• By Company

 

Customers in the Spotlight

"At Disneyland Resort Paris, each visitor should have a magical day. How can we deliver the dream for every single visitor? How do we make 40,000 dreams come true every day? [...] Operational BI using Business Objects solutions enabled us to achieve our dream with an astonishing effect on visitor satisfaction rates, the performance of our teams and the company as a whole."

Christian Perdrier
senior vice president parks, Disney Village and Security

Challenge

With 12.5 million visitors a year, Disneyland Resort Paris is one of Europe's top tourist attractions. How do they manage to make a day in Disneyland truly unique for each visitor? "In principle, it's simple; we have to deliver 40,000 magical days, but we need to know how to relate to the visitors who come en masse in an individual way," says Christian Perdrier. Over half of the visitors to Disneyland Resort Paris come on the spur of the moment without pre-booking. From experience, the Disneyland Resort Paris teams know that the number of visitors will fluctuate depending on the week, the season, the weather, or school holidays and that different nationalities have different tastes and preferences. The forecasts of daily activity are based on data models for typical days, although it is impossible to be totally accurate in forecasting. That's why we need to be able to react very quickly, where the action is.

What the company needed was a system to manage services in real time. This was achieved by implementing operational business intelligence (BI) using Business Objects solutions.

The Business Vision

In a service industry, and especially in the leisure and tourism sector, a satisfied customer is a customer who doesn't have to wait. "We know that a customer's perception of waiting time varies depending on where he or she is. In the car park, visitors have a very low tolerance of waiting time as they feel they are wasting time that could be spent enjoying themselves on the attractions. In the restaurants, visitors are prepared to wait slightly longer in their expectation of a good meal. At the attractions, waiting is generally more acceptable to visitors as they are looking forward to the guaranteed enjoyment of the attraction," explains Christian Perdrier.

During the day, the Disneyland teams have about twenty opportunities to come into contact with each visitor; at the ticket booths, at the attractions, in the shops and restaurants etc. "I wanted to be sure that my teams were on the ground, at the exact place where the visitors needed them," says Christian Perdrier. "I'm not a computer expert, so I explained my vision to our IT team." The computer team's ability to listen and understand what was needed and an excellent level of partnership with the business managers responsible for shops, restaurants and attractions did the rest. Disneyland Resort Paris's operational BI application came into being as a pragmatic, business focused rapid implementation project; its objective being to serve a "zero latency" enterprise. In other words, the time lag between the moment when it becomes apparent that a queue is too long and the time when the company takes action to address the problem, should be as close to zero as possible. Bruno Brocheton

Disneyland Resort Paris took their first steps into BI around 1995. At that time, BI was used to provide consolidated reports from source data in various systems; then in 1998 BI was used to optimize hotel capacity. Disneyland Resort Paris then went on to develop specific data marts, capitalizing on these infrastructures to extend data analysis to other business lines such as shops, attractions and restaurants. Fernando Iafrate, senior manager BI confirms: "We were very experienced in decisional information systems. We were able to produce excellent analyses of past business activity and this allowed us to make sophisticated predictive analyses to model typical days. But there is a variance of -20% to +20% between a typical day in the model and what actually happens in a day. From the moment when the visitors arrive at the park, in the restaurants or shops, we are blind. All we could rely on when problems arose, was the experience of our operational managers to resolve these problems on the ground as and when they occurred."

Disneyland Resort Paris decided to evolve from decisional BI - which analyzed past results - to a "live" vision of the ground with real data captured minute by minute. Commencing in 2004 with the shops and restaurants, the project then continued in 2005 with optimization for the attractions.

Joint command via the Operational Control Center

How did these concepts become reality? Disneyland Resort Paris has set up an Operational Control Center (PC OPS), a real control tower, staffed by professionals who liaise directly with the teams on the ground. Their mission: to supply the right information, to the right people at the right time.

At the start of the day, the experts in the PC OPS center have screens which show a dashboard, modeling the various stages of the day that is just beginning: visitors entering the park, the flows to the attractions, turnover in the restaurants and shops. At start of day, the display is based on forecasts. Then, as the day progresses, actual figures are displayed on the screen, updated every quarter of an hour. If the actuals reveal a higher number of visitors than forecast, PC OPS can alert the operational managers via mobile devices, so that the teams on the ground can be reinforced at key points. If it starts to rain and visitors rush to take shelter, park staff can be redeployed to those areas where the visitors are. Depending on whether the number of entrance tickets sold in the morning or activity in a restaurant or shop is above or below the forecast, PC OPS can ask for a cash desk to be closed or opened and cashiers redeployed as needed. "We have taken our best operational team members to staff the PC OPS center, because we are convinced that the managers on the ground must have utmost confidence in the PC OPS professionals. They act as a team, with some of them navigating while others hold the steering wheel. And it works" comments Christian Perdrier.

BI in Action

Operational BI captures real time data from transactions entered in over a thousand points of sale and information about waiting times at attractions. "We did not have a large budget and we wanted to remain very pragmatic. We discussed very specific problems with the operational teams; for example, how to get real time information directly from the transactions at the counter to the cash desks in the shops and restaurants. We already had a full decisional infrastructure and all the information which enabled us to model activities and visitor flows on an hour by hour basis. Our challenge was to deliver this data to operational managers throughout the day on a minute by minute basis," comments Fernando Iafrate.

For Bruno Brocheton, this is truly "BI in action". "All our indicators are available to PC OPS and also via the operational portal and on Blackberry handheld devices carried by our operational staff. We produce BusinessObjects report data in HTML, adapted for use on the Blackberry browser. The screen shows the indicators and associated graphs: the forecast model graph appears in blue, along with the actual graph in green which constantly updates minute by minute."

"We had to make a break with a certain dogma," confesses Fernando Iafrate. We had to accept that two types of data could co-exist within the company: audited data which has undergone the formal control process on the one hand; and on the other hand unaudited data which reaches the operational process much more rapidly and where its timeliness is of more value than the quality of the data. The unaudited data is not used for its pure numeric value; it is aggregated to arrive at thresholds which become converted to indicators. The indicators are colored to signify the corresponding decision level. Green, for example: it's all going well; Orange: a risk area requiring an increased level of vigilance; Red: a decision and immediate action are needed. The data is constantly refreshed at less than 15 minute intervals. "BI is magic. You mustn't see what's behind it. We wanted to hide the technology from the users, so that all they see is the benefits. We achieved this with a highly standardized approach and integration with the business processes."

Bruno Brocheton continues: "From the systems point of view, we have established a standardized approach based on simple concepts." Between 1992 and 1998, IT management at Disneyland Resort Paris established standards: Unix operating systems, complemented by Microsoft NT from 1998; Oracle and Informix databases; SPSS for data mining and BusinessObjects for business intelligence. Data from ten or more different source applications comes together in data marts which cover the main business activities: restaurants, shops, attractions, visitor entrance tickets, call centers, hotels, human resources, finance, e-commerce. These all have the common dimensions of time, location and product. To facilitate BI project deployments, Disneyland Resort Paris also set up a skills center with six BI experts, a library of best practices, indicators and templates. "This was how we made the most effective use of the BusinessObjects expertise available within the company," recalls Bruno Brocheton.

"With the BusinessObjects technology we were able to show that the early results we had achieved in optimizing occupancy of our 6,000 hotel rooms could be reused for other activities. Business Objects had not only simplified our BI environment, but it had laid the groundwork for the future. Even though the transaction systems were different, the decisional layer could remain intact."

At Disneyland Resort Paris, BI is part of a democratic and business-focused company culture. Its mission is to distribute intelligent information which facilitates the decisionmaking process at all levels. In principle, decision-making for management is global and strategy-focused; managers' decision-making is performance-focused; operational team members decisions are immediate and action-focused. With its "zero latency" organization, Disneyland Resort Paris has aligned all teams at all levels within the company to the same KPIs (Key Performance Indicators). Previously, BI was only available to 700 staff at most; it is now available to 8,000 users.

Disneyland Resort Paris is very proud of the success of its BI project which has become a model for the other Disney resorts. The Disney parks in Florida. California and Hong Kong are all very interested in adapting it for their needs. At Disneyland Resort Paris, decisional BI and operational BI have joined to form a virtuous circle. Decisional BI provides the functionality for modeling typical days; while operational BI allows the typical day to be adjusted every quarter of an hour. The adjustments improve the decisional base which consequently delivers more finely tuned and relevant "typical day" models. This allows the Disneyland Resort Paris team to provide the visitor with more and more information so that the visitor in turn becomes a real decisional player.

• Read more Business Objects customer success stories
• Read more Hospitality case studies

Quick facts

• Industry:
  Hospitality
• Business pain:
  Legacy BI systems did not provide the relevant information to Disneyland Paris employees.
• Why Business Objects?
  Improved income and margins, a 15-point increase in customer satisfaction rating, and 8% increase in occupancy rates for attractions.
• Business Objects products and services
  Standardization on Business Objects solutions, creation of a "zero latency" operational BI system.