As I look at this situation from the outside, two key problems emerge:
- They did not have the right market research
- Their investment was way, way too high
For #2, perhaps this is just a matter of timing. They stated that it would take hundreds of thousands of dollars per plane. However, if a) they built the connectivity hardware into the plan at production, or b) got volume (any reason this couldn't also be done in cars?), then their costs would likely reduce. Seems like they should have just waited until the right cost point.
But, for #1, this just baffles me. How could Boeing not know with a pretty good idea of accurancy what the uptake would be? This is not a revolutionary product or even idea. It's a small, value-add extension of what people already do! Clearly they hosed-up the market research, or, worse yet, sought out the market research that gave them
How would I have performed the market research?
Simple, at low cost.
1. Setup a server with wifi inside some plane. Powered, of course. I would do a) cross-country US and (not red-eyes) b) international routes.
2. Print up some cards announcing the internet service. Give instructions on how to connect, price point, and that customers will need to pay by credit card.
3. When people connect to Wifi,display a page asking them to pay via credit card.
4. After credit card info, display a page apologizing for the inconvenice but service isn't working. Their CC will not be charged, and please enter your email address for a $25 Amazon Gift Certificate.
There ya go. Happy customer, real data (including price points), easy to do, and a marketing database started for when you want to pilot the real thing.
The key is to test actual consumer behavior at low investment. So much better than surveys.
And, if you wanted to do this cheaper, but with less accuracy but still close enough ballpark, just do this:
1. Count the number of people on planes with laptops.
2. Count the number of people on planes with blackberries (they love and value connectivity, after all)
3. Find out the adoption rate at airports for Wifi access, especially international terminals.
Do some back of the napkin math. There's your best case uptake. Doesn't account for cost, which is what I always think is important in surveys.
Have I mentioned I hate surveys?
We did an innovation project on a novel, new-to-market idea while I was at Whirlpool. Tested the idea in the market in a small pilot. Did the post-pilot survey. Over half the respondents said they loved the product and would pay for it. Most said they used it every day.
The problem? When we looked at the usage data, nobody used it.
Surveys indeed.
Now THAT'S insight into consumer behavior. How consumers respond to surveys, that is.
1 comment:
Hi James:
I have similar thoughts about this, but a couple of comments on the post.
There is no doubt in my mind that there is a need. There is also no doubt that there was sufficient latent demand to generate enough usage to make this service profitable. Common sense says that if you look around the plane cabin and see 70% of passengers with notebooks open, there would be heavy usage of connectivity if it was free. The question is, what pricing and what business model is required to get somewhere in between free, and the outrageously high pricing that Boeing offered the service at.
Some problems Boeing had:
1. Market research was a farce. it was done with either a pre-conceived idea of what the answers should be, or to act as filler for the press releases. It was not a well-constructed (statistically valid) test instrument.
2. Basing this service on satellite technology, the cost structure was going to be high.
3. Any realistic market projections would have shown greater than 30% of passengers needed to be using this for the service to make money, yet Boeing made the false assumption that they could get high penetration without a) fixing barriers to usage and b) using a price skimming strateg (which assumes inelastic demand).
4. Waiting until costs came down was not a possibility. most of the cost was the result of leasing bandwidth on the satellites + staffing. It cost $500K to retrofit an existing plane, which included ripping apart walls and seats to install cabling, installing a server and an antenna to pick up the satellite signal on the exterior of the plan. I believe that the bulk of this cost was in labor, which would increase over time, offsetting any minor decreases in equipment costs. If they waited, there was a high probability that land-to-air service which is now being established in the US would suck up enough of the available market to make the cost of providing this only for overseas routes cost-prohibitive.
The problem with doing a small test as you suggest was that the up front costs of leasing satellite bandwidth and developing the technology were so high, that they were forced to make an "All In" bet. But unfortunately, they were not in control of which airlines and which planes got outfitted, so they could not guarantee access to the necessary customers.
I believe that before McNerney came on the scene, the engineers at Boeing were enamored of the technology, and because they thought it was cool, they started believing their own propaganda. They did not consider the very rational conclusion that at a high price, and without access to power in most seats, they had no chance of reaching the required penetration rate to break even. Nor did they treat it as an emergency when no US carriers signed up for the service, depriving them of access to the bulk of the available market.
I attribute these failings largely to Boeing's view that the airlines were their customer, instead of the reality that the passengers were their customers. You can never build a good marketing plan when you don't communicate with real customers, observe their behavior, and use their feedback to improve your products and services.
You can see a detailed financial and market analysis of the mistakes and poor assumptions made by Boeing from the beginning at my blog:
http://thewaythingsare.typepad.com/antimarketer/2006/08/boeing_booboo.html
Post a Comment