Using online behavior to measure the travel industry’s recession-based injury and recovery

By Lincoln Merrihew | March 1, 2010

Reading time: 7 minutes

Abstract

An examination of Web search-related data from 2007 to 2009 charts the travel industry’s recession-induced slump and points at a possible turnaround in the offing.

Research Topics:: Consumer Research | Online Research | Quantitative Research | Web Site Analysis
Industry/Market Focus:: Hospitality Industry | Internet/Web | Tourism | Travel
Content Type: Research Methodology

Share Print

No longer flying blind?

Editor’s note: Lincoln Merrihew is managing director - travel, automotive, petroleum at Compete Inc., a Boston research firm.

The U.S. is in the midst of an historic recession and curtailed spending has affected almost every industry. The travel industry suffered more than others because consumers were more likely to cut back on discretionary/non-essential items and travel is largely discretionary. In a study undertaken for the purposes of this article, Compete Inc., a Boston research firm, found that consumers have responded by changing their behavior: researching/shopping in different places, researching/shopping a greater number of sources, or not researching/shopping at all.

Going into the study, a number of hypotheses were considered.

Hypothesis 1: Consumer online behavior, when truly aggregated across supplier sites and online travel agencies (OTAs), forms an accurate measure of travel industry demand.

To ensure data stability, Compete used a consistent process across all sources and months, using specific activities within sites as behavioral indicators, and avoiding double-counting and false positives (more details on methodology below).

The travel industry has responded to the market and changes in consumer activity in a number of ways, by:

cutting prices, trimming amenities and exploring new fees;
stepping up loyalty-based programs and/or efforts to reach out to new audiences;
cutting advertising (to reduce costs) or ratcheting up advertising (to increase share of voice/advertising ROI);
taking some current capacity offline, as possible, and postponing adding new capacity; and
exploring incremental revenue and income options, such as through cross-selling.

These were aggressive tactics to say the least and on a scale not seen in decades. The industry’s aggressive responses led us to our second hypothesis:

Hypothesis 2: While sales, booking and revenue are under pressure, consumers still exhibited mid-funnel interest in travel online as the industry responded. (Looking at the travel industry sales funnel, awareness of a brand would be near the top, booking would be at the bottom and researching online, as discussed here, is midway or mid-funnel.)

Compounding the industry’s ability to plan for and manage the recovery is that the longer and deeper a recession, the less likely consumers are to return to pre-recession behaviors. The current recession certainly falls into the “longer and deeper” category. Besides its magnitude, this recession also adds an additional unknown: As the first major recession in the digital age, there is no benchmark for recovery. So while the industry will recover, it is in many ways flying blind. Will consumers return to the same brands? Will loyalty programs become more or less important? Will consumers travel less? Will they return to specific brands and suppliers or continue to use third-party sources?

Consumer online travel research behavior and volume provides an illuminating view at bookings potential for the industry, which leads to our third hypothesis:

Hypothesis 3: Online behavior is a leading performance indicator, and the combination of early economic recovery and aggressive tactics by the travel industry should be clearly visible.

Methodology and analyses

To evaluate the hypotheses and help set the stage for charting the recovery, Compete analyzed consumer online research behavior across sites and years. We analyzed the industry and specific travel categories: hotel, cruise, flight and car rental.

The analyses were based on the number of unique consumers that visited sites in aggregate that together represent those travel categories. The sites include supplier sites as well as the category-specific sections of OTA sites. Unique means that we avoid “false positives” by not double-counting people who 1) went to the same site in the same month or 2) went to more than one site in a set of category sites in the same month. These data are all based on observed online behavior - actions, not recollections. The behavior tracking is all permission-based and does not use cookies (so data validity is not subject to the risks of cookie deletion).

Using the above methodology, a consumer who visits a given hotel site more than once in a month is counted only once. Likewise, a consumer who visits four different hotel supplier sites and three OTA site hotel sections in a given month is also counted as only one person. Compete also tracks visits, page views, session and time-on-site but here we focused on unique visitors.

We compared same-month data across 2007, 2008 and 2009 to account for seasonality for the entire industry. We then did the same for each category separately to help reveal the extent to which industries suffered and recovered at different rates.

Note: Results are based on domain-level visits to the sites in each of the categories. Compete also tracks behavior deeper into the purchase funnel and toward bookings, but those are beyond the scope of this analysis.

Findings

Industry

Across the four major categories, the industry is showing a blend of recovery and seasonal patterns. In general, consumers are shifting more of their travel research online, so all else being equal, online research and shopping activity should be trending up. Hence the slight decline in volume in 2008 represents the impact of the recession. Indeed, the lowest volume across the period is November 2008 - when the brunt of the recession took hold (Figure 1). The recovery of growth in 2009 is a key indicator of consumers’ return to travel research.

Category

Compete analyses suggest that the hotel and cruise categories recovered faster than did airlines and car rental, based on comparing monthly 2009 results vs. the same period in 2008. The hotel and cruise categories posted sustained year-over-year gains starting in early 2009 (Figures 2 and 3) with aggregate traffic in 2009 above 2008 levels in all months. The gains coincide both with the U.S. emerging from the recession and, as significantly, with very aggressive price-cutting and deal-creation by these categories. The results also highlight the dip in consumer traffic during the brunt of the recession (Q4 2008) and more so than simple seasonal patterns would suggest.

In contrast to hotel and cruise, the airline and car rental categories took longer to recover (Figures 4 and 5). Sustained airline traffic did not lift year-over-year until July 2009; sustained car rental year-over-year gains did not start until June 2009.

Airlines and car rentals also show the effects of the Q4 2008 recession bottom. Both had 2008 Q4 traffic below Q4 2007, with car rental showing the longest and deepest drop in that period among the four categories shown.

Conclusions and evaluations of hypotheses

In summary, all four categories are showing signs of recovery, and in general consumer research levels are leading indicators, pre-dating the industry’s overall financial rebound. True recovery will of course entail more shoppers but it also needs to include more bookings and fewer discounts/rates closer to pre-recession levels.

Hypothesis 1: Consumer online behavior, when aggregated across supplier sites and online travel agencies, forms an accurate measure of travel industry demand.

Validated: The data clearly show seasonal patterns as well as the pause in online growth during the worst months of the recession.

Hypothesis 2: While sales, booking and revenue are under pressure, consumers still exhibited mid-funnel interest in travel online as the industry responded.

Validated: The changes in consumer behavior to date were less significant than bookings and revenue declines for the industry overall. This supports the concept that impacts were greater at the bottom of the funnel than in the middle. It also reflects the industry’s relative success in keeping customers somewhat engaged as they developed and promoted discounts and loyalty programs.

Hypothesis 3: Online behavior is a leading performance indicator, and the combination of early economic recovery and aggressive tactics by the travel industry should be clearly visible.

Validated: All four categories are showing visible signs of recovery, though at different rates.

Signs of recovery

What are some additional signs of recovery to watch for? While the study results show the emergence of recovery, more data points are needed for a true clean bill of health:

Continued growth in mid-funnel online travel activity. This includes growth beyond what the overall shift to online would suggest.
Trickle-down of mid-funnel activity into the lower funnel. This includes all steps from research to shopping to booking to loyalty program enrollment.
Changes in upper-funnel activity on search engines. Consumer search-term patterns should evolve. Example: use of branded terms could increase while use of non-branded terms (including those that contain “deal” and “discount”) could decrease.

Logical actions

Given that the industry is hinting at resurgence and that there is no benchmark for a digital-age recovery, logical actions for the travel industry to help optimize performance during the recovery include:

1. Time changes in pricing, fees, amenities and active capacity with changes in consumer behavior by site and by industry. Use growth in each step of the funnel as a bellwether in general and within the context of the industry overall. Track and respond to changes cross-shopping within sub-industry and across the industry.

2. Assess digital brand health coming out of the recession. Consumers may or may not return to pre-recession behaviors and many consumers may have found new “brand best friends.” To manage pricing and capacity decisions, brands need to make ongoing adjustments to expectations based on the extent to which consumers return or are available for conquest.

Search engines are among the best “digital funnel” health indicators. For example, compare the extent to which customers shop your products after searching on a brand name or on a generic term.

3. Optimize the balance of quantity and quality of consumer traffic. Traffic quantity and quality often move in opposite directions and the recovery will mean acquisition costs will change and at the same time yields will change. These can be measured based on the extent to which traffic from various sources actually books (on a given site or anywhere).

4. Evaluate opportunities for cross-selling strategies. The industry is placing increased emphasis on the importance of where you book. If this resonates with consumers, it sets the stage for travel companies to expand what they offer to consumers (much like Google has migrated from search provider to phone provider). Understanding how consumers research across categories and brands can inform the best cross-selling opportunities and strategies.

5. Adjust loyal customer yield expectations. Quantify the extent to which your loyal members have also become loyal members of rivals and across travel. The less “pure” a customer’s loyalty, the more difficult it is likely to be to convert loyalty to sales.