Ramping Up: Business Intelligence in Hospitality Today

By Joe Sullivan, CIO, UniFocus

How do you measure success? This is an age-old question, and high-tech has been tripping over itself for years in trying to provide answers. Ever since computers have been placed in front of managers, we have been bombarded with stats, variances, and standards. When I started in hospitality IT in the 80's, my executive team was screaming for more information. In the 90's they were screaming for less information. They wanted information that was more concise, easier to absorb, and more exception-based. Today, I talk to a lot of executives, and many of them are just screaming. So, what is so hard about getting the right information to the right person at the right time?

In today's technology world, we use a term called "Business Intelligence" (BI). Like most popular business terms, BI has taken on additional meanings over time. For the purposes of this article, I am limiting the definition of the term to the reporting technologies that focus on Key Performance Indicators (KPIs), with the intent of improving decision-making.

I recommend breaking up a business intelligence project into 4 phases:

• Determine business processes and outcomes to be measured
• Transform sources of raw data into databases that will be used by BI system
• Determine which behaviors are required from a system to interact and absorb the data (graphing, modeling, color exception-coding, drill-down to more detailed views, dashboard gauging)
• Select vendor/system that will deliver a BI platform and the integration required to ramp up the system

Determine Key Measurements

This first phase is best accomplished before introducing any technology. We need to get our measurements, formulas and variances down on paper first. This enables everyone to focus completely on your business processes and KPIs. For example, if your focus is on Labor Management your team can focus on overtime, accurate labor standards that will deliver your desired level of service, and your scheduling business rules. Once you have your KPI's, do a gutcheck. If you were able to put these measurements in 2-foot letters on the wall that everyone sees when they arrive at work, is this really what you want them to focus on? If it isn't, then go back to your design sessions. The following questions present a good rule of thumb for effective measurements:

• Can I establish a baseline for current performance?
• Can I track this measurement accurately on a regular basis?
• Are there decisions that my staff can make that will impact this measurement?
• Is this metric measuring something that is important to my business?

Transform Raw Data into Analysis-Ready Data

Most data that is tracked by our enormous computer systems is transactional. In other words, it is sequentially added to the database without regard to what was previously added, and its purpose is to capture an event in time. Business Intelligence tools like OLAP cubes, trending tools, and high-level rollup reports do not operate efficiently from transaction-driven databases. Sophisticated BI tools typically "transform" your datagathering databases into new databases that are more structured for reporting purposes. Totals are precompiled, links for trending are saved with the data points, and relationships are established. In general, business applications and financial statements are easier to design, because they are very strict with their hierarchy. For example: all market segments add up to room revenue; all rooms department employees add up to Rooms Payroll; Rooms Revenue minus Rooms Expenses equals Rooms Profit, etc. It's all very straightforward. In BI systems, you may want to create relationships that don't fit into the standard hierarchy. For example, you may wish to compare your laundry chemical expense across multiple hotels and create a ratio to occupied rooms to see the comparative CPOR (cost-per-occupiedroom) for this expense. You may also want to see if your laundry staffing is correlated to your chemical cost. If you wanted to know if some of the hotels in your portfolio are running too many half-loads through the laundry process, a BI tool can answer that also.

System Behaviors

Once you have the metrics and you understand how you will gather the data to feed your BI system, now you can determine how you want to interact with the data. To start, you want to be very clear on who your user-base will be. If this is a group of managers that want quick answers to well-established metrics, then you want a very simple to use exception-based system. This group won't tolerate a steep learning curve. They want to login, get their results, see their improvement, and identify areas of weakness.

If your user-base is more analytical, then you will need a system that has easy export capabilities. All of the screens should have drilldown capabilities that enable the user to start at a high level, then link to more detailed data whenever and wherever they wish. They will also want to define their columns and rows. OLAP cubes are perfect for this group as they get to design their own report layouts in either summary or in more granular detail.

The last group of power-users is professional analysts. They have probably already identified their preferred modeling tools. Statistical packages like SPSS and SAS are ideal for unlimited modeling. However, these types of users are essentially programmers that are designing their tool as they are interacting with the data.

Being Relevant

The final phase is ensuring that before you embark on a costly initiative to track metrics, you need to determine if the metrics will make a difference to your business. For example, tracking the number of people that use your fitness center is only important if there is a relationship between fitness center usage and guest loyalty.

The role of any Business Intelligence tool is to improve your operations through the decision-making that is possible through the effective presentation of metrics. Measurements need to be timely and need to be accessible to the right people who can act on those results.

Conclusion

To summarize, the function of any technology solution is to support the business that pays the bills. For that reason, the success of your Business Intelligence initiative requires an equal amount of attention from both the technology and business process perspectives. Don't get roped in by all the pretty graphics. Rather, you should focus on what is truly actionable for the manager. These simple guidelines will provide a painless user experience and higher return on investment for your BI implementation. The management team will then have measurably improved decisionmaking capabilities.

Finally, here again is that age-old question as it applies to BI: How do you measure success? Quite simply, an effective BI solution should help you run your hotel more efficiently and more profitably than your competitors. And for that you will need to select the right platform and a smart solution provider who understands what your Key Performance Indicators are.

That's the intelligent thing to do.