both the technical aspects of data collection and analysis with the art of communication. It begins with gathering data through techniques like data wrangling, data scraping, and data cleaning to ensure the information is relevant, accurate, and ready for analysis. Data wrangling organizes raw, unstructured data into usable formats, while scraping allows extraction of specific data from web sources, and cleaning removes inaccuracies and inconsistencies. These foundational steps ensure the integrity of the data being used to build a compelling narrative.
Once the data is prepared, different types of analysis help uncover meaningful insights. Descriptive analysis provides an overview of what happened, summarizing key metrics or trends. Diagnostic analysis dives deeper, identifying causes and correlations behind the observed patterns. Predictive analysis then leverages historical data to forecast future outcomes, while prescriptive analysis offers recommendations for decision-making based on the insights gathered. When woven into a story, these analyses guide the audience through the data journey — clarifying the past, explaining the present, and offering foresight into the future — creating a clear, impactful message that resonates with decision-makers.
Ultimately, data storytelling is about transforming raw information into actionable insights that drive growth and informed decision-making. Essentially, it’s about learning from data, and that’s why it’s crucial to embed data-driven storytelling into organizational cultures and learning programs. Data storytelling is an essential tool for engaging, influencing, and educating stakeholders. By weaving data collection and analysis into a narrative, you not only convey facts but also inspire action. If you’re looking to drive meaningful impact, integrating data into your stories ensures a compelling and evidence-based message, backed by the power of data science.