A few months ago I wrote a column giving advice to Larry Page when it was announced that he would be taking-over once again as CEO of Google. Not that Google is especially in trouble, but it is a big job getting 50,000 feet marching in the same direction. In order to make that happen I urged Larry to create startups within Google. And sure enough, as he took over the top job last week and started announcing changes, one of the most radical was something very similar to the “five guys in a rented apartment” scheme I had proposed. Who knows, maybe Larry reads this rag, but probably not.

While I say Google isn’t in trouble, that doesn’t mean the company isn’t stuck. Google is very stuck. Like any successful and mature tech enterprise they are very adept at leveraging market advantages. PageRank, AdWords, AdSense are the big money-makers still. Everything else — everything else — is a page view generator and nothing else. Gmail, all the web apps, YouTube — all they are for is generating page views and displaying ads. There hasn’t been a successful new business at Google for more than a decade (no, Gmail is not a business, nor is Android). So Google is still a fabulously successful enterprise and great at making money, but it is a stuck very successful enterprise.

Google hasn’t shown it is very good at inventing new businesses internally and they aren’t good, either, at buying businesses externally. Name one business Google has purchased and taken to a new level of greatness. They tend to buy companies for the people and then throw away or forget the technology.  Name one CEO or CTO hired by Google with an acquisition who is punching out products today. It can’t be done. Graham Spencer, Rohit Khare, Max Levchin just to mention three: what happened to them?

They disappeared. I’m sure they are plenty busy with this and that, but they are also invisible.

So Larry Page has his work cut out for him and I commend him on his first week on the new job. Streamlining management, making Google’s social business a priority for everyone, coming up with new ways for Googlers to start their own businesses inside the company — that’s all great. But it isn’t enough.

Take that internal startup program, whatever it is being called. In principle it is a great idea, but the implementation is flawed. The internal startup founders, for example, are given two years to make their business work — two years before they have to deliver anything. That’s crazy.

Maybe it takes two years for Google to delivery anything, but Google is now a big stupid company with poor communication skills to boot — in many ways a worse Microsoft than Microsoft. A startup that hits its first deadline at 24 months is a startup that is over-capitalized and too lacking in fear.

Listen, these Google engineers with their startups will have already been thinking about their idea for months or years before they ever submit it to management. They need to deliver a prototype in two months, a solid beta product in six months, and have a full release in 9-12 months, tops. After all, they aren’t spending any time at all looking for money, and that’s what startup CEOs do probably half of their time.

Giving them two years is saying the engineers should take a year on the beach first, then get to work.

And if the startup idea fails then the Googlers working on it should fail, too. They should be fired.  Now there’s a proper incentive to succeed, not this fantasy startup thing.

Maybe Larry has forgotten all this, maybe he’s just slow, but he’s also wrong. This internal startup program may well keep people from leaving Google for awhile, but it won’t generate many new businesses, because it creates the wrong atmosphere — one that actually encourages failure.

But at least it is a step in the right direction.

Now what, Larry?