A classical micromodel in Game Theory is that of the lemonade stands on a beach. Let's say that three people are asked one by one to decide where on a beach several kilometers long they would place a lemonade stand to maximize their profits. The first player would place his stand in the center, the place most accessible to all corners of the beach. The second player, however would be best off putting his stand right next to the first player's; that way anyone coming from one side would come across stand 2 first and drink there. The third player would be best to put his right next to the other two as well, to attract all the costumers from the other side first.
The end result of course is three redundant lemonade stands standing awkwardly all together in the middle of the beach with beach patrons on the far sides having to walk a pesky distance to get a drink. Now maybe if all three stands were not in competition and were all owned by one lemonade tycoon, he might equally space them out for the ease of the lemonade-drinkers, but it does illustrate something interesting about the choices competitive firms make.
Of course things get even more hazy when there are more factors involved. One thing that anyone who has experimented with different internet browsers has realized is that all of the competitors are sure to develop the same kind of interface and ability changes that others do. All browsers essentially are the same program and all have moved to being closer over time; they all have integrated the same abilities; bookmarks, history viewing, searchable url bars, private "porno"mode, autofill, visual bookmarks on the homepage, tabs, add-ons and the same possible interface changes. Whenever one browser adopts an arbitrary interface change, all others flock to it as well. This isn't just with game stopping programing changes, but with little things. The reasoning is similar to the lemonade stand; their best option is to mimic the choice of others and only be marginally different.
This is the best way to conserve one's lead status. So long as your actions are not too different from your competitors, there's not reason that they should abandon you in favor of them.
Of course the incentives for a developing firm are a bit different. Their only chance of being successful is by being different and praying it catches on. This is something inherently risky, which is why speculative business ventures are so speculative. Google distinguished itself by tabulating search results by metadata and by simplifying its user-interface, which was quite different from the clumsy and decorated search engines of the day. As it happened, this caught on as an innovation and once it was realized, other search engines began to imitate the new Google standard. It was a highly speculative venture for Google, but they had no chance of being kicked off the beach as they had no previous position.
Of course once Google gained their position in their online services, they tried to maintain it by imitating competitors in everything they compete in; this way they at least wouldn't totally lose out and the integration of Google services would prove integral for their appeal. Google's developments came only to mimic other services with the attempt of providing them only marginally more efficiently and with the familiar and loveable Google brandname decorating them. Google Maps, Google+, Google Translate, Google Talk, and most of the newest additions to the Google toolbox are examples of this, no to mention the services that were simply bought and adopted into the Google family: YouTube, Blogger, etc.
Even the interface of the Google Search capability has become similar to the old UI of their former competitors. Google is no longer an economy search engine; to load a search page, one must have their Google account data loaded and checked, their history logged in Google, picture previews for the search results are displayed by default and two new toolbars clutter the view with a new extra information bar on the right side. Autofill and prediction abilities, also default, greatly slow search time and make the old search+enter=results Google roll in its grave. This is simply a result of Google trying to master every conformist edge in attempt to hold its position. Despite hating these changes, I have no way of leaving Google for another engine as they all are just as bad. (Update: I now use DuckDuckGo and have never looked back.)
Now I couldn't be totally correct to say that Google has lost its innovative streak due to its newfound position as the leading internet service company; it flatly wouldn't be right to say that of an organization planning driverless cars and a space elevator, but the services of any innovator rationally become more conventional as they have an established position. This is why it's not uncommon to see automobile or petroleum companies buying up patents that might too quickly innovate the efficiency of cars for their own taste; it may be in their long-term benefit to develop more fuel-efficient cars, but if another business were to find a way to do the innovations themselves, it would cause rapid shifts in the market that would damage the status quo and thus their position. Thus we have a situation in which car companies are sitting on piles of various patents for more efficient engines without their immediate costly utilization.