A case against the spreadsheet mindset

Recently, I was having a conversation with a seasoned visualization (dashboard) specialist. I asked this person, “What is the biggest challenge in getting users to adopt a visualization technology?” I was expecting to hear that there was a lack of skills around being able to build complex queries; or experiencing complicated steps in configuring the dashboard user interface, graphs and charts.  The answer truly surprised me.

He answered that the biggest barrier was the spreadsheet mindset. Even with advanced technology available to them, most users struggle with trusting and analyzing data outside of tables full of data points and looking at the raw numbers instead.

He went on to explain that he provided the same dashboarding tool to multiple users-  engineers, sales, and finance employees conducting multiple training sessions and providing examples for them to reference. The engineers produced highly graphical "cool charts" but these analytics were complex and hard to understand. The finance employees, number experts, replicated their excel spreadsheets and did not add any visualization- stating that was easier to see the numbers in tables.  Finally sales provide a mix of both traditional tables and simple charts.  

I was surprised how users are much more trusting and comfortable with tables of numbers rather than visualizations like charts and counters.  Numbers are great when looking into a specific value, but they inhibit analysis and blur the “big picture”. The best example that illustrates how charts enable more insights quickly, I learned from the most respected person in the industry, Stephen Few.

Imagine a table with 12 columns representing the months and 20 rows each with the sales by product line.  Now try to answer the following questions in less than 5 seconds:

1. Which product has the highest and lowest sales?

2. Is there any seasonality per product?

3. What is the overall sales trend?

4. Which month yielded the highest sales?

5. Are there any anomalies over the time period being analyzed?

Now imagine a line chart graph, based on the same dataset. Now go back and try to answer the same questions.

Were you able to analyze the information faster in visual format (chart) or table format? Were you able to answer them in less than 5 seconds?  I’m confident that the insights gained from the chart were both easier to obtain and more efficient in providing the answer.  What do you think?