Advertisement
Airfare: The 'When to Buy' Conundrum

Last week, I reported on the recent findings from CheapAir about how far in advance you're likely to find the lowest airfares. The overall answer: Seven weeks. But the question has another shorter-term dimension: predictions about whether any fare you find is likely to go up or down during the following week.

The newest player in this game is Kayak, the big online fare and rate aggregator, which just introduced a new feature. For a flight search, you log on, enter the usual itinerary details, and the site returns not only the usual array of fares, airlines, and schedules, but also a "price trend" arrow with a forecast about what will happen to fares over the next week. On January 17, I started a test, checking fares that day and again a week later, on January 24. The sample test results:  

Advertisement

For a flight in mid-February:

  • January 17: New York to Las Vegas, $399, connecting; Kayak said to "book."
  • January 24 $272, connecting; Kayak again said to "book."
  • January 17: New York to London, $707, connecting; Kayak said to "wait."
  • January 24: $707; Kayak again said to "wait."
  • January 17: San Francisco to Hong Kong, $977, connecting; Kayak said to "wait."
  • January 24: $1,015; Kayak said to "wait."

For a flight in mid-March:

  • January 17: New York to Las Vegas, $272, connecting; Kayak said to "book."
  • January 24: $272; Kayak also said to "book."
  • January 17: New York to London, $745, connecting; Kayak said to "wait."
  • January 24: $804; Kayak said to "book."
  • January 17: San Francisco to Hong Kong, $968, connecting; Kayak said to "wait."
  • January 24: $949, connecting; Kayak said to "wait."

Bing Travel (formerly Farecast) previously had the airfare-prediction space to itself, and it's still operating. The process is quite similar to Kayak's—in fact, the websites look curiously alike—and it produced similar but not identical results.

Mid-February flight:

  • January 17: New York to Las Vegas, $399, nonstop; Bing said to "book."
  • January 24: $272, connecting; Bing said to "book."
  • January 17: New York to London, $797, connecting; Bing said to "book."
  • January 24: $707, connecting; Bing said to "book."
  • January 17: San Francisco to Hong Kong, $968, connecting; Bing said to "book."
  • January 24: $993, nonstop; Bing said to "book."

Mid-March flight:

  • January 17: New York to Las Vegas, $272, connecting; Bing said to "book."
  • January 24: $272, connecting; Bing said to "book."
  • January 17: New York to London, $797, nonstop; Bing said to "book."
  • January 24: $743, connecting; Bing said to "book."
  • January 17: San Francisco to Hong Kong, 959, connecting; Bing said to "book."
  • January 24: $993, nonstop; Bing said to "book."

The conclusion seems obvious: "Win a few, lose a few." Granted, my test was pretty limited, but it seems to show that the short-term predictors work some of the time but certainly not all of the time—which means you can't fully rely on them. Instead, my base recommendation remains the same as it has been for years: The best time to buy airline tickets is when they're on sale, and the best way to track sales is to sign up for as many email airfare and sale alerts as your inbox can handle. Airlines do them, the big OTAs do them, and SmarterTravel and Airfarewatchdog do them.

On a different question, Kayak doesn't predict changes in hotel rates. But my test revealed one serious problem: The initial rate display for a typical strip hotel in Las Vegas did not include the "resort fee" add-ons. You didn't even see anything about the fee until after you "select" a room, and even then, you don't see the fee unless you deliberately click "read all terms and conditions" far down on the booking page. Clearly, the fault lies with the hotels, not Kayak. Equally clearly, none of the several hotels I checked was complying with the recent Federal Trade Commission ruling on disclosing hidden hotel fees to consumers.

Ed Perkins on Travel is copyright (c) 2012 Tribune Media Services, Inc.

You Might Also Like:

Read comments or add your own insight!
Please enable JavaScript to properly view and use this web site.