Whether you’re a data scientist or data analyst, at one point in your career you’ll have to present your results to an audience. Knowing what to say and include in your presentation will impact your success. After giving many data presentations over the years, I’d like to share my tips on how to increase your chances for a successful presentation.
Know your audience
The audience for your presentation will dictate the level of detail and information you’ll present. There are three main types of audiences:
- Peers – These are data analysts, data scientists, and anyone in analytics that understand what you’re explaining if you drill down to methodology, analytic approaches, or code. Detailed information is preferable for this audience to share your work and for them to understand your approach and possibly leverage for their projects.
- Stakeholders -These are people in the department you support that asks you questions. A moderate level of detail is needed in your presentation for stakeholders to understand the results and make decisions.
- Senior leadership – These are people that manage your stakeholders and occupy senior positions in the company. Only high level information is required that will impact the Business. There is no need to go into details that you would with peers or stakeholders.
Content
I’ve found this format works best for stakeholders and senior leadership. For peers, you can add slides on your approach and methodology because they are able to appreciate this level of detail.
- Agenda – I find it’s helpful to have an agenda at the beginning of the presentation to set expectations for the audience.
- Key findings – Summarize key findings at the beginning of your presentation. I put this slide after the agenda. These are the main points you want your audience to take away from this presentation. Relate your findings to impacted KPIs to show your audience why your findings are important.
- Supporting data – These slides will contain data, graphs, and any other information to support the key findings you just reviewed. Refer to my guidelines for audience types to determine what level of detail you should include.
- Reiterate key findings – Stakeholders and senior leadership are busy dealing with many things and reinforcing the key findings will help them remember the main points you made at the beginning of the presentation.
- Recommendations or next steps – Based on your findings there may be actionable next steps. This may not apply for your presentation but try to associate your results with KPIs because that will resonate better with stakeholders and senior leadership. In the website visitor drop example, the next step is to compare the website changes before and after the drop date to identify the root cause. Failure to address this issue will lead to a decrease in visitors translating to less customers making a purchase and declining revenue. Any impact to revenue will make your audience pay attention because it’s a very important KPI.
Presentation layout
- Slides – Each slide title should contain a main finding and the details in the slide should support the finding. Someone reading only the titles should be able to get a summary of your presentation. For example, if the slide is going to talk about a drop in visitors on the website due to organic search don’t put "Website visitor trend" but instead put "Website visitor decline caused by drop from organic search". In the slide details, show visitor traffic by referral source and indicate the drop was from organic search. Note that upon further investigation a change on the website corresponded with the date when visitors started to drop from organic search.
- Data visualizations – Create graphs to show trends. People can process visualizations more easily than a table of numbers. The first graph below shows daily website visitors before and after the website changes. It’s very easy to spot when the decline started. The second graph is better than the first because I added an arrow to draw attention to the visitor drop. This draws your eye immediately to the part of the graph you want the audience to focus on.


Presentation length
This is my formula to determine how many slides to include in my main presentation assuming I spend about five minutes per slide.
(Presentation length in minutes-10 minutes for questions ) / 5 minutes per slide
For an hour presentation that comes out to ( 60–10 ) / 5 = 10 slides.
Five minutes per slide may sound long but many times I’ve had stakeholders and senior leadership start a discussion about data on a slide and before I knew it ten minutes had gone by. Create an appendix section to add slides that won’t fit within the presentation time. You can refer the audience to the appendix if there are questions not addressed in the main presentation. The best way to figure out if you’ll run out of time is to have a practice run and time yourself. Another reason I advise starting with key findings is because this ensures you’ll get your main points across in case you run out of time.
Presenting in front of an audience can be intimidating and not knowing the right way to present your data results makes it even harder. Now that you know my tips, you’re on your way to a successful data presentation.
You might also like…
How to Deliver a Successful Data Presentation
How I Used a Machine Learning Model to Generate Actionable Insights