“Out-Googling” Google

A recent article in the New York Times discusses a bunch of Google competitors.

Is this me or does this sound so 2002?

I wonder if “out-googling” Google is such a great idea in the first place. If barrier to entry is truly as low as the article claims, then why do it at all? It seems that anything you carve isn’t sustainable.

I’m no business genius, but I prefer the Prego’s Approach.

[Getting from spaghetti sauce to internet search by way of some dancing bunnies after the jump.]

A detour with some dancing bunnies

One semi-infamous part of my latest talk1 was this quote:

“Don’t worry about dancing bunnies and shit like that, because you’ll get your ass handed to you by people that are good at dancing bunnies…”

I didn’t actually mean for it to come out that way, but it sure sounds like me!

I was trying to say that there was no point in trying to do eCards the way Blue Mountain, American Greetings, and Hallmark do eCards—they’ll crush you and, more to the point, you can’t make money that way.

My eCards design philosophy comes from the Prego’s approach.

The Prego Approach

Read this article.

A typical New Yorker article goes from point to point and never says anything. Or, maybe that’s a typical Malcolm Gladwell piece. Hmm, in any case, the point is you have to draw your own points from these articles, because you feel that you learned something even if you can’t point out what that something is.

In my case, it was marveling at how Campbell’s found a niche from Ragú and broke open the entire spaghetti sauce market. Maybe that’s because I grew up with Ragú’s “it’s in there” and I remember when I switched to Prego (I’m a thick and chunky kind of guy).

A lot of great business ideas came from finding a niche and doing it well—so well that the dominant player can’t take it away from you without polluting their brand name. There is a reason why Mercedes doesn’t sell their workhorse trucks in the U.S., why the #1 car company out there created a brand name Lexus to sell their luxury car. The name Mercedes and Lexus convey luxury;2 the name Chevrolet conveys ????

I’m no marketer, but if there is a lesson to be learned from marketing it is: Find something and be that something. (Google likes to call this “do one thing really, really well.” This is how Google carved out “search” from “portal.”)

How does my dad get my coffee now and how did he get his coffee then? Somewhere along the way, Starbucks came and broke off a coffee niche from McDonald’s, Dunkin’ Donuts and all-night diners.

Being Google

If you read the last link above, then you know where I’m going. It’s strange, almost futile to “out-google” Google because they own “search,” it’d be like Campbell’s trying to out-spaghetti Ragú: lot’s of money spent in research, only to find out there is no perfect “spaghetti sauce” out there.

But maybe there is no ideal of “internet search” out there, either.

One day, Joseph mentioned that A9 had launched. He told me about how cool it was, and I’ll admit, even though I had been lightly following what they were doing, I liked some of the ideas I saw. But I’m not like other people: I like to have an opinion, even if it’s wrong.

I distinctly remember speaking my opinion at the time: “Yeah that Ajax stuff has arrived, but why would I want to switch from Google?” I just didn’t find the fraction of a cent that Amazon was paying me to use A9 very compelling.

There is this AOL ad on television right now—remember, I hardly get to watch TV. In it the voiceover says: “You should get your search results in plain English, pictures, music, etc. on the same page.

I have to shake my head.

Is this really what I want?

Finding something and being it

What I want is when I’m looking for a book, to find the book. When I’m looking up an actor, to find the actor, when I’m looking up a term, to find the term. When I’m looking to buy a thing, to find the best price on that thing (and perhaps proof that that thing is the thing I really want). When I’m looking for an air fare, to find the air fare.

Specifically, I’m thinking Amazon, IMDB, Wikipedia, etc.

The category is now called “vertical search” and if I can make a New Year’s prediction (and in keeping with the “I’d like to have an opinion, even if it’s wrong” thing): I’d say vertical search will grow a lot this year, and Google will have trouble keeping their piece of it.

I’m gambling that internet search is like spaghetti sauce because that’s how I use search today. As I type this, I can see no less than three separate search bars on my screen: Spotlight, DevonThink’s search, Google search in Safari. There are a ton more different searches one mouse click or key combination away: Dictionary/Thesaurus, Wikipedia, Quicksilver application launcher, the entire Services menu in Mac OS X…

Maybe search is really more like ketchup, and I’m wrong.

But I’d rather have an opinion and be wrong.

Why not Google?

I’ve already alluded to the branding and marketing problem Google will have above and and in my Krugle article. But more than that, I think there is a technological/institutional hurdle to be overcome.

Google likes to think “Yeah we do that. We do search really, really well. That’s what we are about” But look at Google Labs stuff and Google’s core value: the collecting the world’s information value, not the “don’t be evil” corollary that comes from the reality that collecting the world’s information is fundamentally pretty evil.

It’s fundamentally about gather first, cut later.

It’s fundamental to how Google does anything. My friend Mark got fired for blogging at Google (and thus a legend was born). The most interesting Google internal practice he mentioned is not the one he got fired for talking about: the practice was that everything internal at Google goes into an internal Google to be Googleable. It’s an interesting approach to the problem of intranet informational design and it confirmed a suspicion I always had about how Google organizes the world’s information.

Gather first, cut later.

I built the first internet travel search engine in 2000. I like to think I learned something from the whole experience. My gut instinct is that “collect it all into a database and then cut it the way you want” approach works great in general, but becomes worse the more specialized (vertical) my search needs become.

Vertical search is full of vagaries specific to the vertical. For air-travel it’s the fact that the price changes constantly, your search result is not relevant past the 30 minute booking window (if even then). Each vertical has strange restrictions that make it difficult to simply collect and cut. You have to search with a specific goal in mind, cutting your data as you collect.

It’s like making a good chess-playing program. At a certain point, it can’t test out every combination and it has to resort to some sensible α-β pruning.

Froogle was recently taken off Google’s frontpage. Why? I don’t know, but I certainly have an opinion. 😀 My opinion is that it was fucking annoying to see repititious results spread out across multiple entries with multiple overlapping prices. In other words: at least group things by the SKU, dammit!3

So as I think vertical searching will pass general ones, and I also think Google will get marginalized in this new market.4

That’s a tall order. Google is the Microsoft of internet search. Internet search powers advertising. Advertising powers the internet. There is a saying in the valley: “Google makes money off one thing” to which I like to reply: “Yeah, but oh what a thing that ‘one thing’ is!”

From reading the New York Times the only thing everyone cares is about attacking Google head-on. I believe that is the wrong approach—that approach is like a bug trying to stop a windshield.

I’m a bank-shot sort of guy so this is my article on an alternative thesis. Google can own search, but search seems too generic to be kept to some classical ideal. My opinion is to use the Prego approach, find a vertical, and be that vertical.

And I’d rather like to have an opinion, even if it is wrong.

1 Yeah, it looks like my talk on video thing isn’t happening. I should do something about that. BTW, thanks to Paul Reinheimer and Sean Coates for mentioning this talk in the Pro PHP Podcast, I hope you enjoyed the talk but I bet you went to the Digg talk instead. 😀
2 I heard an interview on Fresh Air last month about a cable TV drama. I don’t remember what drama it was, but in a scene on the show, a street-smart actress ad-libbed her character (a young gangsta hit-man) saying to her boss: “The sales guy said this was the ‘Cadillac’ of nail guns, but he meant ‘Lexus.’” Telling. Don’t you think?
3 If you are wondering how MySimon, PriceGrabber, Shopper.com, and others do this better than Froogle, it’s because vendors actually give access to the database of prices directly to them. In fact in many cases, they pay these sites to give the data to them! (Outrageous. You pay to make it easier for them!) The data is already cut and neatly organized by SKU, etc. with the shipping information rules already provided, etc. That’s so much easier that scraping every page on a shopping site and then guessing what’s the product and the price from the web page. And what if the price changes?
4 This doesn’t mean that Google will “lose.” They will still own general internet search and this means the top right of my browser bar. But I don’t think they’ll ever own the top right of my desktop, the top right of my PIM, the top right of my e-mail, the top right of my IDE, etc. Understand?
Some of you may be wondering how what niche Plaxo eCards owns. It owns the ease-of-use niche. There is no easier way of sending and receiving birthday eCards than Plaxo. We own that. It’s not big, but we own it.

6 thoughts on ““Out-Googling” Google

  1. You are right on man. It really is surprising at first, but vertical search is a weakness for Google. And their competitors know it, or at least Yahoo does.

    I almost went to work for Yahoo’s vertical search team in 2005. They really sold me on how they wanted to rebuild most (all?) of their many, many properties around vertical search. The big problem I saw then was that it didn’t seem like their vertical search team would be able to conquer the huge number of fiefdoms. They had to sell it to everybody like a drug pusher: “Hey man, you want some search?” But they had done a great job of stripping down their search technology from Inktomi and built a pluggable platform around it. It made it easy for them to do a great search for Yahoo Autos (they demo’d this for me back then, I was amazed how long it took for it to make it to the masses) or for Yahoo Local, where they leverage local reviews (I think Google is trying to eat Yelp for this now.) Anyways, they had a great technology, even if they have not been able to execute on it.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.