When running a paid campaign (as in pay-per-click advertising, or PPC), Vizergy often bids on terms that we rank well for naturally (from our search engine optimization, or SEO, efforts), and we may lose some natural traffic to paid traffic. This phenomenon is referred to as “cannibalization." Because of this, some of our clients feel strongly that bidding on brand terms and terms they already rank naturally for is unnecessary. But we’d like to dispel this notion.
What We Wanted to Know
Assumption: We don’t need to bid on brand terms (my hotel name) because someone looking for my hotel will find me anyway.
Proposal: Online Travel Agencies (OTAs) are bidding on your brand, so someone looking for your hotel may end up on the OTA website instead of your website. The best case scenario here would be that the shopper still books with your hotel, but it would be through the OTA, so you’re forced to pay a commission. Worst case scenario, the shopper sees a better deal with another hotel on the OTA and doesn’t book with you at all.
Assumption: Bidding on terms that we already rank for in natural search is a waste of money.
- Natural rankings can drop and fluctuate from time to time. Then where will you be?
- You aren’t the only hotel ranking in natural search, and paid search gives you more real estate on search engine results pages. The more real estate the better.
What We Did
We picked a fairly new client, and with their permission, paused their paid campaign. We didn’t look at any traffic data from before we paused the campaigns; we wanted to start fresh (and we didn’t want to get cold feet if/when we saw the traffic drop).
To prepare for the test, we allowed two weeks with no paid campaigns running while we built very specific campaigns. The data collected in those two weeks provided a baseline – no paid traffic, all natural traffic.
We built one brand paid campaign and one non-brand paid campaign to run, and the non-brand campaign contained only high-traffic phrases (exact match) for which we ranked very high in natural search.
Part 1 – Running Non-Brand Terms Only in PPC
After the two-week period of running no PPC ads to establish a baseline, we launched the non-brand campaign and let it run for two weeks.
We saw pretty much what we expected. On average, the natural traffic went down a little over 3%, while the PPC campaign added almost 12% to our overall traffic. Very impressive, considering the relatively low number of phrases for which we were bidding.
Part 2 – Running Brand Terms Only in PPC
We then paused the non-brand campaign and waited a week to launch the brand campaign, to get back to the baseline. We then ran the brand campaign for two weeks. The results of Part 2 of the test surprised us greatly. Any difference in natural traffic was negligible when compared to the baseline, but PPC added over 8% to the overall traffic. After the two weeks, we paused the brand campaign for a week and then ran it again for a week, just to double check. We saw the same results.
Bidding on brand appeared to take almost nothing away from natural search, while adding significant numbers to overall website traffic. So we started digging into why this might be.
The culprit turned out to be a combination of the Google Carousel and OTAs.
What we found was, when a shopper does a non-brand search (like “hotels in Atlanta”), the hotel appears in the Google Carousel. When the shopper clicks that listing, instead of being taken to the hotel website or Google+ listing, Google automatically does a SECOND SEARCH. This second search automatically runs a brand search for the hotel and serves up another page of results where paid ads for the hotel are listed at the top of the page (the way paid ads are always displayed). The result is Google has added an extra step in the natural search process that increases the likelihood that a user will click on a paid ad to get to your website, rather than a natural search result.
We knew this is how the Carousel has operated since day one, but what we didn’t realize was how many OTAs were bidding on our clients’ brand names. Many even had ads that appeared, to the average searcher, to be an ad for the brand itself, not an OTA. By not bidding on brand terms in Part 1 of the test, we were losing a lot of traffic that originated from a non-brand search.
Spending a few days spot-checking, we found this to be the case for all of our franchised (Marriott, Hyatt, etc.) clients and many of our independent clients.
What This Means for Hotels
For hotels (especially franchise properties), bidding on brand terms and phrases is now doubly important because, thanks to the Carousel, brand searches can now originate (and often do) from non-brand shopping phrases. Your property brand terms have become one of the least expensive advertising channels, yet the most valuable. By not investing in paid search ads hotels can risk missing out on both natural and paid search traffic to their independent website. If an OTA has paid search advertisements on your brand name, and an online travel shopper clicks on the Google Carousel opposed to the search results, then that shopper is directed to the OTA instead of going to your property. The result is the shopper being more likely to book with the OTA and your property could potentially lose the entire booking. And with OTAs bidding on your brand… don’t you think you should too?
Vizergy is an award-winning digital travel marketing company that empowers hoteliers to know more, do more and achieve more – with the ultimate goal of delivering maximum online revenue. The company uses data driven decisions to provide strategic guidance in the evolving world of online marketing. Online marketing services include search engine optimization, pay-per-click and display advertising, local marketing, website and mobile website design and development, reservation software and advanced analytics. Since 1998, Vizergy has propelled thousands of branded and independent hotels, resorts and hotel management companies to success.
Logos, product and company names mentioned are the property of their respective owners.