NYT On Yahoo's US Gains & Google's Endless Betas

We’ve had nearly a year of full-blown search wars, but the New York Times notes in
Search Sites Play a Game of Constant Catch-Up that despite new releases and “me too” matching of
products, Google has still increased its share of searches worldwide from 44 percent in November 2003 to 47 percent in November 2004.

So all’s great for Google? Not at all. Yahoo had a gain too, from 25 to 27 percent. And in the US, it was much more dramatic. Google rose from 37 percent to only 38
percent. Yahoo leaped from 29 percent to 35 percent, not that far behind Google’s share.

What’s behind the gains? The article has lots of quotes from me commenting on how Yahoo shows a desire to define a project, deliver on it and move on. In contrast, I remark
how Google delivers something in beta form, then moves on to something else without seeming to finish the job.

Whether this operational style is behind Yahoo’s growth or if there are other factors, I don’t know. Google responds that it’s hard to say it’s dropped the ball on any
major releases. With respect, I beg to differ:

• Google Images — one of only two NON-beta products out there — was allowed to be nearly a year out of
  date with images. Only now is it showing signs of being updated.
• Gmail remains both still in beta and invite only mode nearly a year after its rollout. How about finishing it up and making it public before moving into other things?
• The Orkut social network service? Ditto.
• Google Desktop still doesn’t support Firefox indexing. We’ve had some notable Firefox developers get hired
  by Google recently — perhaps they could take a minute or two to help make what we view in Firefox searchable?
• The Google Toolbar that has been out even longer than Google Desktop still doesn’t support Firefox.
• Google is literally years behind providing query refinement support to searchers. Google Suggest is
  Google catching up on this front with the other major players. It was a “20 percent” free time project that one of its engineers cooked up and still hasn’t come onto the
  actual site. It should have been an 80 percent time priority to happen years ago.
• Google Catalogs? Another beta project rolled out then apparently abandoned. The most current Ikea catalog is from 2003. Just kill
  it, already.
• Google News nearing its third year of beta? Google cofounder Sergey Brin has said before that it will come out of beta when it’s ready. If it’s not ready after three
  years, when will it be?
• Similarly, in the NYT story, Google says it will keep the beta label on things until time can be found to add “important features” to these products. Three years for
  Google News, over two years for the Froogle shopping search engine — if time hasn’t yet been found to put needed important features on these products, then pull them off the
  home page.

The reality is that Google seems to have no distinction between what makes a “beta” product versus a “final” product. Even the story notes that some of the beta products
from Google have been upgraded with new features over time. Any of those times would have been a time to take them out of beta.

How about this roadmap to follow:

• Alpha: You release a product in limited form to a select group of users. Gmail, Orkut are alphas.
• Beta: You release a product to the general public with the expectation that feedback will be taken over a short period of time (two to three months) before a final release
  happens.
• Final: You put the product out on the Google home page or accessible via the More link on the Google home page. Anything out in front of the public in this way is no
  longer a beta. If it is still a beta, then get it off the main site and back into Google Labs.
• Point Release: After the final, this is when you add substantial new features to a product. Froogle today is arguably Froogle 2 or Froogle 2.5, given that it has had
  several major enhancements.

As for point releases, these needn’t be labeled for the general public. Froogle doesn’t have to be called Froogle 2, for example. But it is useful to use the terminology
for those who are commenting on the changes. It lets us say things like “the second release of Froogle” and know there are substantial alterations that have happened.

Google Groups is a classic example of this. Google Groups was a final product. A new Google Groups 2 came out in limited beta. Then it was deemed good enough to replace the
original Google Groups. That happened — but what DIDN’T happen was removing the name beta from what really
was a final product.

Believe me, I love that Google has a fun, creative process — something I did mention as part of my interview for the NYT story but which didn’t make the cut. That’s
probably because I was much more negative about being frustrated by the lack of completion Google has shown.

At this point, Google is well overdue for an operational pause. Don’t roll out anything new until you bring stuff out of beta or declare it dead and no longer supported.
Then please give me a wealth of new, fun, exciting and technologically disruptive things in the way you do so well — as well as a firm timeline as to when those things will
either receive official, final support or get rolled back out from public view.

More more on this topic, also see More On The Endless Betas Of Google and if you’re a Search Engine Watch
member, my
Breaking Out Of Google’s Beta Limbo that charts when major Google services were launched
and how they they were (or still are) in beta.