Google Needs To Send Flight Search Back To The Hangar (Guest Post)

I’ve finished up my week at APEX, and it was a great one. That being said, I have a ton of stuff to sort through for future posts, so I’ll be pretty busy until next week. Today, I’ve got a guest post from Henry Harteveldt on Google’s new flight search.

Two days ago, Google unveiled a new flight search function. Following its successful acquisition of ITA Software, Google’s move was expected — only the timing remained unknown. But after a bit of testing, I think Google launched the product too soon. Since you never get a second chance to make a first impression, I don’t think Google did itself any favors by launching it when it did.

Like Kayak,, Hipmunk and other similar “meta-search” websites, Google’s new flight search function allows a user to search and compare airline schedules and prices. Google’s flight search certainly offers several benefits:

  • Convenience. Google is the most-used search engine, and adding flight search will only serve to increase the site’s utility to its users. As an advertising-based business, this obviously stands to benefit Google enormously as well — more people using more pages, spending more time on Google.
  • Page download speed. Once I’d selected my travel dates, Google returned results almost instantly. A search on Kayak produced a calendar on the right-side of the page showing the best available fares, but users must wait several seconds (six to eight seconds, in my tests) for all the actual flight results to be returned.
  • Proactive filters. Google’s flight search allows users to filter by both price and travel time. Both of these are smart. Research I did in my previous job shows that a large number of US online leisure travelers — likely the primary audience for this application — allow price to dictate their first-choice destination. The travel time filter is helpful (if not original — first offered this at least three years ago) for travelers who want to avoid long flights, especially if they’re taking a long weekend getaway and want to maximize their time at their destination.
  • Integrated — though slow — booking. After a user has selected both outbound and return flights, a red “Book” button appears (beneath that is an ad for the airline, read on to see how that will play out). Clicking on it will take the user to airline’s website where he or she may proceed to complete the reservation. In my experiences, though, it routinely took eight seconds or more for the airline website page to appear (I didn’t encounter other delays using the Internet or my web browser — Google Chrome — while waiting for the airline page to load).

That said, as analyst I think Google jumped the gun in bringing this product to market. Why?

  • Poor presence on launch day. Hint: If you’re going to tout that you’re bringing something to market, it’s usually a smart idea to let users find it. I did 10 searches on using the suggested “flights from [city] to [city]” phrase. None produced the “flights” link in the left-side navigation menu, as illustrated on Google’s blog. The “flights” link did appear in the searches I conducted today, September 14. Nitpicky? Maybe No. If you’re going to bring a product to market and tout it, make sure your users can consistently find and use it.
  • A clunky map. If there’s anything Google knows besides search, it’s mapping. Google offers two nice features on its flight search map. First, it shows the prices for the destinations based on selected travel dates. Second, the departure airport shows an icon of a plane taking off, while the destination city shows a plane landing. After that, though, Google — surprisingly — disappoints. It’s not at all clear how a user interacts with the map. To change origin and destination cities, a natural response would be to try to move the icons around on the map. That doesn’t work — doing so shifts the map in the window. To adjust the destination city on the map, a user must click a different city. To change the departure airport, users must change the “from” city in the flight search box.
  • Google flight search
  • Incomplete flight search results. Not all airlines serving a city-pair were returned. For example, a search for San Francisco-Washington DC (all airports) returned results for United, but no other airline. With a 6 1/2 hour flight duration, Virgin America’s (VX) San Francisco-Washington/Dulles nonstops should have been included (VX schedules those flights for roughly 5 hours, 10 minutes). On other searches, like Oakland-Houston, Southwest’s (WN)flights didn’t appear. I can understand not showing WN’s fares, but not its schedules, especially since media presented WN flight results. A San Francisco-LA search returned flights for American and United, but not Delta. A search showed 11 flights southbound, nine flights northbound on the same travel dates.
  • SFO LAX no DL GOOG flight search
  • Uninspired user sort controls. I give Google credit for allowing users to sort flight results by airline alliances — that can be very helpful, especially for international flight searches. But the other controls are fairly standard — number of stops, connecting airport, and departure and return flight times. This site has nothing on either Kayak or Hipmunk, whose user controls are more visible. Yawn.
  • A display that emphasizes airline commoditization. Yes, this is a new product. Yes, airline pricing is complex. But airlines are working hard to differentiate themselves, and Google’s flight display does nothing to help them. Google presents the least expensive economy-class fare. Nothing wrong with that, but Kayak presents premium economy fares. Airlines like United and Virgin America can’t merchandise their premium economy offerings on Google’s flight search tool. Carriers that offer amenities like in-flight Wi-Fi, in-seat power, or in-seat audio/video entertainment systems can’t promote those items in the tool, either. Hipmunk shows users which flight have Wi-Fi. These items are important to distinguishing elements for the airlines. Not offering at least some merchandising capabilities at its start places Google tangibly behind its competitors.

When a firm introduces what is essentially a “me too” product, its launch should incorporate enough bells and whistles to create a distinctive, compelling experience to encourage switching. As of now, Google’s flight search may be convenient, but it’s far from compelling.

Finally, there’s been a lot of buzz about how the flight search shows only airline websites. Cranky Flier readers no doubt know that airlines would like to increase their direct sales, primarily via their websites — their lowest-cost sales channel — rather than selling tickets through travel agencies. Don’t expect this to last. Remember that:

  • Google is a publicly-held firm. Like all publicly-held firms, Wall Street and investors expect growth in gross income and operating and net profits each quarter.
  • To help generate those revenues and profits Google sells advertising. Travel is a prime category. Within that category, online travel agencies (OTAs) like Expedia, Orbitz, Priceline, and Travelocity are major customers.
  • The OTAs generally have larger marketing budgets than the airlines. Both Expedia and Priceline have large cash balances, and have — or have had — larger market valuations than many airlines.

The OTAs will not allow themselves to be frozen out of Google’s flight search display. The OTAs are authorized agents of the airlines, so I don’t see how Google can prohibit them from participating in a public channel like its flight search tool. Plus, if consumers are not provided the same choice in shopping channels on Google’s flight search engine that they find elsewhere, Google will not get the traffic it seeks. Bear in mind that Kayak started as an airline-focused price aggregator, and then added OTAs as a booking option. That move must have worked, otherwise would have stopped it.

I understand, and respect, why airlines may not care to have OTAs compete with them in this channel. I don’t expect airlines to sit idly by, either. In talking with carriers, I get the impression that some may be willing to pay Google referral fees for users who book through their websites. Of course, it’s possible the OTAs may pay more. The OTAs can subsidize paying a larger referral fee on air tickets to a meta-search site, since they can recoup that by selling other travel products like hotels or insurance, to the air traveler. Though airlines sell third-party products, they don’t generate the same volume of those products as the OTAs.

Henry Harteveldt is a co-founder of Atmosphere Research Group, where he leads its airline and travel research practice. Prior to starting Atmosphere earlier this month, Henry spent more than 11 1/2 years as the airline and travel analyst at Forrester Research, Inc.

Get Every Post Emailed to You

There are 26 comments Comments


Your email address will not be published. Required fields are marked *