Kayak and Hipmunk get a lot of praise for their UI but if you want to use a hacker interface for finding flights, http://matrix.itasoftware.com/ is the way to go.
It's not as pretty, but is incredibly feature-rich, not limited to the US and regularly finds me prices that are significantly cheaper than anyone else.
FYI - you can access most of ITA's routing language on Hipmunk, if you type :: after an airport code. The advantage of using us is we can make you a booking link to actually purchase the itinerary.
Thanks, that's interesting. The issue for me is that Hipmunk only serves the US market and won't let me enter multiple airports as source or destination. My best results have started with a list of 20 possible sources and destinations which is unfeasible to search with any other publicly available service.
Have you considered watching to see if a user searches for, e.g., flights out of JFK immediately after an identical search for flights out of LGA, and then gently highlighting the feature?
There're a slew of new flight booking startups like http://wuhutravel.com and http://sillk.co that I believe are trying to entice younger travellers by apparently launching features like you just mentioned raldi (based on screenshots I've seen on dribbble from their designers)...
Kayak, Hipmunk, Google Flights, Orbitz, and almost every other aggregator licenses ITA Software's system, actually. And most major airlines use it internally. It's good enough that it's pretty much the de facto standard and framework.
It was a sad, sad day when they shut down the old http://matrix1.itasoftware.com/ interface. It was much more keyboard accessible with faster navigation between inputs, and the forward and back buttons always did exactly what I expected. Also really miss the "weekend search". It was a really great service, but I guess it wasn't exactly making them money.
My guess is that they predict the most common searches and constantly query and cash results for these searches, something I don't think Kayak or hipmunk does.
I don't think you can launch a search product on google if it takes more than .xx seconds to return the result...
I'm not so sure, I looked for a few random flights. However, having built my own very rudimentary internal flight search tool from the global flight list, I can tell you there are roughly 7 million flights per month, a small data set by google standards.
It's not the size of the dataset that is relevant but more that this dataset changes in realtime : price + availability. These requests go through GDSs that are relatively slow, that's why so called real-time search engines like kayak are fairly slow. What do you exactly mean by 7M flights? Keep in mind that most search requests involve a combination of flights (leg), outbound, inbound and sometime multiple stops...
I think that owning ITA allows them to have more data to mine/predict and better access but it doesn't explain how fast the searches are.
Most airfare searches still hit GDSs which are relatively slow to respond. The only way I can think of Google can deliver such a performance is by intelligently pre-caching airfare searches. Google?
That's completely idiotic, and I hope you feel ashamed of what you wrote.
The problem isn't more or less information, it's that insider trading creates two classes of people: Those select few "in the know" who can reap massive profits by getting an illegal phone call, and the suckers.
This ad hominem comment is inappropriate. It is certainly not "completely idiotic". There is a lot of intelligent discussion on the issue in some places. For example: http://www.fff.org/comment/com0306f.asp .
This is not an "intelligent discussion." Did you read this article? He sets up a straw man by comparing specialized knowledge to facts in his first argument. Everyone can know the odds in Blackjack by card-counting, but no one should know the next card in the deck. This is a better comparison of insider trading because using statistics and predictions of market activities based on 'guesses' is perfectly fair, but actually knowing materially negative news that is confidential puts you at an unfair advantage in a market system. I mean, how far would you take the argument... so can the CEO sell his shares before the news? What about the FDA committee chair? How about if a scientist shorted shares before rendering his judgement that the drug should not be approved?
Insider trading is just another way for the privileged and wealthy to get richer and small investors and pensions to get robbed.
You have to ask yourself, what is the point of the market? I'd say to efficiently allocate resources, in order to do that one should be able to use any and all knowledge they have available to them -- it allows the market to find the appropriate price quicker. If small investors and pensions get "robbed" then they'll adapt or leave the market.
Or use whatever advantage you can get to dump the shares in the greater fool.
Having an elite crowd with an unfair advantage is going to make it hard to get future IPOs or companies issuing more stock to generate much interest in their offerings.
The purpose of markets is not to provide a casino. The purpose is the efficient allocation of resources. Don't know anything about sector X? Why the hell are you entitled to a fair shot at gambling in sector X?
You don't understand what an "ad homnimem" is. Simply insulting someone for saying something stupid is not an ad hominem. My argument would need to be of the form "That person is <unsavory characteristic>, therefore he is wrong in his argument".
That is not what I did, therefore it was not an ad hominem.
Please learn the definitions of phrases before you use them in conversation.
Fun fact: Both are located in Norwalk, CT., about a 5-10 minute ride from each other. Priceline is located almost on the Darien border and Kayak is nestled in South Norwalk (SoNo). I gotta believe this proximity lead to board member and executive coziness.
To save you from reading that document, a law firm has decided to 'investigate' whether Kayak's board "breached their fiduciary duties to stockholders by failing to adequately shop the Company before entering into this transaction and whether priceline.com Incorporated is underpaying for Kayak shares, thus unlawfully harming Kayak stockholders."
Breach of fiduciary duty lawsuits are very common for all major acquisitions. For the law firm it's high potential reward with relatively little downside. The mere fact of one or two being filed means pretty much nothing.
I respect what Hipmunk is trying to do, but I really doubt they are filling voids outside of "this" community. I find it much harder to scan, slower to load, and having lower quality tools for finding alternate (or cheaper) fares.
It's interesting, considering they will provide round-trip flights and list the London to NY flight(s) as a separate item, which would make me think the data is there.
Hopefully Priceline will learn some improvements from Kayak and not try to force changes the other direction. Kayak has a great UI overall, but the explore interface is pure awesome: http://www.kayak.com/explore/
Visually that kind of interface is very appealing, but fares may be misleading... once you start the booking process, you will end up with completely different numbers. Typically, a featured "as low as $" air fare is a red eye, 8 hour connection time etc. I was going to output specials myself too, but then commented them out - as they hardly ever apply.
That being said, Kayak does have a great UI, and their iOS apps are eye candy.
I was getting ready to say almost the exact same thing -- thank heavens that Hipmunk came when it did, or that acquisition (if soured) would have left a very large hole in my travel-making tools.
I don't particularly like Kayak, but as a matter of effectiveness, there's no doubt that it was pretty much king of the hill for a good long while. Since Hipmunk's inception though, it has definitely been the 'go-to' place for travel-planning, and beats the nearest competition (that I've used) by a mile.
I would also venture to say that this was perhaps a timely exit for Kayak, as the more and more traction Hipmunk sees, the less off the Kayak founders would have been for an exit.
I'm working on a travel (many)weekend(s) project myself yet to launch, so I won't be too harsh, but... :P I put Warsaw in, and the suggested airports were Okecie (correct) and Babice (it's in Krakow), Modlin was missing. And all the result were in the USA. I like the design theme with all the backgrounds etc. and the logo, cool idea. You should leave some way to contact you with feedback on the page. Keep on, good luck.
sorry, when I've read Babice, I thought Balice (Krakow), it's late. Babice is of course indeed in Warsaw, even nearby me, but it's for small aircrafts and stuff, not passenger airplanes.
modlin now appers, but still don't know what babice are doing there :) yep, as I said, it was just recently opened as a passenger airport, and major european discount brands fly from there so by now it should be in every db.
if you want me to help you sort out these issues, feel free to drop me an email; I've had various ideas around travel websites (the one I'm working on is not competing with yours) so I'll be happy to test it out or just chat.
zalew: airports - the UI lets you add your own. But I will look into Warsaw... I am currently looking at airports within a 50 mile circle, is Modlin farther than that?
About the results: if there's no snow, a ski resort will most likely not be shown to you.
The destinations were hand-picked, e.g. 2012 Best All-terrain, or best night life, or... But the list is growing, we just started to add European locations.
I like it, I've been looking for something like this. Some features (dreaded feature creep!) I'd love:
1. Map interface with little lines from my home base airport to the possible destinations - It's just fun to see.
2. Sliders to increase/decrease budget - Dynamically adding/removing routes as I change the $$ limit.
Maybe these would only take another weekend to add? </joke>
I am hesitant about including budgets, mainly because they can be so misleading, and it's hard to actually get that price once you start booking. I had included "staring from $xxx RT", but then commented it out just prior to launch.
jngreenlee: I was actually thinking about the map yesterday, but couldn't think about a good way to visualize so it doesn't look like another google maps mashup with pins. Your idea with air routes is interesting, hope you don't mind if I just go ahead and borrow them? :)
Main problem, in my opinion, is that most solutions are airport-centric. Say you want to Ski in Alta UT. You can't typically book travel from NYC to Alta, you have to specify "from" JFK/LGA (or NYC) to "SLC", assuming you know exactly where you are going, what the best airport is to fly to etc.
Some let you select "ski" destinations (as Kayak Explore does), so it may send you to SLC (Salt Lake City), show you weather for Salt Lake City, then try to up-sell you on a hotel in Salt Lake City (or maybe Sandy UT).
Mine's is destination centric. Where you fly to, where you check the weather, where you ski, and where you book your hotel (or even buy lift tickets from) are not necessarily the same "place". Make sense?
Thank you, I just "launched" the site a couple of days ago, so nobody knows about it except my FB friends. Airport data is open-source, although it needed some cleaning (to filter out things like railroad stations etc), weather forecasts are from Wunderground, but the main asset is Destinations. They've been hand-picked (a la Mechanical Turk), although the list is growing. Mainly focusing on ski destinations now, so the starting data set is places that I find interesting for whatever reason. I am experimenting with a couple of other data feeds, trying to gauge their relevance. The list of recommendations you are shown is not random, more interesting places (with better conditions at the time) bubble up.
Traditional travel sites are airport-centric. Say you want to Ski in Alta UT. Most would try to send you to SLC (Salt Lake City), show you weather for Salt Lake City, then try to up-sell you on a hotel in Salt Lake City (or maybe Sandy UT).
Mine's is destination centric. Where you fly to, where you check the weather, where you ski, and where you book your hotel (or even buy lift tickets from) - are not necessarily the same "place". Does this clarify things somewhat?
P.S. Just checked real-time Google Analytics, you guys (HN are awesome, stress testing the site :). Do you think I should do a "show HN" at some point?
I love that you're using Mechanical Turk. Thanks for the explanation and congratulations on launching!
I would love to have my own flight search engine, I think there's still loads of room for improvement even with Kayak (a billion dollar + comapany), Hipmunk, Skyscanner out there
This is awesome! I noticed that the "Search for the Lowest Air Fares" page seems different in design from the other pages. Are the forms embedded from another site? Which site did you use?
There is always http://hipmunk.com, which I think has the best UX of any of them (although maybe it is lacking some features, but none I would use anyhow).
I'd recommend www.skyscanner.net, at least for the European users.
Sure, they lack simplistic interfaces of Hipmunk or Kayak but when it comes to actually finding the cheapest tickets and best connections on the old contintent low fare airlines market they're, IMHO, the most reliable.
I'm a big fan of skyscanner. Being able to ask questions like 'what are the cheapest flight from my local airport over this weekend' make for fun trip planning.
I've found Google Flights[1] to be a much better experience than either. It is stupid fast and has a great interface for comparing prices if dates are flexible.
Hipmunk looks interesting, haven't seen it before. Minimalist is the way to go, I think. Can't stand busy travel sites that try to be everything to everyone.
To compare prices between three airports in the same metro, couldn't you use the "include nearby airports" function and just look at the filter list and see the lowest price from each? Just wondering if you've tried that vs. running 3 different searches.
Pretty tangential, but why is it that none of the flight search engines include Southwest? I realize that Southwest doesn't give their fare info to whoever the other airlines give it to, but what's stopping someone from just scraping southwest.com every hour or so for the current fares? Or even if scraping is against their TOS, couldn't they hire one data entry employee to manually go through and add Southwest's flights every day? Or is it somehow illegal to publish Southwest's prices?
I imagine it would be a big competitive advantage for whichever one did it first - it's pretty annoying how every time I want to search for flights I have to first search on Kayak/Hipmunk/GoogleFlightSearch and then separately go to southwest.com and wade through their slow, awful search interface.
Southwest does not believe in distributing their fares through the metasearch or OTA channel. They want to own the customer and make sure there's only one destination to book Southwest fares and that is Southwest.com and it's worked out pretty well for them.
The hard part for scraping is that it's both against their TOS and you wouldn't be able to have accurate availability and price information through manual data entry. The nature of how frequent price changes and the number of possible combinations of fare types/routes/availability is what gave rise to companies like ITA.
> The hard part for scraping is that it's both against their TOS and you wouldn't be able to have accurate availability and price information through manual data entry. The nature of how frequent price changes and the number of possible combinations of fare types/routes/availability is what gave rise to companies like ITA.
Anecdotally, whenever I've checked Southwest prices multiple days in a row they usually stay the same. So I'd imagine it would still be a valuable enough resource if there were someone who did this manually at the granularity of a day, even if when you clicked through to buy the tickets they occasionally wouldn't match the price the search engine told you. You could even have a "report this price as incorrect" button.
Yes, but if you're a website serving millions of customers per day, you need a contractual arrangement to A) have up to the moment pricing, and B) agreement that the airline will honor the prices published on your site.
Because airfare prices follow an upward trend as flight time approaches (although they sometimes taper as the flight date gets really close), you'd inevitably publish prices that are out of date and lower than the actual fare. This is unavoidable because you're relying on scraping, which involves polling, in which there is always a delay due to polling intervals. You'd stand to lose millions of dollars every time a price increase occurred between the time that a customer decides to book a ticket at one price, and the time your polling system picks up the increase.
Circumventing Southwest's desire to avoid this channel is not a good business model.
Interesting, I've never noticed Southwest results in my Kayak searches before (maybe it's new). But when I try it, it doesn't actually give the price as in your screenshot, it just has a link labeled "Info" (which goes to the southwest.com homepage) where the price would normally be. I wonder if it has something to do with you being in the UK?
And where does ITA get it's data. I assume there must be a few competing standards, and the airlines use one of them to give their data to ITA. Why can't other companies get the data directly from the airlines?
In particular, for the nine months ended September 30, 2010, Expedia and its affiliates, including its Hotels.com and Hotwire subsidiaries, accounted for 25% of our total revenues. Also during this period, Orbitz and its affiliates, including its CheapTickets and ebookers subsidiaries, accounted for 19% of our total revenues.
I guess Priceline wants to get a first shot at all that business, and then to get paid by their competitors for bookings they don't get.
Related: Expedia has an API... http://developer.ean.com/. Haven't checked it in a while - looks like only hotels are available now (no flights, cars, cruises, etc.)
Of course, not saying that Kayak went through this route - this is just something available to the hoi polloi.
I looked at Expedia for somewherenice.net, and didn't choose them for that reason. Needed to be able to package the whole trip, from airfare to lift tickets, so ended up working with Travelocity.
In recent times, Kayak has been shifting to push customers directly to airlines and hotels, instead of pushing them through online travel agencies like Priceline.
I suspect this is Priceline admitting that it (and other OTAs) are losing relevance in preference for the meta-search model.
Smart move, and I don't think they're dumb enough to ruin it. Read up on how successful Priceline was with their Booking.com acquisition.
I'm not remotely worried about priceline ruining kayak.
Kayak was the best interface before hipmunk, and it may (or may not be) the best interface now.
But there will be more improvements in travel buying interfaces... whether those improvements come from kayak or someone else, Kayak's current interface will seem kludgy in a couple years.
That's really bad news for Expedia. When you book a hotel room directly on Kayak, over 90% of the bookings go through expedia and hotels is where these travel agencies make most of their money. (Not airfares, they have become a commodity, thanks to Kayak :) )
Since everyone seems to be putting in their 0.02$ on which sites they use for booking, bing.com/travel is also very good. The buy-wait prediction is fantastic.
It's good to see people still using Farecast! Most of the original team is now working on other startups including Decide.com, Medify.com, and Room77.com.
It's not as pretty, but is incredibly feature-rich, not limited to the US and regularly finds me prices that are significantly cheaper than anyone else.