I recently attended a presentation titled "Innovation at Google", by David Bercovich, product marketing manager at Google.
Among the highlights of his presentation, Bercovich shared key lessons learned about product management.
People have a close connection to Google, Bercovich said, because it has opened up information to people and changed the perception people have of sharing information. Before, people who had information had power. Now, in the interconnected world, sharing information has power.
From a marketing perspective, it's interesting to remember that Google didn't do a lot of advertising to attract new users. For most people who use Google's services, a friend said to them: "You've got to try this because it is great." Bercovich notes that as we're more networked, that kind of viral marketing happens more. That's one reason why I believe it is so critical to focus on Net Promoter Score in running a business.
When you think about it, the consumer technology world is innovating because you're one click away from a consumer going somewhere else. For Google, "It's all about innovation, because there are no switching costs in search," Bercovich said.
Here is a summary of Google's product management tenets, as presented by Bercovich:
1) Fast is Better Than Slow.
While there's an old saying "patience is a virtue", users are impatient. Google has found that if they slow search by 1/10 of a second, there are significant decreases in search usage. What does that say about user experience in other parts of the web? People don't have a lot of patience.
There are implications to how to deploy web services. The world sped up, but for many companies how they deploy didn't change.
In the old model: define all the requirements; evaluate build vs. buy; issue an RFI or RFP; select a vendor; conduct a bake off; define an implementation plan; customize the application; build an end user training plan; deploy the application. That takes a long time.
In the new model: get the product out quickly, let early adopters provide feedback and have that shape the product. New product development must be iterative, not a big bang. Launch and improve. Fail quickly and learn.
In the past, safe was better. Now, the speed at which industries are changing makes an important case for speed and taking more risk.
2) Simple is Better than Complex.
Google search today has essentially the same simple experience it did in 1997. Today, all the innovation happens behind the scenes, so the products are easy to use. No user manual is needed.
3) Assume Chaos and Deal with It.
As a basic approach, Bercovich advises: encourage risk taking, don't punish failure. "If you're not making mistakes every month or quarter, you're not taking enough risks," Bercovich said.
In terms of dealing with chaos, Bercovich says that manual categorization and hierarchies are dead. As an example, he has 108,000 emails in his inbox. The old way of putting mail into folders isn't scalable.
"Search is the only scalable option," Bercovich said.
And this example applies to information across a company: rather than structurally manage each piece of content, allow people to search the information they have access privileges to.
Of course, search isn't always the right tool: for example browse and discovery are important for media.
Bercovich notes that there is a difference between not being structured with information categories vs. not being structured in your approach to problem solving.
How does Google approach problem solving? "Google's core philosophy in hiring is hire athletes, not shortstop," Bercovich said. "If you hire a person to solve a particular problem, over time that problem changes."
Organizationally, he advises forming small teams, and moving people around a lot.
"Put a small team on something. Put it on the web. Test it," Bercovich said.
(Photo credit: smanjo)
This Wednesday, Thursday and Friday, I will be a judge for the semi-finals of the GRA/TAG 2008 Business Launch Competition. 
A good article in the Wall Street Journal last week that describes the 
RealNetworks has 
