As a company grows, it becomes easier and easier to rest on past achievements and accolades, and harder and harder to even consider making any dramatic change. This problem has been popularly characterized in the business world as the innovator’s dilemma.

In most cases companies that manage to find a market and a product that gives them satisfactory growth, which is a difficult task in and of itself, do so while riding on a single market vector. Correctly identifying the underlying characteristics of this vector and being able to be brutally honest about the future of where it is headed as well as where neighboring, but different, vectors are trending is critical to ultimately making a decision to fundamentally disrupt yourself.

Many of the people don’t realize AdMob was ever anything other than an iPhone and Android in-app advertising network. They are typically surprised to learn that AdMob was launched in late 2005 and funded in early 2006, well over a year before the introduction of even the first iPhone, and two years before the app store existed. AdMob as a company was predicated on the growth of mobile web (specifically WAP) usage and all of our ads were served in that context. We spent two years aggregating all the publishers and advertisers we could find in the WAP space, and built a nicely growing business with a healthy customer base.

While we were doing this we were frequently asked to build out the capability to show ads inside of mobile applications, which at the time were built on J2ME and Brew, as well as Symbian. In every case where we ran small tests, we learned conclusively that ads in apps simply did not work as well as they did on the mobile web. App usage was rare, and connectivity and latency issues led to a whole host of technical problems when it came to the last mile of actually displaying ads inside of mobile applications. So we learned when it came to this question: mobile web good, mobile applications bad.

On June 29th, 2007 Steve Jobs unveiled the first iPhone and everything about our business changed (though we did not realize it immediately). We were culturally built to be agnostic to any new mobile device, so we quickly built a web unit for the iPhone and continued about our busy lives. Then in March of 2008 the App Store, and the notion of building apps with an approved Apple SDK was announced, and we faced an interesting question.

Everything we knew about mobile advertising told us that apps were bad, and the web was good, but there was something fundamentally different about this device and ecosystem. We were not completely sure the results would be different, but the capabilities of the iPhone and the apps that could theoretically be built on top of it convinced us to give in-app advertising one last try. We immediately devoted a team to building a variety of ad units and SDKs for the iPhone, mainly because we suspected, but were not sure, that what we were seeing was not an extension of our vector, but an entirely new one. The app store was released to the public on July 10th 2008; we announced our in-iPhone ad units two weeks later on July 24th. If memory serves me correctly, the next competitor would not enter the space for at least six months. We had ad units and SDKs for the web as well as for applications. Although we suspected applications would behave differently on the iPhone than other platforms, we were far from sure, and certainly expected the web ads to outperform.

At the time we were serving 34 million impressions a month via the Safari browser on the iPhone, and although that seems to be a large number, it accounted for less than 1 percent of our total inventory. Internally however, even on this small percentage, the growth rates and monetization rates of this inventory were so profound that we began to believe more fully that this new vector was a reality rather than simply a theory. We debated and discussed, but it became clearer and clearer that our world now looked something like this:

As the data every day continued to pound the message home, we became convinced that we were going to have to commit the full energy of the company to this new vector. It was a very hard decision, as we had a number of competitors in the WAP space and we were abandoning markets and battles that were hard fought and still very much active. It was clear to me, however, that while we could win on the WAP vector, it was no longer a fight worth winning.

When I announced the decision to the company, I remember acknowledging the markets and revenue we would be abandoning, and alluding to the plot of the movie Apocolypto. I said that all of what we had been doing thus far amounted to small scale conflict with our neighboring tribes. I told the team that we were the first to see the tips of the sails of the Spanish galleons and all of our energy must, by necessity, shift to preparing for this entirely new world that was about to become our reality. Other than starting the company, it was the single most important decision we made, and led to significant growth and success in the following years.

The key to knowing when to disrupt yourself is keeping a keen eye on the market around you and not getting so caught up in the hand to hand combat that you miss the signs of a dramatic sea change in your world. Business is a battle, and battles cannot be won without detailed tactical plans and knowledge of all the daily minutia. But it’s equally important to remember this: you, and all your adversaries, will all be gone if you ignore the Spanish galleons when you see them on the horizon.