Ask yourself: Is "let's not rock the boat" and too close an adherence to "wash, rinse, repeat" the mantras that are giving you incremental success when big results are what you should be going after?
Stop doing what does not work. Remember Google Wave that was to unify email, IM and wikis? Google Buzz for microblogging? Google Reader for RSS? Google, to its credit, has always know 'when to hold them and when to fold them'; these are standalone products that they have killed. What is admirable is to have deep pockets and know when to stop pouring money into a bad idea, or when to merge a nascent idea into another product as its more natural home. Not enough companies do that! Duct tape empires abound where bad ideas don't go away -- they just get added to.
Ask yourself: How much good money have you spent trying to salvage bad situations, tools that have become less relevant, integrations that have just not worked?
Put the R back in R&D. Google may be able to think of the world as its canvas and do bold expensive projects and perhaps you cannot. But even if your canvas is smaller, are you doing any really new thinking or work on future products? My sense is that most companies just tweak the products that they have, or pick adjacent ones that are new to them but not to their industry. Aligning with customer input is the self-congratulatory "new deal." But the "R" is missing for most in R&D, or it is at best a small "r." The future is truly different from the present -- and few companies think in the long term, which creating for the future demands. Even if the risk you can afford to take is fractional, you need to take that risk if you are to go beyond "me too" products.
Ask yourself: How many new ideas does your company really explore, how much do you lose to the mantra "just don't have the time to look at..."?
A simple recipe for success? No -- all of the above require sustained commitment, rather than a flash in the pan. But the recipe is there. Open the right doors to let the future in.
Sign up for CIO Asia eNewsletters.