Inside Google's New-Product Process
For outsiders looking in, Google's (GOOG) flurry of product releases can appear random and a bit confusing. In the past year, for instance, the search kingpin has unleashed everything from a blog search engine and a finance site to an instant messaging program and online spreadsheets. And on June 28, it uncorked an online payment system that is expected to rival eBay's PayPal (see BusinessWeek.com, 06/29/06, "Google's eBay Challenge").
One thing is clear: Google is toiling to create some blockbuster successes beyond its Internet search engine. But, thus far, its track record outside of search has been tepid (see BusinessWeek.com, 7/10/06, "So Much Fanfare, So Few Hits"). Does Google have a case of product attention-deficit-disorder? Or is there a method behind this apparent madness?
Marissa Mayer, Google's 31-year-old product-launch czar (see BusinessWeek.com, 7/10/06, "Managing Google's Idea Factory"). whose official title is vice-president of search products and user experience, recently spoke with BusinessWeek correspondent Ben Elgin about the search giant's efforts to branch into new businesses. Here are some excerpts:
Google has been branching out beyond its core Internet search offering for four years now. How do you feel the company has done?I'll start with a couple of core philosophies. We believe that we should be launching more products than what will ultimately become phenomenally popular. The way you find really successful new innovation is to release five things and hope that one or two of them really take off. I think by that metric we've been doing really, really well. We should be able to put products out there and, without a lot of promotion, a good product will grow. We like to put products out there early, see what users say about them, what additional features they'd like to see, and then build those out.
We'd rather put something out on [Google's beta site] Labs, have it be a little bit low-profile and grow by word of mouth. That gives the team a little bit more time to scale with the requirements. Also, it gives us some very important indications about whether or not this product fills a core need well, how big the market is, and also how strong our product is relative to others.
How do you measure success of a new product?We rely primarily on our own logs. Also on user feedback. Gmail, I would say, has actually been phenomenally popular. It may not be the size of Hotmail and Yahoo! Mail. We actually have artificially restricted the uptake of Gmail through the invitation model. And I think it's pretty easy to imagine that if we removed the invitation model, we would see ten times as much demand. And if you do that calculation, we would be almost as large as Yahoo! (YHOO) or Hotmail. So, we think that Gmail is actually a very good product.
If you take a look at something like Google News, it actually does quite well. We're now offering Google News in more than 40 languages, in 40 countries around the world. On an aggregate page-view number across all of those different demographics and geographic areas, we're really proud of what Google News has been able to achieve. Its growth year over year has been among the strongest of our mature products. We're seeing traffic almost double every year.
I think the core underpinnings of what we're doing make a lot of sense. That said, there certainly are some products that we've released that aren't market leaders and may never [be]. We anticipate that we're going to throw out a lot of products. People won't be able to remember them all, but they will remember the ones that really matter and the ones that have a lot of user potential.
Often, press and analysts will declare Google's new products, such as Google Checkout, a potential category killer. Do people give too much credit to Google to reinvent markets and take out competitors?People in general, myself included, have a tendency to overestimate the short-term and underestimate the long term. When you look at those headlines, it's that kind of mentality going on. PayPal is a really excellent, mature product. And our service, if you actually look at what we are doing, doesn't really take aim at what they do and what their core competencies are. So there is just a misunderstanding of where the product is aimed.
When you look at a product that is that mature, such as [Microsoft (MSFT)] Excel, it's very hard to hold up a nascent product that by definition is being launched very early in its development cycle and expect for it to have these types of killer outcomes. Do we want to be competitive when we enter a space? Certainly. Will it take us some time and possibly years in the case of some of these very mature markets? Quite possibly.
Google's home page is uncluttered, which is one of its draws. Has Google's thinking changed at all on how to expose products without losing its design edge?It has changed a little bit. We're still not ready to make really fundamental changes and blast all of our products on our home page. [But] there are a few key concepts I've been thinking about in terms of how we can change navigation on our site. One is what I would call the “San Angeles” or “Los Diego” strategy. You take large product and merge them together into the biggest possible nucleus. So if you took San Diego and Los Angeles together and merged them into one mega-city, that's even bigger and more memorable than the two cities independently.
When I look at Google News, where I know we have a user base that is very concerned with current events and likes to see multiple viewpoints, that feels like a really good place to integrate in something like Blog Search and/or Finance. So, we're looking at how we can integrate some of those pieces of functionality. It is hard for people to remember more than 5 or 10 products from a particular company. If we can take each of the products we have and make them even larger and more meaningful to people, I think there's a lot of benefit that could be had by both the users, because they don't have to remember quite as much; and also by us, because we see increased traffic.
Will Google advertise some of its niche products?There's been some really interesting things like the Da Vinci Code quest. It happened in April. There we collaborated with Sony Pictures. We developed something that was I think as good as building hype for their movie as it was for building an understanding of Google products.
For people who played the quest, there was a different Google product showcased every single day. We had several million people who logged in and played at least one puzzle. And I think more than 100,000 actually finished all the puzzles. Including [Google co-founder] Sergey [Brin], who though he was disqualified because he was a Google employee, actually clocked in at I think 10,072. He just missed the cutoff.
Is there enough discipline in Google's engineering organization to churn out steady upgrades of so many existing products?[CEO] Eric [Schmidt] and [co-founder] Larry [Page] acknowledged that we really do need to apply a little bit more organization to some of what's happening here at Google. But, I think it's also important to understand the psychology of what happens with engineers. There certainly are some engineers who tire of working on one particular task and want to move on to a new task. But there are a lot of people who get really deeply ingrained in the space they're working on and they want to build a best-of-breed product.
The lead on our news team, Mike Dixon, was an engineer with very little news background. But now he's incredibly well-versed on news sources from all over the world, viewpoints, every current event. He's definitely become an expert news reader. And it's really important to him that the news product be really high quality and best of breed.
You see the same thing when you talk to our maps engineers. They literally study print maps from all over the world. [They study] the coloration, the contrasts between colors. The names of the streets are actually printed inside the physical roads on the maps, as opposed to be printed across them.
Do product managers need to have more power inside Google?No, I don't think so. We bring together a team of people who are really passionate about [a] subject. I think it's interesting: We still don't do very high-definition product specs. If you write a 70-page document that says this is the product you're supposed to build, you actually push the creativity out with process. The engineer who says, you know what, there's a feature here that you forgot that I would really like to add. You don't want to push that creativity out of the product. The consensus-driven approach where the team works together to build a vision around what they're building and still leaves enough room for each member of the team to participate creatively, is really inspiring and yields us some of the best outcomes we've had.