Tuesday, June 1, 2010

Google - the content producer?

A few days ago I read an article in the New York Times about Google and the antitrust concerns.

There is something called competition law. Basically, companies are not allowed to use their dominant position in one market to offset fair competion in another. A good example is when Microsoft bundled Internet Explorer together with Windows, using their dominant position in operating systems to take over the market for web browsers as well. Killing Netscape.

So what's the problem with Google? According to the article in New York Times:
"Increasingly /.../ Google presents links to its own services, like maps, Youtube videos, local business results and product search listings. Executives argue that providing these easily accessible results clearly benefits users. Rivals claim that this is self-serving, and that Google promotes its content even though there may be better material elsewhere." (my bolded italics)
Sad to say, I have to agree. Don't get me wrong - I love Google. Google has enabled me - and millions like me - to build a business on the internet. Without Google I wouldn't exist. Google provides the infrastructure in almost every aspect of my business. But - it is becoming more and more of a real *issue* for us web developers, and for the healthy competition on the internet, that Google is promoting more and more of it's own content in the search results. Let me give you a few examples from my own business - then judge for yourself.

One of the websites that I run is called Sushikartan.se. If you are a sushi-eater in Sweden, you might have used it. It's sort of the top-of-mind sushi website here, with user-generated sushi listings, reviews and pictures from sushi places.

Let's look at Google, and what the search results currently look like when you search for "sushi Stockholm".

As you see, Google pushes it's own map to the top of the search results. And then, below the Google map, there are two results for Sushikartan.se.

To begin with, why does my website, Sushikartan, score well on Google? Mostly because people like the website and link to it from their blogs. It gets a lot of love, basically. The same logic, however, does not apply to Google's own map: It's at the top of the search results not because the content is so good - but because Google decided to put it there.

If you work for Google, you could argue that the Google map is better and benefits users more - and that self-promotion at the top of the search results is justified. So is that true in this case? I don't think so. I'll give you the facts and you judge for yourself. Let's compare the result on Google Maps with the result on Sushikartan, from four different perspectives:

1. Result page:
  • Google Maps: If you click on the link to Google Maps you come to this page. I think most users would agree that the selection of sushi places is sort of confusing. It does not give you a lot of clues as to where to go.
  • Sushikartan: If you click on the link to Sushikartan you get an alfabetical list of all sushi places in town. You can also order the listings according to the places that have gotten the best reviews. The list is based on many thousand reviews, and it *really* shows you the best sushi places in town.
2. User reviews:
  • Google Maps: Google has user reviews, but there are very few. Most sushi places on Google Maps have not been reviewed at all. A well known place like Roppongi has only 2 reviews on Google Maps.
  • Sushikartan: The average suhi bar has at least 10 reviews. A popular place like Roppongi has 57 reviews on Sushikartan.se - almost 30 times as many as on Google Maps.
3. The number of listings
  • Google Maps: About two thirds of all sushi places in Stockholm are missing completely. It's hard to say why, exactly.
  • On Sushikartan there are almost three times as many sushi places as on Google Maps.
This is obviously a huge difference. A 10% difference would be a competitive edge for most services. Now we are talking about 300%. On Google Maps, a search for sushi around Odenplan in Stockholm gives 15 results.

On Sushikartan.se a search within the same area gives 49 results.

You probably agree that this is sort of a problem for Google Maps? But - there's more.

4. Erraneous listings
  • On Google Maps, for a search around Odenplan, almost 50% of the listings that do appear are plain wrong. Places are listed on the wrong address - like postal address instead of visiting address. Places that have closed, or changed name or owner long ago, still appear. If you go to Street View, you can see some of these errors by yourself. Here are just a few of the errors for this small sample search:

Solna Sushi? That's probably the postal address.

As you can see, that's a bakery, not a sushi place.

That is not a sushi restaurant.

That place is a sushi restaurant. But it's been called Esa Sushi for the last two years. Before that there was a café there.
  • On Sushikartan, the listings are updated continously by the users, and the results are around 99% correct.
To conclude: If you want to find a good sushi place in Sweden it would be a bad idea to base you decision on Google Maps instead of Sushikartan. I think you agree? You would quite likely end up in the wrong place of town, looking for a place that does not exist. You would at the same time be missing out on all the good places that actually do exist.

An what about the Google Map at the top of the search results? I think you would agree that the Google Maps is promoted at the top of the search results not because it is better - but because it is Google.

For a service like mine, though, it's hard to compete with the company that is supplying the infrastructure.

What's more: If you search for sushi in your mobile phone, the results are even more skewed in Googles favour. All the "local results" below point directly to Googles own business listings.

Simply, there's no way for Sushikartan to compete here, even though Sushikartan has a great mobile version, and a bunch of iPhone Apps, too.

As I said before, I think Google is fantastic in almost all aspects. In this case though, I think Google is over-using it's dominant position.

In USA, Google was recently in talks about buying Yelp, the leading local business review website. Personally, I thought that would have been a really bad idea. I was almost losing my faith when I heard about it. When the deal didn't go through, I was happy. A deal like that would effectively have killed the competition in the US for local business review websites.

The thing I love about Google is that it gives raison d'etre to many millions of small, diverse content producers all over the world. But when Google step by step is taking ownership over the content in the search results - it's effectively killing many of those wonderful small markets that is once helped create.

I hope someone at Google is reading this. (@Matt Cutts, @Stina Honkamaa)

If you ask me, Google should avoid owning content - and instead focus their energy on creating even better ways to find, make sense, and organize. Stay true to the motto: "Organizing the world's information" - and stay away from "Owning the world's information".

In the case of local business reviews, there's an opportunity for Google to do what it does best: To aggregate, organize, and make searchable all the worlds local content. Not a lot of companies can do that - Google can. Instead of owning content, Google can encourage content producers to supply the listings with geographical meta tags so that Google can make all sorts of "semantic" sense out of it and put together a killer search function. And I would guess that's where the real money is, too.

To put it short: Innovate in search, don't imitate in content.


Unknown said...

Good analysis and very well put. It comes down to the fact that specialized services for your specific need at a given time will always be better than "come all, serve all" services.

The other side of the problem though, is that for times when there isn't a specialized service, the Google alternative can be really useful. And add to that, the Google results are very accessible.

But as you said, the most relevant search result should always be put on top. So if Google could evaluate competing web services against their own objectively and list them in their results accordingly, I think we'd have a solution. But I guess that's really hard to accomplish.

The best solution for the users would be to integrate your and other specialized services directly into the SERP, but that's a whole other problem.

Gabriel Bergin said...

It all depends on how their product "web search" is defined. If the product only consists of the aggregation and ranking of web pages according their algorithm and then matching these pages with a user query, then you are probably right. But, if the product is just search, no matter the content or purpose, then their Maps application would just be another way to rank and display that content. Pushing that application above the results would then be the suggestion of another content display (the user reviews on google maps are mostly aggregated from other sites anyway), not bundling.

I'm neither a lawyer nor a web developer, so this is just layman guesswork. I would like to know what the judicial definition of their search product is (and is map search and web search two different products?).

Ted Valentin said...

@Gabriel: In this example, Google is not only promoting "another way to search". It's also promoting it's own local listnings pages, like this one, for Roppongi: http://maps.google.com/maps/place?cid=7629099842301367084&q=sushi+stockholm&cd=1&ei=zOMETL7NPJzGOInVyZIM&sig2=_mgEOsEslrAY_0xtUtRH2w&dtab=0&sll=59.332711,18.054558&sspn=0.037733,0.053771&hl=&ie=UTF8&ll=59.359971,17.963505&spn=0,0&z=13&iwloc=A

... That is a page where they are more and more buildning up their own content. User reviews etc.

On the mobile phone, all of the links lead *directly* to these pages.

Relevansanalys.se said...

Nice post.

Guess most people agree with your thoughts about Google and their own content being displayed in excess. So do I at least!

Unfortunately, I don't see any changes coming anytime soon..

gobezu said...

Ted, thank you very much, well put and detailed enough for anyone to understand the issues at stake

With all good intentions and naive views we are touting a single actor, in this case google, on the internet in many services on the internet

The side effects of smashing out of business such ventures as yours is really an issue as well

So lets stop being naive about these issues and make our voice heard. And lets stick to the facts.

Annie said...

Great blog post, interesting read! It's been a while since I've been to the land of Yahoo search, but I did a quick run through all the search engines besides Google to see if any others showed their maps first, and they don't! Yahoo actually integrated TripAdvisor as a tab in their search result which is really cool. First, you're served with the search result and you have the option of clicking on Yahoo Travel or Trip Advisor on the left hand side (use Sushi Stockholm as your search). Considering Ask.com owns Citysearch.com they don't integrate that with their search results either.

"Google can encourage content producers to supply the listings with geographical meta tags" I totally agree, let's just hope they listen to this post and all other local search providers! :)

Matth said...

How do you draw the line between content, and aggregation of content though? In particular, between the market for content (which you're suggesting google shouldn't be allowed to enter), and the market for aggregations of content (which it's assumed they may continue to compete in?)

Your argument in this case of sushi is quite persuasive and seems to be based mainly on the quality of the aggregation.

Because aggregations of content are themselves content, you're saying that google should cede to a competitor where a competitor is able to aggregate content better than they can.

Interesting argument -- but whether it works from a competition perspective rather depends on where and how you draw that line between content and meta-content / aggregations-of-content. There's a certain blurry circularity involved.

Joel Dietz said...

Great post and in general I agree with you and what should be the focus of Google's business. That said, I think Google is generally justified in placing the map up top since for most searches (e.g. "car dealerships stockholm") there probably is no distinct user review site.

Also, what this indicates to me is that there needs to be a better way of aggregating reviews of this sort (like Yelp, but beyond).

Ted Valentin said...

Hm, I think there might be some confusion around what the Google Map actually points to.

So, to clarify:

* In the web results, there are some links to external web pages in the list. But what I'm referring to is the links at the top and bottom, going to Googles own "local business listings". (Recently renamed "Google Places"?).

If you are on a mobile phone, there are no links to external pages - all links point to Google Places.