As a compensation professional, going to conferences about compensation like the World at Work Total Rewards conference is exciting for me. And as a nerd, I like to try to find ways to analyze the available information for an exciting trip. When I take my family to Disney World, I buy the 800+ page guidebook, check websites, read reviews, etc. But for a professional conference, there aren’t as many visitors and there is not going to be as much to data prepare for the conference as there is for Disney World.
As a nerd, I want to make the most of the information that is available to me. What is available to me, in this case, is a detailed schedule on the conference website of session titles, description, speakers, and times.
The downside to conferences is that there are more breakout sessions than I can possibly attend. This year I pulled data (excluding all the general and vendor sessions) for the 91 breakout sessions from people that are likely as nerdy as I am about my profession. These 91 choices are spread across only 9 unique breakout session time slots, so there is a lot of overlap. I am honored to be speaking in two of those 9 slots, so I’m going to get 7 unique slots to check out what amazing things so many of my peers and mentors are doing.
The common problem is “Who is going to talk about what? How do I sort through the different options?” Some speakers are going to talk more about the topics that I want (or need) to hear about than others. I could get out a spreadsheet, and I could go through the list of hundreds of presenters, and co-presenters for the 91 sessions. I could read each person’s bio and look up their company, and try to figure out if they have spoken at the conference before. Or I could do what most people do, and go through each time slot, and just pick a first and second choice of session to attend and hope that I can get a seat.
But then I decided to dust of some topic extraction analytics that I’ve done before, and I used this data analytics workflow which is basically an implementation of a fairly unstructured NLP (Natural Language Processing) system. And I put the last 4 years of Total Rewards conference session titles, descriptions, presenters into it to see what the top 10 most common themes of words used were:
I then took the data that came out of that, loaded it into Tableau, and created an interactive visualization story. You can walk through a brief explanation and use the interactive story charts, so you too can see who talks about what the most at the conferences. You can look at the most common presenters on a given topic. As you get into the dashboards, you can hover over the charts, see names, titles, and topic trends for either the person, or the company they work for. (If it doesn’t display below, you can click here instead)
Is this overkill for a 3 day event that is very different each year? Probably. Is it nerdy, definitely. But if we want HR to get nerdier, and be more data driven, then we have to be willing to expirement and try to learn new skills. I’m sharing this because it’s something that around 1,000 people could find useful for a few days. But more importantly, I am hoping a few people read this and say “I’d like to talk to someone who is that nerdy about compensation and data, so I will seek him out at this conference.” Each year, I look forward to reconnecting with the old friends and meeting new ones. See you in Dallas!