Event optimization case study - SurveyEngine GmbH

Maximising Attendance and profitability of an Outdoor Event

CASE STUDY
First published August 2005

An events organiser’s hunch is confirmed and Bondi Beach becomes a Winter Wonderland.

Over half a million people attend the Winter Festival in Sydney, Melbourne and Brisbane. It receives international coverage and dominates the niche of winter events in Australia. The festival is an unqualified success.

The Problem

The organisers knew their product well and had all the necessary skills to manage the event, but because it was a new concept there were still plenty of questions such as:

  • Should admission be charged?
  • What events should be included?
  • What should we charge for food and drinks?
  • Where and when should it be held?

Now you might be thinking, “that’s easy. Just run a survey or just ask people.”

Why Traditional Market Research Fails

Asking people directly, either verbally, or through a questionnaire works well if there is a single dimension – such as “would you vote for Mr Smith or Mr Jones?”.

But when there are multiple dimensions, traditional market research has to resort to clumsy methods such as ratings or rankings that often offer no more insight. This method would lead to people wanting all of the benefits and none of the costs.

A Combinatorial Approach

Think of all of the possible Winter Festivals that could have happened but didn’t. The ones where the rink was too small, the music was bad and the food too expensive (or too cheap). In fact, there are over 4 Million possible Festivals that could be constructed.

Testing every one of these ‘hypothetical’ festivals is clearly infeasible because of the large number of combinations. However using Choice Modelling, the core technology in SurveyEngine, it is possible to achieve the same result by modelling the data in controlled trials. The models identify systematic behaviour, if it exists, in how people are making decisions.

What Mattered

The most important factor was the size of the rink followed by the location. The model predicted that using this information alone an increase of 800% in attendance would be achieved.

The model also allowed fine-tuning of other attributes such as the price of food and drink and activities.

The Results

The apparently risky decision to stage a winter festival on an Australian beach paid off. The event was a huge success and spawned a number of other festivals in Brisbane and Melbourne. 

Scroll to Top