Pattern Recognition: Google growth drivers; BigCo Innovation; Atwitter about Twitter

My goal is to write one 'Pattern Recognition' a week. Just the top 3-4 stories that stayed under my skin. Here's what stuck this week:

1. Why can't big companies innovate? Ryan McCarthy at Reuters looks at the woes of Nokia, Yahoo and others, and theorizes why big companies can't seem to build new products and services that create new revenue growth. I have blogged on this topic in the past HERE, and would add that one key variable for big companies is that they are aligned in silo'd business units, whereas most game-changing innovations come from holistically, deeply integrated approaches. In fact, if you look at the past 30 years of innovation, you see products that BEGAN life as one concerted effort, but over time were broken into piece part efforts, overseen by different teams in different divisions. Hence, the effort becomes more about coordination and avoidance of internal co-option than break the bank re-think. As to R&D, the seminal truth there is that specifically because most research efforts are unencumbered by the forced disciple of shipping and monetizing, you end up with lots of "concept cars" and no Mustangs. Case in point, look at Microsoft's myriad of 'test the water' endeavors with tablet and phone prototypes, and contrast that with Apple's approach that it 'doesn't exist until it ships.' Both companies are exercising lots of research, and building lots of prototypes in the effort, but only one has a binary, all-or-none measurable. Is it any surprise that only one of those companies generates a material portion of its revenues from products that didn't exist 3 years prior?
2. Three takeaways from Google's earnings: Robert Hof of Forbes live-blogged Google's earnings call this week, and captured the following focal points for assessing Google. One, Google has serious religion about becoming an integrated advertising machine. By that, they want to connnect the dots between ad types (e.g., branding campaigns vs. direct response) and target environments (e.g., desktop vs. mobile). Part of this is that the desktop is relatively weak, with shrinking CPCs and less effective ad units, whereas mobile CPCs are like 1999 in terms of search. Two is that the company is making a push to improve the quality of ads served, noting  72 quality improvements in ads this quarter, including, better geographical targeting, better phrase matching, etc. Finally, the comapny is focused on chasing the enterprise, seeing it as a future growth engine. Underlying many of these efforts is Google’s new Knowledge Graph, which has 500 million things, people, and places, including 3.5 billion facts about them, which is consistent with making Google contextually aware of what you are looking for, and serving it up before you've even completed the thought. While none of these things make you jump up and say, "YES," they are the right types of problems for the company to be focusing on, especially given how dominant they are in the PC universe, and how rich mobile is becoming. 
3. What Twitter could have been: In lamenting Twitter's de-evolution from its tremendous promise as a real-time cloud API company to instead, being a tightly managed proprietary service (something I wrote about HERE), Dalton Caldwell grabbed the lectern, and proposed morphing his App.net service into a fee-based competitor to the Twitter APIs. This prompted noted VC Fred Wilson, an early Twitter backer, to write a piece defending free/ad-support models like Twitter. Of course, Caldwell promptly rebutted. As my experience is that very few in the tech business understand just how hard it is to execute a platform strategy, especially one that developers can build a business around, this is one thread worth wrapping your head around.