Google, ESRI and FUD
August 12, 2005 22 Comments
Link – ESRI and Google and major uphill battle Update – Link seems to be back up now
I usually try and not use FUD in my posts because I think the term has been overused, but in this case I think it is appropriate. Berlin Brown seems to have gone off the deep end in regards to ESRI and the “threat” of Google to their business.
While, there is no doubt that any serious GIS user would consider google maps, not even really GIS. But, google has done things that ESRI and other companies haven’t done. They are agile about their approach. They started with just the street mapping, pre-rendered images and then moved on to satellite imagery.
Of course one would have to ignore the plain fact that ESRI has been producing web mapping applications for probably longer than Google has been a company. Actually the problem has been ESRI hasn’t pushed their web services well enough to the point even GIS professionals didn’t know much about it. People fail to realize that Google Maps is just Beta and to try and build a application around a Beta map service is just asking for trouble (well actually you can’t make money off of your Google Maps applications so that isn’t a fair comparison, ArcWeb Services is really the only solution).
GIS is more than just the image data. You have to worry about the geodatabase schema, uploading data, editing features, on and on.Google hasn’t touched this yet, but already they have momentum, so it might be possible.
Actually this is the part that Google might never figure out. While at Where 2.0, when Jack Dangermond talked about this, all the Google Maps folks weren’t paying attention. Basically they don’t care or don’t understand this. Google is all about selling ads, not providing services. Anything they do is driven by the need to place ads on the web page. Can you imagine a company such as HP using Google Maps to show locations where to buy their products and have Dell Computer ads show up? Of course that is why this product will never move much beyond either Beta (which means that they can change the product at any time without giving warning to programmers) or it will only be used by “Google Maps Hackers” who just want to place points on a map.
But, here is where ESRI has failed. One: they have not capture the WEB. Two: they are still operating like a big, bad company. High prices, shady reps, no real value, not agile. Eventually, users are going to go from ESRI to Google. Google is not GIS, but if given enough pressure, they could do it. Already, google has mapped the world with limited resources. Do you really think that they can’t help you create a magical GIS shop.
I don’t think ESRI really cares too much about competing with the Yahoo! Maps and Mapquests of the world. They specialize in providing services that others can use for their maps. They haven’t given up on the web, they aren’t in the same consumer space as Google is. Now to call ESRI a big bad company is a little unfair. I’d love to see some backup to support this. I have never had any problem with ESRI consumer reps or support. Heck the fact that you can actually call ESRI over their web mapping tools proves my point about Google. If I need help I’m supposed to post in Google Groups? Not likely! We all know Google Earth is a nice application, but it isn’t even 1/100th of the power of ArcView let alone BusinessMAP. Oh and Google hasn’t mapped the world at all, they have just bought mapping from other companies who all probably use ESRI software to create these maps. Heck even Keyhole was a huge (and probably still is) ESRI user. The “Professional GIS Software” market is so small, I can’t imagine Google wanting to jump into that market and even if they did, how would they provide ads to the desktop? I can’t imagine having to deal with Google Ads in my applications.
Bye, bye ESRI, it has been fun
I’m curious, what products do you expect to move to? Probably none of them will be Google. If you need to perform GIS analysis, there are tons of open source products out there that are wonderful, but none have the Google name on them.
I will even help you guys out. You need GIS momentum. Drop the 1million dollar price tag for software. Opensource some of your products(not all). Market. Embrace the web.
Wow, with an ending like that, how can we seriously take the rest of the post? Sure one could spend a million dollars on ESRI software, but you’d need to purchase a whole lot of product to get there (about 667 copies of ArcView 9.1 would do it). While I do think ESRI should open source Avenue and Map Objects (when they “kill” it), but why should they open the rest? There are tons of great open source software already available if that is the route you want to take. Should ESRI mass market GIS software? I’m not too sure about that, consumers love Google Earth, but why do they love it? Not because they can perform GIS analysis on it, but because they can see pictures of where they lived as a child. You can’t build a business model around that. And there are few companies that have embraced the web as well as ESRI has. They have been producing web based GIS solutions since pretty much day one of the world wide web. Could they market better? I sure think they could, but that doesn’t mean they don’t understand their place in the GIS world.
Berlin seems to have bought into the Google hype, but he isn’t looking at what Google is offering. Sure visualizing your data is wonderful, but ESRI is much more than eye candy mapping.
About James Fee
James, you are dead on.
There are lots of good things out there to visualize space, but the nitty-gritty processing and preparing of data lies in the uber GIS world.
If organizations can afford it, ArcWeb is looking like the way to go for the use of slippy map technology. I would speculate on the use of MapPoint, but I don’t understand that technology as well as AWS and ESRI products. I think the future process for cranking out good wholesome data is going to be building data in ArcGIS, disseminating information to AWS and also into KML/GML via ArcIMS or ArcGIS Server into 3D viewers like GE/WorldWind/ArcExplorer 9.2.
I think that ESRI has niche and Google has a niche. Heck, even Microsoft has a niche. So, there’s enough work and new technology for everyone. I don’t think anyone is going the way of the dodo either.
Just keep on truckin’, all y’all.
Brown needs to take a little timeout from Google Maps. Did you inadvertently knock him off the web? I cannot reach his site today.
Google Maps and Google Earth can’t do everything. They’re certainly not tools that can be used for processing or creating spatial data. But here’s the thing: Google does what Google does much better than the way ESRI does what Google does. Google is beating ESRI on the ground where they compete.
ESRI doesn’t have an answer (yet) to Google Maps and Google Earth, and with all of the interest we have seen in those products over the past few weeks, how can anyone can think that ESRI shouldn’t be concerned? Do you think that this next generation of geospatial developers who get their start on the Google Maps/Earth API is going to spend the kind of money ESRI demands when they look to set up shop in a few years? They’re getting used to getting things free or doing it themselves.
Spatial algorithms are not top secret and there is a lot of open source software floating around. The outrageously high cost of ESRI software is likely to motivate a few smart programmers to put together a functional alternative spatial data cruncher sooner or later.
“Do you think that this next generation of geospatial developers who get their start on the Google Maps/Earth API is going to spend the kind of money ESRI demands when they look to set up shop in a few years? They’re getting used to getting things free or doing it themselves.”
But Google won’t let you use their API in any product that you want to make money off of. I can’t charge users for a product I make with their API. Its one thing to create a simple mashup with Google Maps, but it is another thing to actually create an application that does something useful.
“Spatial algorithms are not top secret and there is a lot of open source software floating around. The outrageously high cost of ESRI software is likely to motivate a few smart programmers to put together a functional alternative spatial data cruncher sooner or later.”
Explain to me what product ESRI offers that is outrageously high in cost? Maplex was really expensive back in the day, but its market was so small. Is ArcView too expensive at about $1,500? I’d say that is a bargin if Google wants to charge $400 for Google Earth.
Today one could use open source GIS to accomplish almost every task that ArcGIS can do, but people still use ESRI software in the end. I do think in most circumstances, ESRI software is actually cheaper for many companies to implement than open source.
Let’s not beat around the bush: Google Earth has spectacularly succeeded where ESRI has failed or has not even bothered to do: develop an effective client as means of communication and information sharing.
In contrast, ArcIMS, so far, has been a dissapointment. “Out of the box”, it’s a rather raw and ugly thing. Customizing it is a laborious and time-consuming process that eats away resources for other things. Beyond the government/educational dataMILs that deseminate their data in outmoded and exclusive shapefile format, ArcIMS has very few applications that can justify its high costs.
Whereas with ArcIMS, I am responsible for my data, the client viewer, and the server, with GE I can just worry about my data.
Step out of the GIS cathedral for a minute and think about the users: They can have one universal client to manage and organize their realm. This is what potentially makes GE the next killer app. Why have them visit 10 different ArcIMS ArcWeb-subscribed sites?
ArcWeb Services? I guess that would depend on the scope of your work. For me and many other developers with site-specific concerns, ArcWeb Services is probably overkill.
I didn’t know whether to laugh or cry during the 2005 ESRI UC plenary session as Jack and co. demoed the AO reincarnation of ArcExplorer. Perhaps I don’t “get it”, but it looked like a knee jerk response to Google Earth and it was, by comparison, lame. As a long time customer (some would say survivor) of ESRI tools I resent them trying to compete in the commoditized space that Google, MS and Yahoo are fighting over. Jack kept repeating: “Did I mention this is all free?”. I ask, free to whom? What about customers who are paying big $s and are stuck with products that need plenty of work with respect to quality, performance, manageability, interoperability and documentation? ESRI has a large and healthy niche in the professional geospatial world. They also have a responsibility to the customers that define that niche. Watering down their efforts chasing Google, MS, etc. is futile and a disservice to existing customers. Jack, would you please stop chasing windmills?
“Step out of the GIS cathedral for a minute and think about the users: They can have one universal client to manage and organize their realm. This is what potentially makes GE the next killer app. Why have them visit 10 different ArcIMS ArcWeb-subscribed sites? “
That would be the new ArcExplorer Web client in 9.2.
“Jack kept repeating: “Did I mention this is all free?”. I ask, free to whom?”
Actually the client will be free to download. You should be able to connect to any WMS server to get free data or any of the free AWS.
“Actually the client will be free to download.”
More like subsidized by customers who are not receiving full value for their annual software maintenance…
“More like subsidized by customers who are not receiving full value for their annual software maintenance…”
Which product (other than ArcIMS) aren’t you getting value from? I can’t argue with you if you say 9.x ArcIMS isn’t of value.
“Which product (other than ArcIMS) aren’t you getting value from? I can’t argue with you if you say 9.x ArcIMS isn’t of value.”
If you re-read the very section of my previous post that you quoted, I called into question the full value of my annual software maintenance. In my original post I complained about problems with quality, performance, manageability, etc. I have these complaints with ArcSDE, and anything ArcObjects based. Anyway, when issues like these are not fixed through multiple releases, I take a pretty dim view of ESRI’s ArcExplorer initiative. ArcExplorer may be free to download, but it is bought and payed for by every ESRI customer on annual maintenance. I want the core products I rely on daily to be fixed first.
Fair enough… I can’t really argue that point with you. Value of maintenance means different things to different people. I sounds like 9.2 will finally answer some of the issues GIS pros have been having with desktop and ArcIMS over the years, but until we get our hands on it, the jury is still out.
As for your comment about ArcExplorer being thrown together for the conference, I doubt that since it did work pretty well and at least what they showed looked pretty tight. I’m sure the development started before we had Google Earth in our hands. In fact I had heard from a couple ESRI folks that they almost didn’t want to demo it for the fear that it would be compared to or though of as an answer to Google Earth. I didn’t think Jack did a really good job explaining why ArcExplorer is important so many around me just went “huh?” after the demo. What I saw after the demo on Monday made me feel better about the program. ESRI should be showing more of ArcExplorer, but their bloggers are pretty silent about it. Guess that is to be expect considering how ESRI feels about “leaks”.
I’m late to this conversation, but I have been most interested in the response to Google Earth, not just here but in general.
The conversation within ESRI regarding Google Earth and associated mash-ups has been positive from the beginning. From our perspective, a greater number of people now understand that you can do more with maps than get directions. You can map data and perform dynamic analysis. This is a hard concept to explain, and Google Earth has made that conversation much easier.
Google Earth is a consumer product, and ESRI is a business-to-business company. Their business model is to sell advertising, and ours is to sell professional GIS software. There is room in the marketplace for both of us, and we both fill a need.
Well said Lisa, you pretty much hit it right on the head. I think in time Google will value ESRI more as a partner than ESRI will of Google.
James, I appreciate this entry. I think the hype over Google Earth has gotten out of control. The imagery they are serving is pathetic outside of the major metropolitan areas. My hometown is reduced to some crappy Landsat imagery. Why would I want that when I could get statewide 2m Color, statewide 1m DOQQ, or countywide 6inch imagery? Lets give Google credit. The Ajax implementation is very innovative and the use of tiled image caching is very powerful for static maps. ESRI recognizes the power of tiled image caching. Disk space is cheap. I heard at the UC that they are planning to implement the tiled caching in future releases of at least ArcGIS server (hopefully ArcIMS too).
Like it or not ESRI is a major player in web mapping. One of the projects (http://www.mapdex.org) I have been involved with has indexed over 1650 ArcIMS servers, serving nearly 30,000 map services, containing over 400,000 GIS layers. There is so much more out there…
Correction to the BusinessMAP link:
http://www.esri.com/software/busmap/
Check it out.
OK, I fixed that link. I thought it was the same as what you posted, but I changed it to what you had.
“Google Earth is a consumer product, and ESRI is a business-to-business company.”
Lisa -
I think the fact that Google Earth is a consumer product is the biggest reason we should be paying closer attention to it. The next generation of GIS is consumer GIS products. GIS is convoluted and complicated, but maps are generally simple and self explanatory. So how can we get using a GIS to be more like using printed maps?
It may be an uncomfortable subject for many to consider a GIS world without GIS Analysts, but I think that’s where we should be headed. In order for the true value of GIS to be realized, it needs to be simple enough for anyone to use. Google Earth is simple and intuitive, and very powerful, like printed maps.
Google Earth is also now so ubiquitous, that overnight it’s become the defacto GIS client. To direct people at ArcExplorer is silly thinking. More people than have ever heard of GIS have Google Earth already installed, which client do you think we are going to target? If 99% of the people hitting our website have Firefox or Internet Explorer, why would we want require people to use Opera?
Make sure you guys check out Brian Flood’s blog to see what is being done with Google Earth, it has gone way beyond aerial photography. Google Earth client with ArcObjects on the backend, it actually should have been the demo at the UC. I guess there is just no room at the top for partnership…
In my opinion, Google Earth is not GIS. As others have stated in this thread, there is a lot more to GIS than just presenting maps in a cool viewer. GIS was used behind the scenes to create and manage the data that is presented in Google Earth.
Keep in mind that for a very large number of people Google Maps and Google Earth were their first introduction to spatial data. It is only natural that most of these people do not know what goes into capturing and preparing that spatial data. However, GIS people should know better.
It always comes back to the data. Google Earth was being demoed to a client of mine. It didn’t take long for the usual question, and it came from a non-technical person: “When was that image taken?”
So far I’ve seen no business model in Google Maps or Google Earth that would be viable for a company with much less resources than Google or Microsoft. Suggesting that ESRI (or MapInfo et al) missed the boat is a bit premature. The cost of the bandwidth alone must be huge. I’ve heard suggestions that when the satellite imagery was launched in Google Maps it was consuming about one third of Google’s entire available bandwidth.
Now, to these complaints about the cost of ESRI products. People, it’s horses for courses. There are choices other than ESRI, both commercial and open source. If you need a desktop GIS for a single user, or small workgroup, have a look at Manifold (about one tenth of the price of ArcGIS Desktop). MapServer is a robust web map server. The cost of implementing a solution, obtaining data, and paying people to run it, will usually far exceed the cost of the software. If you are not getting value spend your money elsewhere.
I have some sympathy for those complaining about the complexity of ESRI products. However, it’s a bit like Microsoft Word. Most people only use 10% of its capabilities, but different people use a different 10%. ESRI products are created to handle a wide range of situations, and they handle some of those situation better than others. That creates complexity, but it also provides room to grow. You have to decide if the ESRI suite will provide value for your situation.
I have more sympathy with those complaining about lack of support from ESRI (or their representatives in other countries). Yep.
“In my opinion, Google Earth is not GIS. As others have stated in this thread, there is a lot more to GIS than just presenting maps in a cool viewer. GIS was used behind the scenes to create and manage the data that is presented in Google Earth.”
Google Earth is medium to display the information. The same way plotter paper, or a web broswer, or ArcView all are.
Now we could put our head in the sand and pretend Google Earth isn’t out there, or we could exploit it to show our data in an extremely quick and intuitive interface.
How the data gets in there is a totally different subject. You could write a simple dynamic KML that displayed point data from your database using Google’s geocoding on the fly without owning a single piece of “GIS” software.
Or you could write a suite of integration tools to plug into your GIS. Hey wait, that sounds like a good idea.
Or we could just ignore Google Earth and try to explain to people how it’s not GIS, even though its incredibly useful, intuitive, and fast. But its not GIS, not GIS, no not GIS.
Or lets rewrite it and call it ArcExplorer Web or whatever…
“In my opinion, Google Earth is not GIS. As others have stated in this thread, there is a lot more to GIS than just presenting maps in a cool viewer. GIS was used behind the scenes to create and manage the data that is presented in Google Earth.”
Google Earth is medium to display the information. The same way plotter paper, or a web browser, or ArcView all are.
Now we could put our head in the sand and pretend Google Earth isn’t out there, or we could exploit it to show our data in an extremely quick and intuitive interface.
How the data gets in there is a totally different subject. You could write a simple dynamic KML that displayed point data from your database using Google’s geocoding on the fly without owning a single piece of “GIS” software.
Or you could write a suite of integration tools to plug into your GIS. (Hey wait, that sounds like a good idea.)
Or we could just ignore Google Earth and try to explain to people how it’s not GIS, even though its incredibly useful, intuitive, and fast. But its not GIS, not GIS, no not GIS.
Or lets rewrite it and call it ArcExplorer Web or whatever…
Hi Phillip,
I’m not suggesting that Google Earth be ignored (I’m working on ways to get data into it). Nor was I suggesting that it is not useful, or beneficial to the geospatial industry as a whole.
The point I was trying to make was that just because Google Earth is a useful tool does not mean it will replace GIS, or be the end of GIS software companies, as some are claiming. You are right, I should not have said “Google Earth is not GIS”, because the two are not mutually exclusive. Presentation is a critical part of GIS.
Slapdown for poor expression deserved.
Andrew