Give me a lever and the place to stand, and I shall move the earth.
See Wikiquote - Archimedes for the Greek and the actual phrase.
In trying to plan for the future, to perform the strategic planning that is a core activity of Enterprise Architecture, you must stand in the right place.
Specifically, you must stand in the future, in the envisioned target state, and look back, to determine the plan.
If you stand in the mire of the current state, and you try to look forward, your boots will never get unstuck.
Hindsight, as they say, is 20/20.
If you stand in the present, and look forward, it's like looking into mirrored one-way glass. Looking forward from the present, you see only the current state reflected back at you. Start from your current set of running projects, and all the many many details of your organisation's resources, and tasks and operations, and you will never stride out of that minutiae to reach the future.
But when you stand instead in the future, and look back, you can see right through the one-way glass, you can see all the steps you needed to take to get there. You have to, therefore, plan by first creating the future, envisioning it in target statements, mockups, use cases, whatever is a meaningful description to you of what your organisation does, in 2012, or 2015, or whenever.
Then look back... in 2015, we're providing semantic search for chemistry, so how did we get there, well we must have created an index of chemistry terms out of the literature, and we must have gotten some semantic search tools to extract and unify this terminology, and a nice user interface so chemists could enter queries... so that means we must have been holding full-text articles and had the rights to index them... and we must have developed a basic search capability... and on back in time until you reach 2008.
This can be done.
The alternative is to stand around in 2008, and say "we're doing a bunch of stuff, and then... um technology changes somehow and... then the entire competitive environment shifts and... then we do a bunch more stuff... and then err, a miracle occurs and then we have semantic search in 2015". You can't get there from here. You can't get to the future by speculating forward from the present.
Don't predict the future, create it.*
* Alan Kay said "The best way to predict the future is to invent it."