Innovation at Google

February 10, 2009 · Posted in innovation · Comment 

Google is great example for people interested in innovation because they do many things right as well as wrong. They create a lot of great innovations but mostly through brute force and a lot of luck.

By examining Google I hope to show you how to achieve the greatest possible innovation success. If you haven’t read my Blog then you might be asking, “Who are you and why should I listen to what you have to say?” Fair enough.

For over 25 years I’ve help create and bring to market dozens of innovations worth billions of dollars. As result of my experience I discovered a system to reliably create targeted high value innovations on demand using resources you already have or can easily acquire.

I have a special connection to Google. Larry Page, one of the founders of Google attended U of M. He was an undergrad when I was there preparing for a PhD in a new area of Information Theory. He used some of the exact techniques I helped develop when he started Google.

In 1992, before there was such a thing as the web, I was at Apple writing white papers describing what the web would look like and how to build it. Some of that work led to the software development methodology called Agile. Google is an avid user of Agile.

In 1994, years before the founders of Google even met, I was implementing the web at Hewlett Packard. We already knew that search was a major concern. We solved the problem internally at HP by using the Topic Real Time engine, the same technology the CIA used to scan and categorize millions of documents per second. One of the insights we implemented was sharing search criteria so people could benefit from the human contribution to search. This is the exact same innovation, although implemented differently, that launched Google. So that should answer a little bit about my understanding of the web, search and why what I have to say about innovation will help you achieve the greatest success.

In the presentation, Douglas Merrill hits an early innovation home run when he says, “Google tries to answer the question you meant”. That is exactly the point of innovation. Give you what you want. The search term “Apples” plural returns pages about the fruit. The singular term “Apple” mostly returns pages about the computer company. This is a great understanding of the desired Outcome.
The customer doesn’t care how to do it, they just want the result they desire. Inventions are about how, Innovations are about what customers’ want.

The best definition of innovation is:

“Satisfy customer’s unmet desires.”

When you get the definition of innovation correct you suddenly see where a lot of confusion comes from and how to fix it.

Douglas says there are three types of innovations:

  • Incremental
  • Incremental with side effects
  • and Transformational.

That is non-sense. Customers don’t care about those distinctions and neither should you. All customers care about is how well their desires are met. Period.

As an innovator all you care about is how many customers you can satisfy for the least cost and effort. Sometimes the so called “transformational” innovations are the cheapest and easiest way for you to get the greatest return on investments. Other times the small improvements are the way to go. That is more of a strategy decision than anything fundamental about innovation. But for you to make those strategy decisions you need to know all the possible innovations for your industry. That can only be done with the Predictive Innovation Method. You can find out more about that on my blog or at

Another quote from the presentation is, “Innovation for it’s own sake isn’t useful.” The heart of that statement is absolutely correct. If a change doesn’t satisfy customers’ unmet desires then its not innovation. It might be an invention, but its not innovation. If it fails to satisfy an unmet desire it doesn’t innovate. Ideas that could be innovations still can fail because of poor implementation. So innovation is more than just ideas. Innovation is thinking plus doing. In fact there are 7 essential elements to a complete innovation system and idea formation isn’t even the 1st step.

Now here is the biggest mistake most people make about innovation.
“Users don’t know what they want, but they know what problems they have.” Wrong Wrong WRONG! Users know exactly what they want. What they don’t know is how to get it. If a user expresses a problem they tell you “I want this, you give me that, and its not what I want.”
In the broadest sense customers desires are

“What I want, where I want, when I want, the way I want, with whom I want, for the price I want with no hassle.”

All innovations progress towards that ideal vision. Each new innovation gets closer to that ideal for one or more aspects of the product.

He said customers didn’t know they wanted some of the features that Gmail offered. Not true. Gmail did not do anything that people haven’t wanted for their written communication since the beginning of time. It did some of the things better but it didn’t do anything new. In fact the specifications for Gmail could have been written thousands of years ago.

Whether its e-mail or letters written on parchment the essential aspects of reading written communication have not changed and never will. The desired Outcomes are the same for reading all written communication. You want to find what you are seeking when you are seeking it and process the important information in a way that is useful for the task you are doing. The only thing that changes is how well that is done and the methods for doing it. Any product that does not at least partially achieve all essential Outcomes will fail to satisfy customers.

The simple formula for what will be an innovation is “it achieves each of the essential outcomes for the task the customer is trying to do and achieves one outcome or more better than what is on the market.”

Too much bureaucracy kills innovation. Companies rarely make money from doing something new. They make money from doing the same thing many times. So the types of controls for a manufacturing or sales organization won’t work for innovation. But there can and should be effective controls.

Google got lucky to hit it big before the Dot Com bust when money was easy to get. They can afford to “wait as long as they can to kill innovations”. That method just won’t work for most businesses. You don’t have the resources to fiddle with ideas until they fizzle out. Plus, you might not have the time to experiment before competitors beat you to market. When he says, “wait as long as you can to kill innovations”, he is really saying they don’t have a reliable method for recognizing innovations early on.

One thing that is always true about innovation is it will be different from what has come before. That means the entrenched experts will hate really good innovations. But, customers will love it. What you need is a way to let real customers tell you as quickly as possible if you are doing the right thing. There are two parts to doing this.

First, start with an outcome based innovation method, such as Predictive Innovation. The outcomes never change so if the idea satisfies all the outcomes then its good. If it doesn’t satisfy all the outcomes kill it right away. Second you need to directly ask customers “does it satisfy the outcome better than what is already available”, “Do you care about satisfying this outcome better at this time”, and “are we satisfying it properly?” Only customers can answer those questions. Ask the customers as soon as possible and as frequently as possible.

The spelling correction and search suggestion feature of Google search engine came from watching what customers actually do and like. Regardless of what experts say the only real expert is the customer. They might not be able to tell you what they want but if you watch them you will see what they want. Find it out and deliver it.

Three suggestions he makes really boil down to the same concept. He says, “Allow people to work on things they are passionate about.” All innovations are projects. They have a beginning and an end. Passion is very important for pursuing anything through to completion. Passion is especially important for things that are new. So this rule is very meaningful to moving from the idea to the delivered innovation.

Passion can’t be forced on someone, it only thrives when people have freedom. Give employees freedom to follow their passions. Give customers freedom to walk away and also freedom to personalize what you offer them. Their improvements will often point you to the next big innovation.

Use your products. He presents this as a rule but its not always appropriate. Unless you actually benefit from using your product your input won’t be meaningful. Of course if you could benefit from your product and you don’t use it then there is a serious problem. But if you really aren’t the target market of your product you should be intimately involved with people who are and do use your product. Talk to them, watch what they do and don’t do. Make sure you understand what they really want to achieve, then figure out how to help them do it.

And he finished up by saying that leadership is essential to creating diverse teams and that is important to innovation. Different talents, skills and points of view can be very helpful to innovation. In the real world most companies have to use the resources they already have. Fortunately, if you use the right system you can achieve great innovation success. In the end making the decisions that create success in the truest sign of leadership.

Predicting the Future

October 7, 2006 · Posted in innovation · 2 Comments 

One of the most difficult aspects of the current world is the speed that things are changing. And to top it off the speed of change is increasing. It’s impossible to survive by merely reacting. Simply being faster isn’t enough. If you’re working on tomorrow’s product and you make a mistake you have missed your chance and you go to the back of the line.

You have to be working two steps ahead.

The only way to stay two steps ahead is to know in advance what customers expect. You need to predict the future.

When I worked at Apple we designed something called n-tier distributed client server computing. That’s a very technical term for breaking a task up among any number of levels of computers. That improves performance, reduces cost and increases functionality. The concept wasn’t completely new but no one had done it very well. The reason we wanted to do it was to solve the two out of three problem.

You’ve probably heard the two out of three problem, “you can have it fast, good or cheap, pick two.” Its most often said as a sarcastic joke to an unreasonable request. The two out of three problem is an impossible problem to solve, unless you do something deceptively simple but I’ll leave solving impossible problems for a future post.

What my team at Apple realized is if we made such a system it would be made in a very different way. So we not only designed a system, we created methods for designing those types of systems. As it turned out, only a few years later n-tier distributed client server computing was popularized by someone else and is now called the web. Because Apple had seen ahead to develop methods for building web systems they were able to keep up and eventually re-gain the lead again with products like iPod and iTunes.

So staying two steps ahead requires predicting the future. What we want to predict is what customers will buy in the near future. It’s essential to predict the near future and not something way off in the science fiction future. If you are too far ahead of the market you will waste resources on marketing. Success comes from giving people what they want when they want it.

I’m now working with OutCompete. They created a system for figuring this out. We call what customers will want in the future their Emerging Expectations. How do you predict customers Emerging Expectations?

First understand what drives customers desires. One basic truth of humans is we always want more. We want better, easier, for less cost and with less risk than yesterday.

As it turns out we follow a predictable cyclical pattern starting with new features moving toward lowest price then starting again with new features.

So look at any product and you will find it’s at one of those stages. Lets look at portable music players. You might not remember but transistor radios were the first portable electronic music player. The next improvement was adding the ability to choose your own music. The tape player did that. At first tape players were big but eventually Sony made a pocket sized tape player. Next came digital music on CDs. At first CDs weren’t portable but soon Sony made portable CD players. The first ones skipped a lot but improvements were made and they became more reliable and smaller and then cheaper.

Now there is the Apple iPod. The iPod improved on portable CD players by allowing you to build your own lists of songs to play along with a store to buy individual songs. When iPods first came out they only played music, Apple made incremental improvements offering smaller ones and eventually competitors came in with cheaper products. Now Apple moved to the next level with the Video iPod, introducing a new feature and is moving steadily through the cycle.

Looking back each step in the progression was very logical. Once you become accustomed to analyzing the more, better, cheaper, with less risk cycle future progressions are just as logical.

By knowing what customers want today and what they will want in the future its possible to sketch out a map of each possible step. With the map of all the possible steps you can always stay two steps ahead.

In future posts I’ll talk about: