Phil Buchanan / Making Analytics Useful for Event Planners

Making Analytics Useful for Event Planners

Article tagged as: Design Process, Research, Ideation, Data Visualization

A really bad pie chart with dozens of equally sized slices and overlapping text is all but useless
An example of one of our old charts. It doesn't even have a legend or labels. Not our best work.

We were loosing customers because of our analytics platform. And rightfully so, when we were outputting useless charts like this one.

It would have been easy to make a few quick changes for a quick win, but we wanted to focus on delivering new value to our customers, not just covering up our mistakes of the past. We needed a better understanding of the jobs our customers were looking to hire analytics to do.

What Do We Offer Today?

Before talking with customers it was important for us to know what they were currently using. No one on the design team even worked at EventMobi when the original analytics was built, so we had to find out for ourselves where the pain points were.

Our system was basic, offering a few metrics organized around our system's architecture and features. It didn't offer much in the way of insights or actionable data. The metrics we did provide were mostly vanity and required our success team to schedule debrief calls to help customers actually understand the data. Some data wouldn't even load.

Armed with this understanding, we began our research.

Research

Our first step was to combed through data from our product analytics systems and databases to learn about usage patterns. We wanted to know things like how often analytics were being looked at and how often reports were being generated.

We also conducted some competitor analysis, looking at how other systems handle analytics gathering and reporting.

The outcomes prepared us for the customer interviews we conducted. We wanted to hear from a wide range of customers so interview candidates were selected based on a number of criteria including those who love our system, those who have had feature requests in the past, and even some customers we recently lost.

Our interviews revealed three primary insights:

  1. The most critical insight was was that event planners measure the success of their event based on event goals. Every event has a different goal like education or networking. By bundling various features, our customers were tailoring our product towards helping achieve one or more of these goals. Customers didn't care how one feature performed compared to another. We realized that we could group our analytics around the same goals planners had rather than our individual features.
  2. Our customers were manually comparing events year-over-year, exporting raw Excel data from each event and combining it together to tell the story they wanted. This is a tedious, time consuming process—something computers and technology are already well suited for.
  3. Often times the reports our customers exported weren't for themselves. Instead, they were shared with CFOs, Marketing Directors, Speakers, Sponsors etc. Each stakeholder is looking for very different information.

Documenting Learnings and Preparing for Ideation

We knew event goals were the key to success for our analytics. Together, the product team listed every goal we could think of, grouping and organizing them until we had a list of the most common categories. We then mapped each of our product's features against those categories. For example, our chat, group discussions, activity feed, and gamification features could be combined into attendee networking analytics to help planners understand how attendees were connecting with each other.

Organizing event goals into categories in a mind map

Ideation

We decided to run a group ideation session involving team members from product, sales, customer success, support, marketing & engineering, in order to solicit the widest range of ideas.

The session began by sharing and discussing our research findings and allowing participants to contribute additional insights such as issues a support member had heard, or questions customers were asking during a sales call. All ideas were posted to a whiteboard to ensure everyone had a reference and shared understanding of the problem.

Searching for the broadest set of ideas possible, we used the crazy eights sketching exercise. We followed up with silent critique to avoid having strong personalities in the room control the conversation. After silent critique, we regrouped to critique and provide additional feedback and clarity. Finally each participant was given one post-it and asked to vote on their favourite idea. The final sketches served as the basis for our prototyping and testing phase.

Heat map and voting on sketches produced during the ideation session
Heat map and voting on a few of the sketches produced during the ideation session.

Rapid Prototyping and Testing

At EventMobi we have an incredibly powerful tool for creating rapid, high fidelity prototypes (it's an HTML/CSS/JS prototype, powered by our product style guide). We used our prototyping tool to rapidly build and test the winners from our ideation session.

We utilized our internal event planning team for initial user testing. After a few rounds of testing and iteration it was time to return our conversation to real customers. The reception was better than expected and we generated a lot of excitement for the future release of the updated platform.

Some winners from our testing included:

Grouping Analytics: Event planners loved the idea of grouping analytics based on their goals. And why wouldn't they? We we're finally putting their needs first.

Comparing Data: Comparing data year-over-year was already something planners were doing and was a tedious process with our old system. In our redesign, we made that a simple and delightful experience.

Analytics comparing events year-over-year

Reports: Remember the dumpster fire charts of our old system? We improved them considerably with a set of logical guidelines defining how much information we display and how we display it in charts and legends. Now charts are consistent, predictable and usable.

A redesigned report chart with clear legend and capped number of items

Excited About Success

The whole project took about 3 weeks to complete, including scheduling and conducting interviews, running the ideation session, prototyping, and testing. The events industry runs on an extended timeline where buying decisions are made many months in advance of usage, so we'll be watching this closely over the next few quarters so see how it performs. In the conversations sales has had, customers seem very excited about the process of reporting on the success of their events.