Website revenue tracking – the conundrum

Website revenue tracking ? the conundrum

As marketers, we love to generate and analyze performance data. Getting a better understanding of which initiatives are working and which are missing the mark affords us the ability to refine our focus and generate the best possible ROI. Tracking, however, is still a very imperfect science. Reviewing one metric or one source of data often leads to narrow analysis.  When decisions regarding spending on media are made based on one metric or one source of data, money is left on the table.

We have come to understand that the hospitality industry too often falls into the narrow metrics trap. The past 18 months have seen a dramatic rise in the focus on return on ad spend (ROAS). On the surface, the focus on this metric could not be more appropriate. ROAS details the amount of revenue directly attributable to a specific marketing effort. It is a way to account for the return on every marketing dollar.

For example, ROAS is being used to track the effectiveness of individual terms in a Google paid search campaign. Imagine a hotel is bidding on the term ?Boston Hotel? and ?Downtown Boston Luxury Hotel.? At the end of the month, the hotel evaluates how much was spent on each term, and how much revenue was generated from each term. The next month, terms that did not produce an acceptable ROI are turned off in favor of terms that performed better.  Simple right?


Decisions about where precious marketing dollars are allocated are serious. Make the right decision and you gain market share; make the wrong decision and you miss opportunity. To base this decision on a single metric is fool?s gold. Here are just some of the flaws in ROAS tracking as it stands now:

A future guest searches for ?Boston Luxury Hotel? on Google.  They find your property?s ad and visit your website. They like what they see but are not ready to make a purchase. The next day, they return to Google, search for your hotel?s name and ultimately make a reservation. Which term will receive the credit for the revenue?  Answer: Unless you utilize Omniture for your website analytics (an expensive proposition) the hotel name will get the credit.

o   One of the stated advantages of paying for Omniture is that they allow for ?Keyword Stacking.? Keyword stacking is the process whereby Omniture recognizes someone as a repeat visitor and credits revenue to the original search term.  However, as we have tested this functionality we found significant flaws in how their data integrates with the website booking engine, leading to highly inconsistent data. 

?         A future guest performs a search, finds your website, then calls the property. No revenue would be assigned to that keyword.

?         A future guest searches for and finds your property while at work, then goes home at night to make their final purchase on their home computer. 

?         A future guest has their browser security settings turned on to not allow you to track them. No matter what they search for, the revenue from their visit will not be attributable to their inbound marketing channel.

?         To track ROAS, your website must interact with your website booking engine. SynXis, TravelClick, GenaRes and many of the other primary providers have recently offered ROAS tracking.  However, the methodology for how each provider tracks ROAS is different. They support different browsers, have different requirements for how data should be passed to them etc. Each of these variations has strengths and flaws, leading to incomplete and inconsistent tracking. 

We have a recent example of how ROAS tracking can lead to narrow decisions. We developed a Google Paid Search campaign for a luxury hotel in the United States. The campaign had over 1,200 target keywords ranging from ?Name Terms? to ?Broad Market Terms? to specific ?Event Terms.? The campaign was launched in the midst of peak season and we instantly saw a 40% spike in website revenue. No other changes were made at the time to the property?s marketing efforts, pricing or merchandising. However, the campaign as a whole only generated a 300% ROAS. The goal from the property was to target a 700% or greater ROAS.  We worked on the campaign for two months. While we continued to see the property website revenue grow, leisure transient market share grow and overall visibility grow, we simply weren?t able to move the ROAS to 700%-plus. 

Told that ROAS was the #1 metric the property would review, we were left with no choice but to change our strategy. We shut down everything but ?Hotel Name Terms? and let the campaign run. One month later we were heralded as a great success. ROAS was well over 1400%. The fact that website revenue had declined suddenly and precipitously was disregarded.

Tracking ROAS is important, and should be considered with other important metrics to measure the effectiveness of your marketing efforts.  But understand it is a flawed metric as described above.