Taylor Rodgers
  • Home
  • Books
  • Free Tools
    • Requirement Doc Templates
    • Quality Checking Guide
  • Blogs
    • Data Science
    • R Programming
  • Consulting
    • Data Science
  • Contact
  • Home
  • Books
  • Free Tools
    • Requirement Doc Templates
    • Quality Checking Guide
  • Blogs
    • Data Science
    • R Programming
  • Consulting
    • Data Science
  • Contact

Most Analysis Presentations Are Boring - Here's What You Can Do to Fix Them

6/8/2019

0 Comments

 
Analysis presentations are often treated like dashboards read out loud. The presenter will go through a series of slides, reading off facts such as “Impressions were up in July” or “Conversions remained steady.” This will bore the audience, who will inevitable check out and start looking at their phones, begging for a distraction.

Analysis presentations don’t have to be this way. These presentations are a unique opportunity to do something we often say we want to do with data – tell a story.
“Telling a story with data” is the big cliche in business intelligence and data science.  I’m guilty of using that line in job interviews. Now I cringe when I hear other people say it.

Truth is, we often don’t get any opportunity to tell a story with data. Most stakeholders want dashboards for reporting so they can find and tell the story themselves. The closest anyone in the data visualization development world gets to telling a story are people who work for news outlets, such as FiveThirtyEight and the New York Times. The data visualizations on those outlets are intended to tell a one-time story that’s usually out of date after a few days.

Dashboards, on the other hand, are meant to automate reports. Since you’re wanting something that holds up over time, you don’t focus on one-off questions. You focus on several KPIs and must design a dashboard, not to tell a story, but to make it easy to read so much information on one screen.

Data analysis presentations though are where data professionals can and should tell a story. You have a captive audience at that moment. They are expecting a more detailed analysis that goes beyond simple KPIs.

Sadly, BI professionals often don’t make use of that opportunity. It’s not because they lack public speaking skills, although that is often the case. It’s because they don’t know how to structure the content for a presentation. No one has ever shown them.

There’s many ways you could structure this content and it depends on the moment, but there’s a simple ‘hack’ I’ve learned that applies most of the time. I use the inverted pyramid of journalism.

Most newspapers, from the Wall Street Journal to your local paper, use this pyramid to structure the content of their stories. You start with the who, what, when, where, and why at the top. Then you explain the background. You leave other interesting tidbits of information at the bottom.
Picture

The reason journalists do this is that they know most people will lose interest as the story progresses. This allows the reader to still know the key events, even if they have no desire to keep reading.

You have the same challenge with analysis presentations. Usually people are paying the most attention at the start of the presentation. As it continues, they lose interest. By focusing on the key details up front, you can ensure they take action with your analysis at the start.

For example, many people would outline a presentation like below, where there's a series of slides covering each topic.
Picture

This presentation outline goes through various subjects, such as website and campaign performance. All that information may be useful and even hold actionable insights. However, the content flow doesn't emphasize those insights. It just tells straight facts. By the time you got to social media, you've probably lost your audience's interest.

Using the inverted pyramid method, I would take that same information and organize it like the following, where I focus on providing the key takeaways first.
Picture

Typically, I'd do this by providing three questions I asked during my analysis, such as "why did website traffic decrease, but revenue on the website increase?" or "why did owned social produce so much engagement, but so little click-through?" I would then immediately answer those questions without even using data or charts.

The reason I do this is that my audience is the most captive in the first ten minutes. As the presentation moves forward, their attention will decrease. By telling them the key takeaways up front, before I even show them the data, they will be more likely to remember those takeaways.

Then I'd start showing them charts, graphs, tables, and any other data point that supports my analysis. If there are other performance metrics that didn't seem relevant to my key points, I'd create an appendix section for the audience to use on their own later.

My team and I recently revamped our presentation format to follow the inverted pyramid method. I’ve sat in on one of these presentations and it was amazing how much more effective the engagement was among the stakeholders. They started talking among themselves during the presentation about their business strategy. The discussion actually led to a decision to invest more in paid media.

And that’s what we want with our stakeholders. We want them engaged and to actually make a decision with the data we provide.

Other Thoughts

In addition to analysis presentations, BI professionals can use a similar method for dashboard content. Most experienced developers do this intuitively, sometimes without even realizing it. They’ll present the KPIs at the top of the dashboard or on the left. More granular metrics are presented at the bottom of the dashboard or on the right.

0 Comments



Leave a Reply.

    ABOUT

    A blog about the non-technical side of data science.

      SUBSCRIBE

    Confirm

    ARCHIVES

    December 2022
    April 2022
    March 2022
    October 2021
    September 2021
    August 2021
    March 2021
    February 2021
    January 2021
    November 2020
    August 2020
    June 2020
    May 2020
    April 2020
    February 2020
    January 2020
    December 2019
    November 2019
    October 2019
    September 2019
    August 2019
    July 2019
    June 2019
    May 2019
    March 2019

    RSS Feed