I’ve had a web app idea in the back of my mind for the longest time. For reasons I explain, I haven’t tried to do it myself. It has continued to bother me, though, especially in those times when I wish it existed so I could use it myself.
Considering its extreme relevance to Valentine’s day, I’ve decided to write it up and just put it out there.
Problem
Nowadays, people have friends living all over the place. They have their high school friends, their college friends, their friends from when they were an exchange student in Germany, and their friends they met on vacation in Guatemala.
Families, too, are getting more spread out; a sister might live in Seattle, and her brother in New York. Where should they meet for a reunion?
When you want to meet up with a friend or family, you don’t necessarily want to fly straight to their home city. Sometimes you want to meet half-way and have a short vacation.
Of course, people in long-distance relationships especially want to find places to meet up. Friends can FaceTime, but lovers probably want to meet in person, and they might not be especially picky about where they meet. A destination is okay as long as it could be a moderately entertaining city, because the city is less important than the person they’re meeting.
So what do these people do? They find some place that is cheap for both of them to fly to, attempting to minimize the combined flight cost.
Missed Opportunities
To solve this problem, you would probably start out with countries or cities approximately equidistant from both people, doing two flight price searches per city and manually adding the prices. This works okay for a few cities, but you might miss the best options because you never thought to check them.
You see, the topology of flight pricing is not directly related to distance, and it’s sometimes counter-intuitive. It might be cheaper for someone in Florida and someone in Tennessee to meet in Chicago because United has a hub there. It might be cheaper for someone in Malaysia and someone in Hong Kong to meet in the Philippines, because Air Asia is running a limited-time promotion. Certain longer routes can be cheaper than much shorter routes simply because of increased competition on certain stretches of that route. For the adventurous and flexible (destination-wise) traveller, this gives opportunities to explore a new city and save some money.
Currently, the best way to find these things is to just obsessively search as many destinations as you can, or maybe use something like KAYAK’s Explore or Skyscanner’s all flights search for each person’s city, but there has got to be a better way.
Solution
The solution, as I see it, is to leverage the power of the Internet flight search engines. If one of these providers had an open-use API (i.e. allowing programmatic searches from third-party apps), it would be extremely simple to write an app that, given two locations, starts brute force querying all likely cities and tabulating total costs.
Of course, that would be a dumb and extremely search intensive method. The smarter way is to start to build a graph based on past cached search results to try to minimize the number of fresh searches required, performing searches only on especially promising or especially volatile routes.
The final result would be a “Rendezvous Report” with a list of potential meeting cities and prices.
Complications
Data Source
It’s a pretty simple idea, but the difficulty is in the execution. By far the biggest complication is getting the flight pricing data. All this flight pricing data is locked up behind proprietary interfaces and data agreements.
To implement this app you need to do dozens of searches per user. Most of the flight providers won’t be happy with you freeloading on their searches, especially when your primary monetization methods are probably competing with theirs.
KAYAK, my favorite flight search engine, used to have an API, but we know that it was expensive to operate In fact, they shut it down due to “costly misuse.” I haven’t searched recently, but last time I checked most of the other providers (and I think KAYAK is the savviest of any of them) don’t have APIs either, or have only extremely locked down or expensive APIs.
This is why I think the best place to implement this is within one of the existing travel search engines like Kayak, Google, or Skyscanner. They’ve got the expertise and the data, and they’re most able to optimize the searches and caching because they already collect historical flight data from their primary business. They’ve also got the monetization part solved.
Searchability
The second problem is searchability. How do you let people know about this site? This is a problem that a lot of people have, but not one that they necessarily go looking for a solution to.
When confronted with a problem, I usually look on Google and try to find a technological solution, but I don’t expect most people to act like me.
Even if they did, what do they search for? There’s no quick and easy keywords, only abstract explanations of the problem. How do you buy the keywords for a long search query like “finding the cheapest city for two people to fly to.” It’s difficult to Google, with lots of generic “cheap flights” noise, so it would be difficult to help those people to even know you exist.
This is another reason why an existing player would be great. They can put up a splash message on their front page, and (my favorite idea) have a huge valentine’s advertising push: “Looking for somewhere to meet up with your sweetheart??” They might even be able to detect search pattern and offer Microsoft paperclip-style advice: “It looks like you’re trying to find a city to meet, want to try our special search?”
Other than that, you’d probably also want to advertise on places like ex-pat forums, and maybe even online dating sites, though that potentially gets a bit sleazier.
Name
I like “Rendezvo.us” because it’s clever and Web 2.0, but most people don’t know how to spell rendezvous so it’s kind of a non-starter. Other ideas: HalfwayPoint, LetsMeetHalfway.com, MeetYouHalfway.com. These are all mostly bad.
My friend thought “Rendezvous Report” would be a nice name, but I think it’s too long for a domain. As a sub-brand of another provider it might work.
Novelty
I haven’t found anybody else who had quite this idea. There are a lot of sites which will find a convenient location for you to meet using driving directions, but I couldn’t find any that would look for flights or help you choose a city based on pricing. I imagine it is because of the data problem, because the idea isn’t that far out and I expect there are others who have thought of it.
This goes back to the other main problem, though: searchability. If this already exists I can’t find it. If you know of a site like this or another blog post like this please contact me (@divergio), I would love to know about it. If you see any other fatal flaws in this plan, I’d be interested to hear them, as well.
Conclusion
I still think it’s a good idea, and an existing player with just a little bit of engineering time and expertise could easily roll something like this out in time for the next Valentine’s day.
My best bet is Kayak, because at least historically they were pretty savvy and disruptive, and even have a Google-like “Labs” project page. Though recently I haven’t seen many exciting enhancements come out of them besides their well-done iPhone app.
I looked at Google’s flight search and it also looks more developed than I thought. Someone could do it as a 20% project, though in that case I’d be worried about the typical Google ADD-style product abandonment.
I hope someone can do it. Those of us in long-distance relationships or living far away from friends and family could really use a tool to help us meet up with those we care about.
Addenda
A few other non-essential additions:
For bonus points, you could integrate with other transportation networks. If it’s cheaper for me to take a hydro-foil to Macau and meet there than for her to fly direct to Hong Kong, then the app would find that solution. If she can take Megabus to D.C. and save me $100 on a flight, then the system would recommend that possibility. Unfortunately, I’m not aware of any successful provider who has really got multi-modal travel search (with pricing) working well even in the normal case of point A to point B.
Another idea: integrate hotel pricing, and maybe even other cost of living factors. If you’re price conscious enough to be using an app to find a cheap flight, you probably don’t want to go somewhere where all the hotels are $300 a night and meals average $30 a person. Kayak has some of this information.