Assessing the impact of rebranding in the hospitality industry

Using a panel dataset on hotel rebranding, we quantify the impact of rebranding on the occupancy rate and other performance indicators of the hotel properties. Our annual data consist of franchised hotels that changed brand names (173) and those that did not rebrand but were located in the same market although not the same zip code as the rebranded hotels (1972), over the time period 1994-2009. Our analysis controls for hotel property-specific, time-invariant unobservable factors (hotel fixed effects), hotel-specific time varying observable factors (such as additional facilities like restaurants, maintenance and marketing expenses etc.), market, time and market-time-specific unobservable factors (to account for e.g., varying economic conditions over time and across geographies). To account for any residual hotel-time specific unobservables that could be correlated with the rebranding decision (e.g., unobserved competitor reactions), we use the variable part of the annual brand-level franchise fee proportion paid by the hotel to the brand as an instrument for the rebranding. An increase in franchise fees (that are common to all hotels in a brand) increases a property's costs and could influence its decision to rebrand independent of its local demand conditions. We find that rebranding results, on average, in about a 5% increase in occupancy rates with increases in other metrics such as revenues per room. We then decompose this effect into a component attributable to the brand identities before and after rebranding and a component reflecting the average "match" of the hotel property to the new brand (relative to the previous one) and find that about half of the rebranding effect can be attributed to each component. We extend our analysis to look at heterogeneity in rebranding effects along dimensions that are relevant to the hospitality industry including (i) differences in effects for economy /midscale versus upscale / luxury properties---we find stronger effects for the former group of hotels; (ii) differences within and across brand "umbrellas" (e.g., various brands in the Marriott umbrella)---we find rebranding across umbrellas to have a slightly larger effect; (iii) across branded and independent hotel "flags"---we find that rebranding from one branded to another branded property and from an independent to a branded property increasing occupancy; and (iv) up to 2 years versus longer duration effects of rebranding---we find over longer durations the match value effect is bigger than the brand effect (with the caveat of possible survivor bias with this finding). We assess the robustness of our results to various model assumptions as well as to using matching estimators for our analysis and to exploiting rebranding as a consequence of hotel mergers as a means to measuring rebranding effects. Finally, we consider the impact that rebranding might have on competitors' properties.