Here's the backstory and my thinking process:
I haven't traveled outside of the US for almost 3 years and I've been itching to take a long vacation. At the same time, my grandpa has been sick in Shanghai, so it was my goal to go see him.
I decided I might as well combine my visit to Shanghai with also seeing Singapore and Hong Kong (two cities I've been wanting to visit for a while and had even thought about moving to).
I then decided to maybe also stop by Japan and whatever else cities that I can check out. (I like heavily densely populated areas.) So my trip basically evolved into visiting as many Asian Global Alpha+ cities as I can, and it required several flights to hop between the countries.
As I thought about the number of plane tickets I needed to buy to accomplish my vacation, the engineering side of me kicked in. At work I frequently deal with requirements. One thing I've been taught is that requirements (the goal) shouldn't define implementation (how to reach the goal). That is, you set your goals (e.g. I need to make a website that sells water), and you let the implementation part be separate from the goal (e.g. the website can be built on PHP, Djanog, Ruby on Rails, etc.). The idea behind this is that the people who set the initial requirements might not be the people doing the implementation, or might not even know how to to do it, while the people doing the implementation might be experts at what they do.
In my case, my requirements were to:
1) visit Japan, Singapore, Hongkong, and Shanghai, within 21 days,
2) with Shanghai being the last stop and taking up about 10 days while the other cities took 2 or 3 days each,
3) and do so at the cheapest price.
Those were the requirements and not much else mattered. (What airline, which city I visited first other than Shanghai being last, how many stops/layovers, etc.)
I wanted a site that allowed me to search like that, and the closest thing I found was http://matrix.itasoftware.com. The site allowed me to enter numerous cities as starting places and numerous places as destinations. The site was useful, but since it didn't search all the airlines (just like how Orbitz doesn't have Southwest), I often used it as a double check to see if I was getting a decent deal.
After some goggling, I luckily came across this site http://www.theflightdeal.com, which told me that there was a deal going on for flying to Asia through using Orbitz, and after tweaking around some Orbitz searches, I found that leaving LA on a Thursday or Friday and coming back on a Monday was the cheapest. I also found that I could set my return flight be from Shanghai to LA while having my "To" flight be a different city. Setting the "To" destination to be Tokyo Narita was the cheapest ($692 for LA to Narita and Shanghai to LA). This more or less caused my initial destination to be Japan.
Then I discovered that setting the To destination as Tokyo Haneda (Haneda = another airport that is closer to downtown Tokyo than Narita airport) was only a few dollars more ($716), and then I found that setting the To destination as Osaka was only a few more dollars than that ($725), but automatically included an 8 hour layover in Haneda. I decided that I also wanted to see Osaka, and train or domestic flights from Tokyo to Osaka would have been more than $9. At that time, my other searches for flights leaving Japan to my other destinations (see spreadsheet below) also revealed that there was a Tokyo to Singapore flight that leaves at midnight, whereas there wasn't one for Osaka to Singapore. I speicfically wanted to leave at midnight so I could sleep on the plane (save time and lodging costs). So I bought the tickets to Osaka and was planning on taking the train from Osaka to Tokyo, and then leave from Tokyo to Singapore.
After more goggling, I also found that that low-cost airlines of Asia that flew internationally were Tiger Air (Singapore), Air Asia (Malaysia), Fly Scoot (Singapore), Spring Airlines (China), Jet Star (Austrailia).
Also, Kayak did a pretty great job of revealing airlines that other flight search aggregators did not reveal. For example, it introduced me to Fly Scoot!
Once I was already set on arriving into Asia at Japan, I then needed to narrow down where to go next, and I made an excel sheet to jot down the cheapest flights I found, like so:
|172, asia air
|100? tiger air (128 or 144 on orbitz and ITA)
|33, air asia
|111, ita (134 air asia)
|18, air asia
|130, air asia
|165, air asia
The Tokyo to Singapore ticket for $172 (totaling out to $195 after i selected the option to sit in the quiet/no-babies area) on Asia Air was hard to beat. So Singapore became the next destination. The funny thing was, the Tokyo to Singapore ticket had a 6 hour layover in KL, yet it was cheaper than the Tokyo to KL ticket for the same timed flight. So I bought the Singapore to KL ticket and planned to visit KL. But if 6 hours was not enough and I don't get back into the airport in time, I also bought another KL to Singapore ticket (those 2 tickets together was still cheaper than the Tokyo to KL ticket).
Here are the prices i got the tickets for
- LAX to Osaka (with 8 hour layover in Tokyo), Orbitz, $725
- Tokyo to Kuala Lumpur (7 hr flight, sleep on plane), Asia Airline ($195.62)
- KL to Singapore (1 hr flight), Asia Airline, ($18.80)
- Singapore to Hong Kong (4 hr flight, sleep on plane), Fly Scoot ($117.90)
- Hong Kong to Shanghai (2hr flight, 7pm to 9pm), Spring Airlines ($140.11)
- Shanghai to LAX included in the $725 Orbitz price.