It barely existed about three years ago. Now, many fliers consider inflight WiFi a necessity. And with Gogo, the leading provider of inflight WiFi in the United States, testing out new, more expensive pricing plans, some on the Internet are up in arms.
Economic theory would tell you that Gogo is sometimes too cheap.
Airplane cabins are a great place to study economics. You have people from many walks of life traveling for different purposes. On many airlines, you can literally see the line dividing the haves from the have-nots.
To understand Gogo’s economics challenge, you first have to know that there is limited bandwidth to each aircraft. With the most widely deployed technology today, there is peak 3.1 Mbps of bandwidth to each plane. That relatively tiny sliver of bandwidth is shared among all of the passengers. A single Virgin America Airbus can hold 149 passengers. (For comparison, I regularly get 15Mbps-25Mbps on my home broadband, just for me.)
According to Virgin America spokeswoman Abby Lunardini, “The average uptake across all of our SFO-JFK and SFO-BOS flights (this includes red-eyes!) is now 26%.” If you assume an 80% load factor, that means on the average longhaul flight, 30 people are competing for Gogo’s bandwidth.
Raising prices is just a way of better matching supply with demand.
“As many Gogo users know, bandwidth to the plane is not unlimited and it is a shared resource,” said Steve Nolan, a Gogo spokesman. “On routes where use is heavier, we are testing prices to make sure that we are reflective of demand and that those users who choose to connect have a consistently satisfactory experience.”
What would happen if Gogo priced WiFi at $1 for a 6 hour flight? Everyone with a laptop, WiFi-capable phone or iPad would want to use it. The service would be so slow it would be unusable for everyone. If Gogo priced it at $500 for the same 6 hour flight, almost no one would buy it. Though there are scenarios where they might: part of the Facebook IPO debacle might have been avoided if NASDAQ’s CEO had flown a WiFi-equipped plane back to New York on the day of the IPO. I bet he would have had a very high willingness to pay that day. (He was on a United flight from SFO to Newark that didn’t have WiFi.) I frequently do live television interviews. Television networks would be willing to pay a premium to ensure that a live interview from an aircraft was smooth.
Any given aircraft has people who are trying to get serious work done and others who are just looking to kill time.
Consider these more common scenarios:
- A consultant who charges $800 an hour is traveling to New York and needs to keep up on client work. Most of the usage is email and IM. He downloads contracts and specs. He reviews changes and uploads the finished specs.
- A whiny tech blogger wants to stay connected with the news while he travels. He reviews various tech blogs. He watches clips of Bloomberg West. He checks out slideshows of new gadgets. He files a story.
- A leisure traveler is going to visit friends in New York. He is just looking to kill some time and watches The Daily Show on Hulu. Five and half hours to go, he starts watching full episodes of Breaking Bad.
Each of those travelers has a very different willingness to pay. It can make economic sense for the consultant to pay $50 an hour for inflight WiFi, because it’s still a net win. Relative to the return, it’s a modest price to pay to turn otherwise unproductive time into work time. The whiny tech blogger might be willing to pay $5 an hour for WiFi. And the leisure traveler is probably willing to pay $1-$2 an hour.
But — and it’s a big but when it comes to scarce bandwidth scenarios — the scenarios with the lowest willingness to pay use the most data! The leisure traveler watching videos might consume more bandwidth in 30 minutes than the consultant does the entire flight. (For the record, all three of those people could be me, depending on the reason for my trip.)
In a scenario where you have finite resources, it makes the most sense to sell the product to people willing to pay the most while consuming the least of the product. It makes perfect sense for Gogo to price its product above what the leisure traveler who wants to watch video will pay. The only reason not to price above the whiny tech blogger is that he’ll whine about it on Twitter, blogs and elsewhere. (Idea for Gogo: klout-score based pricing!)
One of the challenges I’ve noticed when I’ve used WiFi in flight is that it can really bog down. When I’m trying to work, it can be really frustrating. In a survey I conducted earlier this year, one respondent wrote, “Latency, poor connectivity and bandwidth often lead to frustration.” By pricing WiFi high enough that people just looking to just kill time drop out of the market, you increase the quality of the service for people who are willing to pay more. Raising prices might actually increase demand from people with a higher willingness to pay because they can count on it more than they do today.
It also makes economic sense to sell Gogo in smaller chunks. Instead of $20 for an entire flight, some people might want to use it just for an hour in the middle of the flight to check email. In the past, I’ve seen Gogo offer increments of 15 minutes or less.
Because the variable cost to Gogo is negligible, it would also make economic sense for Gogo to have pricing that varies by flight. A flight full of leisure travelers from San Francisco to Orlando has, on average, a much lower willingness to pay than a flight with eight bankers from San Francisco to New York.
It would never fly, but the economically optimal solution would be for Gogo to conduct a Dutch auction on each flight. In the first 15 minutes after the plane reaches 10,000 feet, passengers would bid on one of 15 slots for Internet access. The lowest of the top 15 bids would be the price for all of those passengers; everyone else would get no Internet access. Because they would be ensured of better performance than they do today, there would be an incentive for people who value the service the most to pay more.
I suggested the idea to Nolan, but he said it’s not on the table: “While we are testing several pricing scenarios to achieve the right balance between affordability and delivering a satisfying customer experience, we are not currently considering a ‘Dutch’ auction.”
Given the current political climate, I’m sure this post will trigger some blowback about income inequality, fairness, etc. Sorry, but the Declaration of Independence didn’t talk about “life, liberty and inflight Internet.” It is incredibly costly to deploy the infrastructure that Gogo has put in place for its service. It’s only reasonable for the company to charge market prices.
Gogo is trying to offer solutions that cater to the leisure traveler looking to fill time. On some American Airlines and Delta flights, it offers access to a selection of video content such as TV shows and movies. Because these shows are cached on the aircraft, they don’t need to use the scarce 3Mbps link to the ground. Gogo typically charges 99 cents for TV shows and $3.99 for movies. That’s a much better use of scarce resources than allowing everyone to stream whatever show they want from the ground.
There is additional hope for leisure travelers and whiny bloggers: Gogo is implementing a revision to its technology that will increase peak bandwidth to 9.8 Mbps. If supply outstrips demand, prices should fall.
For Internet and mobile advocates, Gogo is a microcosm of the spectrum crisis we will face as more people try to connect to the Internet wirelessly. To achieve true mobility, we’re going to need the FCC to set aside more spectrum for mobile.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.