We put a lot of thought and effort into coming up with new ways to help our clients reach their customers efficiently and effectively through Hotel Price Ads. We might be nerdy by nature, and we may tend lean toward huge data sets, analysis, technology, algorithms, and models to build a picture of a user… but we also know the value of sitting down and just observing and taking notes.

With that in mind, we’ve done an informal study on how real users search on Hotel Finder. Our goals? First, by better understanding the end customer, we believe that we can better structure our campaigns and optimize those campaigns both directly and indirectly. Second, we want to see how users are liking the new Hotel Finder design and feature set. By paying attention to this, we can better understand how successful Google might be in capturing additional hotel comparison traffic in the long term.

Methodology

We recruited a handful of folks that had experience booking travel online, met our desired (but broad) demographic (25-40, >$40k income, intermediate internet experience,) and sat them down for a brief introduction to their task and started them off on Google.

The prompt: You’re planning a vacation for 4 days and 3 nights to Boston, Massachusetts in early October. You have a total of $1,400 to spend on your hotel while you’re there, including taxes and fees. Your goal is to find the “best” hotel within your budget, however you’d define that.

(Note that we also made sure testers did not click on any advertisements during the test.)

Once the users were familiar with the interface and interacting freely with Hotel Finder (it was each user’s first time on the site,) we provided them with some tasks to complete specifically:

1. Try to find the best hotel for your trip, doing whatever research is needed. If you don’t find what you’re looking for on Hotel Finder, you can leave the site. Just return to Hotel Finder for the next task.
2. Find a hotel that has the specific amenities most important to you.
3. Try to find a hotel near a place you’d like to visit or a landmark in Boston.
4. Without clicking on the room booking buttons, choose an offer that is best for you.

And finally, on completion, we asked a number of follow up questions to get a few more details:

1. Which hotel did you choose, and why?
2. What features might have helped you make an even better decision?
3. What were the most helpful features on the site?
4. Would you use the site again to book a hotel, or would you use another site? If another site, which?

Observations

Overall, user experiences with the new Hotel Finder interface were very positive. Each user was surprised that Google had that kind of hotel rate search functionality. The booking path for each of the users was roughly the same. They didn’t interact with the list of hotels in any signifiant way before starting to apply filters. The first filter (likely influenced by our focus on a set budget) was price, and then immediately users began to focus on ratings and reviews. Each user described themselves as at least fairly confident in their selection given the data available.

• Users seemed surprised that they weren’t able to sort by descending price. The default action when you select “Sort By Price” is to show the lowest price first. Since there were so many results and the relevant hotels were at the bottom of the list, it was interesting to see each user double take at this functionality. They quickly adapted to the price range filters, though.
• Another consistent was that users either did rough calculations in their head or pulled up a calculator to determine what the maximum they could spend per night was in order to stay under budget. Each user noted that it was strange there was not total price available for them to check against, just the average nightly rate.
• Watching users filter by the amenities important to them brought up a common issue: filtering by amenity made properties disappear too quickly! Each user with relatively few amenities was able to get down to an extremely short list of properties.
  • The suggested solution here was to show the number of properties that would be available once selecting or deselecting an amenity. We couldn’t argue with that!
• Users found a bit of shortcoming in Hotel Finder when looking for attractions around a property within Hotel Finder. Every user went to the “Location” filter to try to select a major landmark, but that is not currently how the filter works. Trying to type the landmark into the search box on Hotel Finder was not fruitful either; the autocomplete was not graceful here and the results were poor.
• It is pretty well documented¹ that booking travel is a complex process taking place over many sites, so it wasn’t a complete surprise to see users check the reviews in Hotel Finder against reviews in other sites like Yelp and TripAdvisor, but it was certainly enlightening. One user paused to reconsider her decision when she saw that the property she chose had a 1-star lower rating on Yelp than it had on Hotel Finder.
• When choosing an offer to ultimately select (in this case, which Hotel Price Ad they would have selected in a real situation,) both low price and brand recognition played a role. In one case, a user chose a slightly higher advertiser because he had “used them before and been happy with them.” We also noticed that users didn’t seem to “get” that there were many more offers behind the more button. None of them interacted with the brands to see what other prices were available.
• When talking about what hotel they ultimately chose and why, the following keywords and themes came up: right price or rate, right location, lots of reviews, good value, brand recognition… No big surprises here! The features that users wanted to see more of: searching by landmarks, adding keywords to search… The most helpful features: photos, reviews, amenity list, filters, dynamic map… And maybe most importantly, would they use it again: mostly yes, one no because she preferred a competitor for the warmth of the site and liked to be able to search by landmark.

Next Steps

The users we tested (and of course, this is too small a sample size to make any really broad conclusions!) were pleasantly surprised with Hotel Finder and maybe even a few will become loyal users in the future. But how can we use this insight to improve our digital campaigns?

It’s great to see that Google is iterating on the product. It helps give us confidence that investing in Hotel Price Ads campaigns is a smart play for travel brands. Reducing the amount of research a user has to do across multiple sites could help reduce costs.

This experience is a great reminder of the direct and indirect impacts to click-through and conversion rate. It will be interesting to dig in further to see when it is worth it to jump out of the “More” section of the offers to capture traffic if users just aren’t getting there by themselves. We’d like to see some testing around showing the total price to determine how that would affect advertiser conversion rate. It’s also clear that a user is still going to be hopping from site to site, so brands need a good remarketing solution in place to be able to bring them back in. Having some indication that brand recognition can help secure a booking even if the fees are a little bit higher certainly has some interesting implications as well.

If you’re interested in seeing some of the live user testing videos, get in touch and we’ll be happy to share.