I love seeing old YC companies continue to try to discover product market fit. I remember this blog post which made me think they were toast because pg told them: “But until you’ve got a major discovery to show, or some serious growth, I can’t recommend you to investors”<p><a href="http://tomhoward.co/part-1-reality-check" rel="nofollow">http://tomhoward.co/part-1-reality-check</a><p><a href="https://news.ycombinator.com/item?id=3941754" rel="nofollow">https://news.ycombinator.com/item?id=3941754</a>
The title is a polite way of putting it. In reality, what they are backing is a startup which claims that the data they gather will allow airlines to gobble all consumer surplus by being able to offer compensation equal to the reservation price of the customer: I.e., the compensation that leaves the passenger indifferent between flying on this flight vs giving up their seat.<p>I am going to claim that the airlines backing this and the investors of the startup have not studied Ultimatum Games[1]:<p><pre><code> One of the most debated results in experimental
economics comes from studies of the so-called
“ultimatum” game. In this game, one player (the
“proposer”) goes first and offers a split of a pie
of given size. The second player chooses whether to
accept or reject. If the second player rejects, both
get noting. If he accepts, they split the pie as was
proposed. The ultimatum game has a unique subgame
perfect equilibrium where the proposer gets
(essentially) the whole pie. It also has many other
Nash equilibria, where the proposer offers a more
generous split, fearing an aggressive offer will be
rejected.
The experimental results are in stark contrast to
the backward induction solution.
</code></pre>
[1]: <a href="https://web.stanford.edu/~jdlevin/Econ%20286/Experimental.pdf" rel="nofollow">https://web.stanford.edu/~jdlevin/Econ%20286/Experimental.pd...</a>
A good start, but I think this could be taken much farther.<p>If we look back 20+ years, we built up very strong rituals around scheduling because we couldn't make the system work any other way. But in-person plans have become much more dynamic as we become better connected. Air travel is still very much in the old world there: we book far in advance, slotting ourselves into fixed-schedule, fixed-capacity options, even though we are often much more flexible.<p>For example, every August I go back for a big camping trip with extended family. I could fly from a few different airports into a few different cities over a few days. Since working remotely for a day or two isn't a problem, the return's even more flexible. Why do I have to buy a specific down-to-the-minute ticket now if I want a good price?<p>I'd be happy to commit now, and then let the airlines (or airlines negotiating automatically with my booking agent) work out exactly when I'm going. They could use much more data to maximize efficiency, giving us lower total prices.
>Volantio’s solution: Gather non-identifiable data about purchasing behaviors to learn what type of consumer is more likely to accept what type of offer.<p>Call me a skeptic as much as you want, but the "non-identifiable data", must <i>somehow</i> end up being identifiable, or at least the passenger will be personally tagged as "type A", "type B", Type C", etc. in order to propose him/her the "best deal".
There's a much simpler solution to the overbooking problem: eliminate the fraud exemption that airlines currently enjoy, allowing them to sell the same seat twice.<p>The downside is that, if airlines are unable to resell unused seats, more flights will fly partially empty, reducing overall efficiency of the industry. But if they offer standby "tickets" at steep discounts (instead of making <i>every</i> ticket a probabilistic standby seat), that problem would solve itself.
I'd often be perfectly happy to take earlier or later flights on the same day, and potentially even change routing (OAK vs. SFO, or even something crazier like MIA vs. FLL). Wouldn't even necessarily need to offer cash; frequent flyer miles or upgrade to C or F or even lounge access while waiting would be fine.<p>Rare other times, I'd probably want >$5-10k to change my flight (missing a speaking slot at a conference or a customer meeting or something when booked close).<p>Being able to surface customer preferences like this seems valuable. I've had Alaska/Virgin call to offer compensation to change (or cancel) flights before, either due to overbooking or downgauging, and I've had international flights change and offer free rebook/cancel as a result, but nothing has really been technically efficient or easy to use yet.
I'm surprised that airlines wouldn't/couldn't build this functionality themselves.<p>Does anyone here have insight into the 'buy vs. build' decision that any of these airlines went through in adopting this company's services?
As far as I can tell overbooking is an almost exclusively US phenomenon. In Europe I don't know of anyone who has been 'bumped' off a flight because it was overbooked. Low cost airlines like Ryanair manage to fill flights without overbooking. There is a lot to dislike about Mr O'Leary's company but I don't think anyone worries that their ticket is not going to be honoured.
Just FYI, if you get asked to swap, there's no guarantee you get the same plane type.<p>I opted to switch in exchange for $300, but the plane was an older 737, down from a 787.
Maybe I'm cynical here but any technology that allows safer overbooking will be used to squeeze more bookings on a plane again increasing the risk of upset passengers and PR disasters. Markets lead to exploring boundaries and the boundary to overbooking is the willingness of the company to suffer the consequences. That willingness is cultural and will not be changed by technology.
The investment by non-US carriers intrigues me, because there is no significant overbooking problem worldwide: its a mostly US phenomenon as I understand it.<p>The rebooking-because-cancelled, thats a huge problem anywhere. So, I suspect the JV partners are looking to benefits for unexpected surge loads and overflow from late/dropped flights in a tight jet schedule.
While this is a step in the right direction it just serves to show how far behind airline companies are in terms of tech.
Swapping out passengers on the fly and handing out virtual miles in exchange is seen as "innovative technology"?<p>Hopefully this will pave the way for technological disruption in the airline business.
There's already a well-proven technique for price discovery: it's called a _market_.<p>But because there's no real market for airplane tickets--the airlines basically forbid secondary transactions--we're forced into a world of centralized planning, and by an entity that's fairly incompetent at that.<p>This entire thing would be solved if plane tickets could be bought and sold in some sort of marketplace. The airlines could even take a cut of every transaction!<p>It's baffling to me that the overbooking problem is solved <i>at the airport</i>, minutes before take-off.<p>Imagine instead a world where you could buy and sell plane tickets.<p>- On purchase, I could set a reserve price for which I'd be willing to sell my plane ticket, and have the price go up as the time of flight got near.<p>- Late to the airport--perhaps I could trade tickets with someone on a later flight that wants to leave sooner?<p>- I have a flexible lifestyle. Standing order for tickets to the beach|family|wherever at a low price, as long as I have 5 hour warning before flight time. Gonna miss your flight? I'll buy it.<p>Let the airlines take a cut of every secondary transaction and let app developers & market makers figure out the use-cases.