In the world of PowerPoint, Prezi, Canva, Tableau, Looker, PowerBI and millions of other tools that should make organizing and presenting data easy, poor data presentation skills are incredibly common. And yet people generally don’t know how to present data. How many times have you heard the awful “Sorry, this chart is a little hard to see on the slide” or a poorly labeled monstrosity of a graph with fifty different colours, like the one below?
There are many poor graph and colour choices that can be made, but my most despised graphs are poorly labeled stacked graphs. Those are incredibly energy draining to decipher, so I tend to ignore them if they are being presented.
So I hope that the following 5 tips will benefit anybody who will ever have to present a slide sometime in the future. From analyst level to senior executives. Don’t think this is a problem at junior levels. During my MBA studies, I have seen experienced managers and directors present unlabeled wobbly lines in class.
Following tips are not necessarily about graph choices and colour combinations because that is a matter of taste. Although there are some things you should never do. However, it is more about the purpose of every graph, every colour choice and how they are used to convey your story.
Data presentations shouldn’t be about the data
How to present data better? Don’t make it about the data!
The data in a form of charts, graphs, tables is not important at all. It is the story that it tells, that is the key of every presentation. If one line can summarize the story better than 3 convoluted charts, go with one line chart.
This is a mistake I made a lot during my first years working as an analyst and I constantly see people make. I thought when I went into meetings and presented as much data as possible, I would be rewarded. Look at me, look how hard I have worked on this and how much content I produced. I stuffed my slides with fancy graphs, trying to cover every point. I would create a number of “Appendix” charts just in case our CEO had some twisted question and bam I would pull up a chart that would bedazzle the executive team.
But in reality the opposite would often happen. Instead of amazement I would see puzzled looks and the worst sounding words a junior analyst can hear “We see all the work you’ve done but let’s summarize major points here..” Not impressed.
Data is meaningless, every chart, every graph needs to have a very clearly defined purpose.
Use PowerPoint notes to summarize stories
Use PowerPoint notes under every slide and write out the following: “The data here shows:__”. For example, the chart here shows the decline in sales over the last quarter that requires review of the following stores” or “The graph here summarizes the success of the marketing initiative implanted last month”.
When you pose this question to yourself, you realize that maybe the chart you have is not the best suited to tell “the story”. So expect multiple revisions as a result.
Never use Absolute Numbers
When you present an absolute number like Sales last month: 1000 units, unless I am directly involved in sales or it is a small company and everybody knows exactly how many sales the company makes every single month, this information is utterly useless to me. Especially, if I am from a different department.
I’ve seen this happen many times. Finance department comes in and says we posted a profit of $1M. I have no idea how to react, is it good and we should celebrate or is abysmal and we should completely reevaluate our whole strategy?
Stay away from absolute numbers and always focus on TRENDS. Record quarter results? Show me a bar graph with the previous quarter’s trend.
Now obviously there are exceptions here. If an absolute number is so astonishing that everybody in the audience understands it without context, definitely use it. For example, the first 1 million users or the first billion dollar year etc.
It is not about the story either
Okay so now we are all set to tell better stories with our data. We know the purpose behind every graph and we can clearly tell the story. We focus on trends and not individual results. But we can never forget that we are not in the storytelling business, we are in the actual business!
People are not paid to tell beautiful data stories but to either MAKE or AID in decision making. Every story must lead to action.
Therefore, wherever possible we must think one step ahead and connect Data to Story to Action. Imagine a Senior Director sitting through a presentation where a team of analysts, smart as they are, do all this amazing analysis, present the story and then just leave it at that. My boss used to say “We are paid to think and act, not to compile information”.
Add at least one possible action item in the slide notes
In the notes, under every slide after you write: “The data here shows:__”. Add another line called “NEXT Steps”. Jot down what you would recommend the next step will be. In the case of falling sales, it can be simply reviewing the worst performing stores and retraining the sales team and so on.
If you are a junior person reading this and you say my company doesn’t appreciate recommendations from junior associates, reevaluate your relationship with your employer.
Labels, Titles and Highlights save mental energy
Audiences have limited mental energy. The later in the day your presentation is, the lower the energy level in the room will be. Don’t waste whatever shred of energy your audience has on comprehending your unlabeled charts.
Label things. Never forget the units if the space allows. Especially if you are dealing with thousands, millions etc. The worst thing that can happen is if you yourself forget the units during a high pressure presentation. Sometimes it is not clear if your number makes sense in thousands or millions and you get asked about it. You will look foolish if you make it up and end up being wrong or if you have to say the dreadful “We will get back to you on that..” for such a simple question.
Label to the point where one quick glance at the graph and everything makes sense. Follow the less is more approach:
After spending a few seconds looking at a graph your audience needs to see the major point. Use colors where possible. Play with tones to separate highlighted points.
When generating graphs in Python use Annotations from Matplotlib.
Make One point per chart
To further expand on the point of highlighting key points, don’t overwhelm. One chart means one major point. More points if they both support the same message but don’t cram opposing points on the same chart in hopes to impress your audience. In fact having too many points on a slide, means that you failed at the job of generating insight because you couldn’t evaluate which point is more important.
Know your audience
If possible figure out in advance who is going to be at the meeting. Is it just the sales team then you might want to exclude the financial graphs. Will the CFO attend, well then you might want to keep the financial slide.
These are key questions to ask yourself before prepping for your presentation:
Who is in your audience?
Knowing who will receive your information is crucial. This feels like common sense but you have to tailor your presentation to your audience.
What are their major goals?
As I mentioned before, the goal is to make better business decisions. But for different departments that means different things.
What do they know?
How well versed is your audience in the data you are presenting? Do you have to simplify or the opposite, do you need to go in depth?
What keeps them up at night?
Is there a major concern that the audience will want to focus on? How does your presentation relate or address that concern.
What is their energy level?
Will you be the last to go in a stream of long presentations from others? If that is the case, maybe simplifying even further might be a good idea. Or spending extra time making sure that all charts are so easy to comprehend a child can understand them at a first glance.
Key takeaways that I had to learn the hard way myself is that the quicker you can get to an actionable insights the better it is for everybody. To get there some data presentation techniques can do a better job than others, but the focus should never be on the data itself or the graph, it is always about what the graph is saying!