Dave Naffziger's Blog

I had long assumed that all mapping services were based on the same data since there are only two providers in the market (Navteq & Teleatlas). In fact, all 4 major services (Google, Yahoo, Mapquest & Microsoft), use Navteq data solely for their street maps.

Since the 4 services all use the same data source, I just need to find evidence on one of the services of new road construction that isn’t evident on the others. That source would have the most current maps. There is a great discussion of mapping in general and online mapping services in particular on this great blog that details how maps get created, updated etc. Apparently Phoenix maps had 4000 changes in the last year. Seems like a good city to test with. I dug up the AZCentral article mentioned in the Map Room blog, and learned that Chandler Arizona is a high growth area in Phoenix. A little bit of browsing around Chandler, and I found my first discrepancy: S. Gardner Dr.

Street Maps

1. Yahoo: Yahoo appears to be 1 or 2 quarters more current the the other mapping services. Using Phoenix as an example, this probably equates to ~1K to 2K changes.
2. Google, Mapquest & Local.live.com: All of these three services are running the same versions of the Navteq data (at least as far as I can tell). I was able to see the new road in Google’s satellite map (making that more current than the street data). Both Mapquest and Local.live.com recently updated their mapping services. Mapquest launched a much more user friendly version of their mapping services a week ago and up until a few weeks ago, local.live.com showed a Navteq copyright of 2005.

I believe that the currency of the mapping data is largely an indicator of operational excellence (which is why I’m surprised that Google isn’t at parity with Yahoo). Everyone is likely paying for the same data. Yahoo is just quicker at getting the most recent data live on the web.
Screenshots:

The Best:

The Rest:

Map APIs

The most recent data available via API in rank order:

1. Yahoo: uses the same data (Navteq) as their website. This is currently the latest and greatest map data available on the 4 major sites.
2. Google: uses Teleatlas for their API data. That means that Google delivers lower quality mapping data to API users than they make available on maps.google.com. There is a good discussion on O’Reilly’s blog about Navteq and Teleatlas data, and what is made available through the Google APIs. There is also a good FAQ on the Teleatlas site. It appears that it is pretty clear that the Google APIs provide the lower quality Teleatlas data.

I didn’t investigate what if any APIs are available from local.live.com or mapquest.com. I didn’t comment on the features or overall quality of the Yahoo or Google APIs. The O’Reilly post lightly touches on this topic and indicates that the Yahoo API is very basic, while Google’s is more fully functional.

Our users at Judy’s Book often suggest corrections for the businesses that they have reviewed. One user pointed out that our map showing the location of a business showed an old location. The area around the coffee shop was new and new roads had been built.

I was surprised to find an entirely different map when I checked the same address on Google Maps. Not only were there new roads, there were new parks, schools and other buildings. Take a look at the differences:

Map retreived from Google Maps API (for a coffee shop on Judy’s Book)

Current Google Map

Is this driven by licensing requirements, technical considerations or just poor product planning? I wonder if the Yahoo maps API does the same thing?

UPDATE 1

I asked this question on Google Answers, and Google took it down! Here is the question and their response:

Why does Google Maps API show outdated maps?

I recently discovered that they show different maps through their API than they show on their website. In fact, it looks like they show older maps through their API.

You can see several screenshots here:
http://naffziger.net/ Why does Google do this? Is it due to license restrictions from their data providers, some intentional dithering (GPS anyone), or something else? Does yahoo do the same with their maps api?

Their response:

Thank you for your question ID 755934, titled “Google Maps.” We’ve removed your question because you can find the answer on our main site, free of charge. All publicly available information about Google is available at: http://www.google.com/about.html.

For additional questions about Google, please visit: http://www.google.com/support

Thank you for your interest in Google Answers. Please visit us again.

Sincerely,

The Google Answers Team

OK, so they took down the question. Pretty weak. I searched pretty hard and couldn’t find the answer on their site. I guess I’ll try on Yahoo Answers.

UPDATE 2

I asked the question on Yahoo Answers and received a bunch of useless responses. For those that are interested, you can find it here: Yahoo Answers Responses. Shortly after my question expired, I received a really helpful email from a user (possible Yahoo employee?). Here is the response below:

From: ab

Subject: Why does Google have outdated maps?

Message: Some useful information, since your question is no longer open for answering.

Basically, Tele Atlas must’ve offered better terms on using their map data on the API than Navteq did to Google. Within the industry, Navteq is considered to have better data; they’re also a tougher negotiator. When map data is made available via an API, the data companies expect much more compensation for its use, as it may replace opportunities to resell that data elsewhere.

Yahoo! prodominately use Navteq data on their beta consumer maps offering and APIs - the tiles are being
pulled from the same servers. As of a week ago, 2nd quarter Navteq data - the latest - is being served.
(I helped put it out on the server; that’s why I know :) ).
Google uses 1st Quarter Navteq data on their consumer maps, who knows on their API.

Check out http://developer.yahoo.com/maps/

and http://sidebyside.ning.com/

I hope this helps.

Wow, pretty interesting. I somehow assumed that since Google/Yahoo/MSN/Mapquest were all using data from the same providers that they would have equivalent quality. The email above seems to assert that Google is being ?cheap? and not delivering the highest quality data possible. I’m going to dig into this more.

Garrett Rogers of Googling Google, points out that Google has a relationship with Mark Monitor for the registration of its domains. Mark Monitor further has a relationship with Data Docket, a company that holds many domains that may be indicators of Google services to come (that do not contain the Google name).

A lot of noise has been made about Google’s upcoming gDrive service, as well as their Pagetastic service. The pagetastic domain is hosted on a network range owned by Google (despite the fact that DataDocket has the domain registered). A few searches across other IPs in that range reveals these domains also owned by Google: drivetastic.com, filephile.com. Kind of cool names. I wonder if they’ll be used instead of the name Gdrive?

3/27/07 Update:

I was reading Philipp Lenssen’s post on a Gdrive wishlist reminded me of this post. I did a bit of research and confirmed that Google still owns these domains. The current registrant is:

Registrant:
DNStination, Inc.
(DOM-1291683)
45 Fremont St.
Suite 1400
San Francisco
CA
94105
US

Domain Name: gdrive.com

Registrar Name: Markmonitor.com
Registrar Whois: whois.markmonitor.com
Registrar Homepage: http://www.markmonitor.com

Administrative Contact:
domain admin
(NIC-14299489)
DNStination, Inc.
45 Fremont St.
Suite 1400
San Francisco
CA
94105
US
+1.2083895761
Fax- +1.2083895740
Technical Contact, Zone Contact:
domain admin
(NIC-14299489)
DNStination, Inc.
45 Fremont St.
Suite 1400
San Francisco
CA
94105
US
+1.2083895761
Fax- +1.2083895740

This entity has also registered gdrive.com, and shares the same fax number as Data Docket (confirmed Google domain holding company). It also is registered through MarkMonitor.com, Google’s registrar.

Gmail recently dumped this message in my spam box:

Invitation to Participate in Google Publication Ads Pilot

For now the test is limited to roughly 30 automotive magazines, however I wouldn’t be suprised to see it expand from there.

A few early thoughts:

• Media buyers will be out of a job in 5 years. Start looking for a new career now. Google’s interface speaks to the power of aggregated information, a point and click interface and the future simplicity of buying magazine ads.
• Google will begin to feature demographics within AdSense/AdWords. They’ve spoken about this already, but demographics are clearly compelling to ad buyers. We know they are avaliable on large websites that already run adSense - they just aren’t available to advertisers…yet
• If there was any doubt about Google’s ability expand beyond the Internet, this removes it. Lots of hurdles remain, but this clearly represents an ability to execute outside of their core (or their core is something different than they bill it to be).

A few screenshots:

I’m fed up with Google’s overuse of the ‘Beta’ classification for nearly all of its products.

AdSense publishers may have recently noticed that Google now offers Electronic Funds Transfers (EFT) instead of monthly checks. This functionality is certainly a welcome addition and clearly a good idea for both Google and for publishers.

The new option was released to all publishers (AFIK). However, this is what the publisher is greeted with:



OK, Beta products are a necessary component of any product launch. Bugs need to be ironed out. Broader beta tests are also needed for ground-breaking cutting edge products (which EFT is certainly not).

However, anything that involves transfers of significant funds should NEVER be made Generally Available (GA), unless it is bulletproof. Can you imagine any halfway reputable bank or online trading firm offering ‘beta’ payments?

It is quite simple: if a product is ready to be made GA, then launch it and identify it as such. If a product isn’t ready, don’t launch it. This is doubly true for anything that involves transactions.

Am I missing something?