Essential steps to present data persuasively
It's not enough to have conducted research or created a winning program at work - you need to share results in a thoughtful manner
“The graphic method has considerable superiority for the exposition of statistical facts over the tabular. A heavy bank of figures is grievously wearisome to the eye, and the popular mind is as incapable of drawing any useful lessons from it as of extracting sunbeams from cucumbers.”
-Arthur Briggs Farquhar and Henry Farquhar from Economic and Industrial Delusions
In a previous column on industry trade ClickZ, I shared that data visualization is your secret weapon for storytelling and persuasion. I wanted to follow that up with actionable tips for how you can be more successful with your own efforts sharing data with your team.
I’ve been in countless meetings over my career where clients or colleagues presented data to me …painfully. When I say painfully, it wasn’t because they didn’t work hard on a campaign or project and have amazing insights to share. But the actual presentation/delivery of insights and actions could have been far more effective. And very worth the effort.
Presenting data persuasively is a ‘last mile effort’, because you’ve already done the work, mentally feel ‘finished’, and are now just sharing learnings and next steps. This is understandable and basic human nature. The work feels finished, which is why many content creators throw slides together haphazardly as they are ready to move on to the next thing. Today I’d like to persuade you not to do this and instead to take time to present your data thoughtfully.
If you follow the following steps, your next meeting with a boss or client will not only go smoother, you will have much higher odds of influencing that person to take the action you want. I’ve included a few brief examples to illustrate my points.
Before you get into the data, set up the situation
Never just start presenting to a group by throwing charts and graphs on a screen. And never present to a group just by sharing a Google Sheet or projecting Google Analytics reports on the overhead.
Make a quick deck – not only is this better to present it, but gives the team a takeaway that’s easier to circulate internally and senior execs who get forwarded the content are far more likely to open and click through.
So you’ve started a presentation. First, spend time to take people through what you did (visually, ideally). For example did you just run an A/B test on a new landing page? Great, show a slide with the old page and new page before you dive into results so we can see what you did, along with a goals of why you ran this specific test in the first place.
The benefit here is now everyone will begin your presentation immediately following your logic and you’ll get far fewer questions. Your team may have initially signed off on the project but it’s more likely they don’t remember everything. Help them out.
As an example, if my colleague Krista Seiden was going to present the results from her project redesigning the Google apps for Business site, before diving into the data she should start by showing what the old site looked like.
Have one clear takeaway per slide, fix broken, confusing or misleading visualizations
Too many times I’ve seen young account executives throw all their data on one slide (perhaps not even visualized just as a list of numbers) because they feel a need to get everything across up front and more data must be good, because it looks more official, right? Wrong.
Never have a slide with more than one chart – it’s just too much material, no one is going to absorb it and people’s eyes will gloss over. This is simply too much information for wrap our mind around and still listen to a presenter.
When you try to say everything, you say nothing. An example of what not to do (four charts on one slide, just far too much):
Even more important, never mislead your audience! This visualization scale in the below chart is not only totally broken, but the “cute” attempt to use people icons to represent the bars in the chart simply adds more confusion. It’s difficult to take anyone seriously who uses these sort of graphics, however well-intentioned.
Some graphs are just hopeless …the below not only has 3D bars which you should never, ever use as they are incredibly confusing, but the icons above the chart add even more confusion. Not to mention the fact that this chart has no label! The only thing this chart would persuade someone of is that the person who created it probably shouldn’t be presenting data to your team. And by the way, this was a real chart someone presented in an agency meeting (edited to protect the innocent).
Via Ian Lurie
Present data as simply & cleanly as possible (now let’s see some good examples)
If you wanted to compare the growth of different revenue streams or consumer preference in a sector, you don’t need to do anything fancy. Just plot the data in a simple line graph with labels, sources and titles. Easy and the results are very clear!
We know from looking at the below chart that between 2017 and 2018, online video service revenue is projected to eclipse box office. There’s no room for confusion or any way to misinterpret.
It happens to us all – you created a chart that’s not very clear. But, this is fixable: figure out what point you want to get across and remix the data to portray, clearly, what you want to communicate.
The huge takeaway of the below chart is the massive 216% growth of internet as a source of news over the last 10 years for consumers. But you can’t see that very well in the first chart. Bold the one datapoint you want to use to make your point and also consider calling out percentage change to focus your audience on how big the trend is.
Source: flowingdata via one of their ‘Visualize This’ challenges to make data visualizations better
Always present data with context – never assume people know what you do, even your own team
So, for example don’t just show a trend of metrics via this month, but overlay what last year looked like to quickly see what this means compared to previous timeframes.
Google Analytics makes this super easy to do and it’s far more helpful than a chart that looks pretty and is going up and to the right. Sure, that looks nice, but what does it mean? Are these numbers good? We don’t know. Context answers this.
Additionally, never simply show a chart like this by itself – instead, have a text box below with something like a percentage change call out you want to draw attention to and the reason for it. Tell us what you want us to take away from the data.
And remember, someone should be able to click through your slides and get a clear understanding of what you wanted to communicate without needing you to present.
As analysts interpreting not just the ‘what happened?’ but ‘why?’ and ‘what does it mean?’ are what separates the good from the great. A hint for making your life easier: use annotations in Google Analytics to add callouts to interesting events right in the product, and later when you need to share data you won’t have to worry about forgetting what happened during that timeframe.
Have additional details in an appendix, but you don’t need to go through it live
You don’t need to go through every single KPI and indicator metric in your presentation, especially the ones not core to your project. It’s actually what you don’t show that makes presentations better. Not only does more data not help you tell your story but it’s going to tire everyone in the room out and you’ll lose attention.
There’s only so much we can absorb in one sit down, plus it’s our job as analysts and marketers to share only the key information. Offer an appendix to both CYA but also provide detail if someone does want to understand more. They can do this offline and not waste the group’s time.
Have a ‘next steps’ slide where you summarize exactly what you want accomplished next
You’ve presented your project, goals, results and insights. You’ve made your points and have everyone persuaded to think the way you want them to (seeing reality, hooray!).
Now summarize with what, specifically you are going to do with this data in the form of a task list and team deliverables (such as running a new test, getting people started on that badly needed new site design, or removing products no one is buying from your ecommerce catalog).
Review your slides/presentation before you present
Practice makes perfect with presenting, and while you likely don’t need to go to the length of someone preparing to keynote an event, you should do at least do a quick dry run of what you want to say.
It’s worth the effort, even one or two rehearsals through something will not just get you comfortable with the material but also make it clear what you can remove from your presentation.
Bonus: I gave a talk in New York on Data Driven storytelling in 2017, that goes through the above and more. If you’re interested in learning more on this topic you can view here or via the embed below.
A version of this post originally appeared on my analytics column on ClickZ.com