The people who are offering the Zen-like answer ("the big thing that can be predicted is not the real big thing, it does not have true buddha-nature") are being a little too skeptical. Historically, there's been a fairly good record of people seeing "the next big thing" (NBT) a respectable fraction of the time, so it CAN be done, without 20/20 hindsight. Here are some successful big-thing anticipations that I am cherry-picking to illustrate when, why and how you can predict NBTs.

1. The moon race/Apollo was the NBT of its time and everybody knew it. It trained a generation of technologists and spawned a huge ecosystem of offshoots we use to this day.
2. Distributed computing was the NBT of the 70s, and Xerox invested in it for a decade to make it happen.
3. As the man told Dustin Hoffmann in The Graduate: "Plastics." Another NBT that plenty of people saw coming.
4. Container shipping. The entire shipping industry saw it as a grand vision that could change everything, and they just had to wait for the people with the right talents. See my review of the history of the revolution: http://www.ribbonfarm.com/2009/0...

What caused the early warning visibility in each case?

For 1, the science had been done since Newton, the basic feasibility had been demonstrated (V2s, Sputnik), and it was a matter of money. Once Kennedy committed the money, succeed or fail, everybody knew it was going to be the NBT for a decade and dominate technology.

For 2, many recognized that transistors made dirt-cheap Moore's Law computing possible, and as Alan Kay said, there was a lot that could be done by "wasting transistors" instead of viewing computing as a scarce resource. We focus on PARC because that happened to work out, but HP for instance, made major early NBT bets on hand-held calculators for the exact same reasons. They were a little too early in a sense, and many of the things HP worked on—such as hand-held devices and touch screens—are only maturing now, market-wise.

For 3, once nylon had been discovered and proven useful, the race was on given the supply of cheap oil (the main raw material) post WWII from the Arabian Gulf. It took genius to invent the first of the new materials, but once that was done, it didn't take superhuman vision to see that there was huge potential in a class of materials capable of competing with everything from steel to cotton.

For 4, necessity was the mother of invention. The demand was there, the concept had been recognized early on as a solution, and there were no obvious impossibilities. It was mainly regulation that slowed things down. Many saw containerization coming, and placed bets which paid off beautifully. That's how NJ killed NY in the harbor race.

On the Web, streaming TV was always going to be big... too many people had spotted the opportunity in collapsing bandwidth costs for it to not happen. It was a matter of timing and execution. Search was also anticipated (which is why investors DID get in on Google early). Facebook was slightly more surprising when it happened because the sector seemed to have been proven of minor value by Friendster, Orkut etc. Twitter (to me at least) was a complete blindside. I didn't expect anything like it. But I am guessing others did.

So the themes are what you might expect:

• a huge pot of unallocated money earmarked for an area,
• a game-changing basic technology that reopens many vast areas of technology/breaks assumptions underlying existing industries, or
• a problem with obviously huge demand that had no clear impossibility barriers (unlike say "anti-gravity" which nobody has an attack for, despite the obvious demand).

So where might you see similar things today?

• First, follow the money (healthcare, education are two places currently attracting investment).
• Second, look for unexploited recent scientific breakthroughs (this is why nano, genomics are so attractive... there's unexploited basic science there. Feynman said about the former, "there's plenty of room at the bottom"... costs are dropping faster in genomics than they ever did in electronics or video transmission).
• Third, follow contours of demand for which there is no obvious impossibility barrier: an example is the needs of retiring baby boomers. Huge numbers of people gradually exiting the workforce, healthy enough to do stuff a little longer, and hit enough by the recession that they need the money. Something's gotta give for the gray-haired and there's no obvious impossibility that says they're going to all have to starve to death. Also cleantech, since it is by now obvious to everybody that peak oil is approaching and money is starting to pour into alternative energy as well.

There ARE cases where the NBTs aren't predictable, not because predictable NBTs aren't available, but because even BIGGER NBTs appear out of nowhere and render the predicted ones moot. These are black swan NBTs.

One great example is xerography, the process behind photocopying. The idea truly came out of nowhere (dry copying on plain paper, based on an obscure physics phenomenon called photoconductivity) and completely disrupted the typewriter industry, with vast "typing pools" of working women drying up practically overnight, and the office being transformed beyond recognition.

Another blindside NBT was penicillin, which suddenly disrupted healthcare, since everybody at that time believed in the magic bullet theory (one medicine per disease) and didn't know what hit them when it turned out that one drug could fight a lot of diseases in "broad-spectrum" ways.

I have my own opinions about the NBTs of the Web and in technology generally. If you are looking for a single answer like "mobile ecommerce" then you are thinking about this wrong. Technology forecasting is a portfolio game. That's why the investment industry is structured around portfolios.

Whether you are an innovator, researcher, entrepreneur or investor, you need to have a list of potential NBTs with expected value/probability attached. Some hypothesized NBTs will be very sharply defined ("mobile ecommerce") while others will be broad ("something big in education"). And you need enough unallocated risk-taking bandwidth to jump into black-swan NBTs when they appear. And you should be doing some of your own original exploration in obscure areas to look for your own black swans.

You also have to ask yourself what YOU know that nobody else does. Anything broadly considered an NBT is not going to be a bargain to get in on, due to oversubscription (information spreads rapidly in an efficient market...). People mistake unexpectedness for big-ness because the unexpected makes a few peripheral people look like geniuses.

So pick your options, place your bets.