Thursday, August 21, 2008


I have always been hesitant to write on this subject because I felt like it was one of our company secrets that gave us a competitive advantage. Data bases, of course, are no secret. Neither is what you can do with them. Most hotel property management systems (PMSs) have a guest history feature. The key is that they are PMSs, not sophisticated relational data bases!

What's the Difference? The difference is that in a relational data base one can ask for a list of people that stayed more than once, arrived on a Sunday night, paid rack rate, for three or more nights in January, said they would return, live within a two hour drive, and were not smokers. Yes, we know that for our hotels and we can send a personalized letter with unique references for each of them within a few hours of deciding to do it.

We collect the necessary data in several ways. In hotels without PMSs, registration cards and folio data are input into a simple data base file with the necessary fields by the night auditors. These data are subsequently appended to our main history data base in our corporate offices on a monthly basis. In hotels with a PMS, the necessary data are either imported to our office local area network (LAN) by modem or disk and manipulated until ready to be appended to our history data base.

Historical Database

I use the term manipulated because guest history data on a PMS does not typically include the necessary fields we need for our subsequent data queries. For instance, a PMS will probably have a field for arrival day and number of nights but not for the days of the week that encompasses. While a PMS knows which rooms are non-smoking for the purpose of doing room assignments, the fact a particular room is non-smoking is not saved to the guest history record. Additional editing work includes verifying American and Canadian postal codes, getting all the names associated with the proper gender for personalized addressing and to get capitalization, spelling and spacing just right.

For many of these things the software we use, Smartware II by Angoss, will do the work for us with either standard formulas or special queries. For instance, "TRIM" will clean up all the spacing issues while, "PROPER" will take care of almost all capitalization issues. Queries which effectively say, "If the arrival was on a date which was a Sunday and the number of nights was two then code the record as a Sunday and Monday stay." Or, the query might say, if the guest stayed at such and such a hotel during a certain time period in one of the following rooms then the record should be coded as a smoker, etc. Yes, we know that is not foolproof but it is a start and our experience is that smoking rooms are more costly to clean and maintain. We also are careful to remove duplications but not multiple stays. Multiple stays are coded as necessary. At some point, we want to begin noting how long its been since the previous stay so we can use the information to our benefit.

While many PMSs have guest history modules that collect much of the same data we are and can merge it into letters, we have found that the combination of information and the unique lists we want to extract cannot always be done. They also need purging as their storage capacities are filled over time. We believe that the best PMS focuses on what the hotel needs in reservations, registration, guest accounting, etc. As long as we can get the data off the PMS we can make up for the PMS's data base and word processing shortfalls relative to what we can do in the office.

Comments Database

The information in our history data base is also screened against our guest comments data base. This data base is a collection of all the information collected on comment cards in our hotels. At each hotel the comment cards which do not have complete information regarding the guest are researched and the information added to the card whenever possible. The General Manager reviews all comment cards at the weekly staff meeting before forwarding them to the corporate office where they are reviewed and then entered in a data base.

Each week for every comment card which we have a complete address for (over 95%), a personalized letter is sent to the guest automatically on that hotel's stationery. The content of the letter varies based on the guest's experience and comments. Each letter appears to be personally signed by that hotel's General Manager. Monthly summary printouts are prepared and distributed to the various hotels for additional review by the leadership team. We also use the ideas and comments from the guests to plan capital budgets and service upgrades.

We then run the history and comment data bases simultaneously and matching key information to identify the guest as being the same person. We don't match entire names as the room may have been registered in the husband's name, Joe Smith, and the comment card completed by Mary Smith. By matching room numbers, dates of stay and the last four letters of the last name we get a pretty accurate match so that the history record can be coded appropriately from the comment card record.

This extra work allows us to do some pretty amazing extra things in our direct mail campaigns. In the case of the Econo Lodge in Minneapolis which benefits from some hospital-related business, our system, which involves some other coding on the registration information, keeps us from confusing those stays with commercial or tourist guests. We just would not want to confuse a hospital stay in late November from Fargo with a Holiday Season shopping trip!

We can use our analysis of guest comments from certain types of rooms or time periods and correlate them to the corresponding types of guests in general. These two data bases, along with some good letter or post card copy writing are key components of our hotel guest history marketing efforts. They are also crucial in deciding where to advertise and make sales calls. The front office staffs also uses extracts of the information along with reports from the central reservation systems of the hotel chains we are franchised with to more effectively manage our room inventory and rate programs.

The history data base also carries nonguest data records. These include Chamber of Commerce membership lists from local communities, travel agency information (combined from what is paid by the chains centrally and what we pay direct), snowmobile clubs, or whatever we might accumulate.

One feature of a relational data base is that two or more data bases can be linked together in useful ways. We have created several views of our history data base that include tables that are actually the continuous call report for a particular record. In this way we can look at a travel agent's record and see both the extent of our marketing efforts (direct mail and sales calls) along with the total revenue they have generated for us and the details of their bookings with us regardless of which hotel it might be.


Our third data base is our prospect and business contact data base. This data base is used in the marketing of the management company. It, too, includes views which have tables including the history of our contacts with a particular person or company. This data base includes trace dates, quality ratings, source information and attributes.

Quality ratings help us keep track of hot, warm and cold prospects, etc. Attributes are alphabetical and numerical codes assigned to each record. A record might have more than one attribute. For instance consultant, "C" and owner "W" in the case of a California consultant who owns a New Orleans hotel. In this case we need to market for both referrals and possible management business. There are insurance companies that might lend, take ownership positions and/or use our consulting services.

On at least a monthly basis our Director of Business Development, Bill Hauge, runs a special routine on our prospect data base that includes sorts and multiple queries so that this large data base is broken down into indexes by geography, attribute, trace date, account responsibility within our company and other characteristics. This information is used to help manage our marketing efforts.

As I look back over many years in our industry, I sometimes wonder how we managed before we had computers. Twenty-five years ago computers were still in their genesis in our industry. They were typically only found in central reservation systems, high end telephone switches and the fancy new electronic cash registers. Now these items are relatively inexpensive, have features to enhance service and marketing opportunities that we only dreamed about and in many cases are beyond what we dreamed about.

Using these many technological marvels effectively, both hardware and software, can take creativity and be time-consuming. Those that don't learn them and use them will be left behind in the battle for market share whether they are a small player in a local competitive skirmish or a large chain trying to maximize the benefit of a frequent stay program.


No comments: