Skip to content

Business Intelligence Begins with Data; Data Management Begins with BI

As business intelligence begins to permeate the hospitality sector, one thing is certain: You can take a “buy” approach, implementing a BI module from a hotel systems provider (for instance, as part of your PMS, CRS or Sales & Catering system) or from a third-party solutions provider.  You can take a “build” approach, securing the funding and setting out with your IT team to implement something from scratch.  In either case, the certainty is that hotel business intelligence begins with data.  Data is the lifeblood of the hotel reporting, analytics and dashboards that are the output of any such BI initiative.  And while high-quality data (complete, accurate and uniform) won’t alone ensure the success of your hotel BI initiative, low-quality data (full of holes, errors and inconsistencies) will certainly ensure the failure of your initiative if not addressed.

I emphasize “if not addressed” because too often I’ve seen hotels or hotel companies allow their concerns about the poor quality of their data to prevent them from embarking upon a business intelligence initiative at all.  This makes sense on the surface – why invest in BI tools when I know that we have data issues, and therefore the reports and dashboards in my BI tool might be inaccurate?  Here’s why: BI exposes the flaws in your data like nothing else can.  BI pulls data inconsistencies – from the highest to the most granular of levels – “into the light of day” so that they can be acknowledged and addressed.  So while it seems that getting your data in order is a precursory step to implementing business intelligence, the truth is that data management and business intelligence exist in a symbiotic relationship – each strengthening the other in a continuous process of refinement.  The mistake to be avoided is to wait until you have “perfect” data to begin.  Clearly, one needs to be careful with early interpretations of hotel reporting and dashboards in a BI environment suspected to have data quality issues – but this can be managed through selective exposure of users to BI output and through prudent communication.  And that elusive state of “perfect data” may never arrive in any case – though you can get very close with the right approach.

You’re likely to find all types of data-related issues when you embark upon a hotel business intelligence initiative.  That’s because data is the outcome of our business processes (or lack thereof) and the consistency of those processes (or lack thereof) within or across business units – in our case, hotels.  And data is affected by the integration of our various hotel systems as well (an easy example being how PMS data is affected by the configuration of the PMS interface with the CRS – as reservations flow in, values for room types, source codes and the like are translated, with the accuracy and completeness of that translation being a function of how the interface is configured).  This all means that the correction of data issues – “data management” if we want to call it that – is not a “quick fix”!  Rather, it’s a long process that will most likely involve asking human beings to change how they are doing certain things (our business processes – for example, which fields are filled in – and how – when taking a reservation or checking a guest into the hotel ) and may require attention to the configuration of our hotel systems as well.  These issues take time and focus to address.

At the enterprise level as well, business intelligence provides fantastic opportunities to get started on data management.  Frequently, before a BI initiative, a hotel company may have allowed methods of market segmentation, room type classification, channel classification and such to be determined at a property level, with each property doing what makes sense for their own hotel.  Now finding themselves in a position where corporate-level hotel intelligence is desired (for instance, needing to know how much company X produced across our hotels last year), they are confronted with this mix of data standards.  Moving from this type of scenario to one with corporate data standards is facilitated by business intelligence.  Through the mapping and/or aggregation of values, a BI platform can serve as a virtual “layer” of standardization across properties – producing one corporate view with standardized market segments, channels, room type classes, etc.  In many cases, this “virtual” standardization is just what is needed until configuration of the source systems themselves can be standardized across the company, which is a lengthier process with more dependencies.

The rewards of a business intelligence initiative (properly executed) speak for themselves – we all want increased visibility into the factors that are affecting our business, so that we (and those around us) can make the best decisions possible.  Similarly, most (if not all) of us before starting a BI initiative – or deciding if we should – have some rather large degree of doubt about the accuracy and completeness of our data.  Those BI practitioners among us know not to wait for “perfect data”, but rather to get started now on a path to near-perfect data – a path that begins and ends with business intelligence.