The GeoWeb is easy, right?
OK, love or hate the term GeoWeb, it does reflect the reality of our workflows today. We are taking our applications off of the desktop and running them on the web, we are combining data silos right inside the browser and we are giving tools to that used to be reserved for technicians on UNIX workstations to users via their mouse. It is truly a new way of working, but I see it running directly in a wall.
Licensing
Here we are at the tip of the iceberg changing the world (maybe a self-serving statement), but we’ve got a chain around our necks limiting our potential. Google and Microsoft’s (among others) APIs are sold the same way IBM sold software before there was the world wide web, large companies can cut great deals, smaller users are left paying full price because we don’t matter. ESRI’s ArcGIS Desktop and Server licensing doesn’t reflect how users are using the applications in the real world (sure, allowing editing on ArcGIS Server Standard instead of Advanced is a step, but it is just one in a long list of problems with the licensing model). Don’t even get me started on Oracle’s licensing. Arbitrary levels of licensing that have no real world basis are killing innovation and requiring consumers of the services to look elsewhere or limit what they can do with technology. I’m not advocating abandoning any of these companies here because there are great business cases to use their software, but their customers are not able to continue business as usual.
Won't someone please think of the users?
Just last Friday, I was talking to a client who was paying for ArcInfo because one hour a month he needed a function that it handled. The rest of the time ArcView was good enough for him. This isn’t just limited to ESRI, most software companies sell software this way and users are expect to pick up the cost just to get some functionality that some committee, in a dark room in their software design center, determined that only “power users” would need is crazy. Sure there are ways to get around most of these limitations using other software, but all these companies are doing is pushing their customers away.
So what do we need here? Google offers their products as SaaS and so does Microsoft. Why does this make sense for “Office” applications and not Geospatial software? Now these efforts of course don’t replace Microsoft Office and that isn’t their mission (well at least Microsoft’s). But what do they do is allow users to extend their collaboration further than the office conference room. Geospatial software is well set up to take advantage of services. Pay for what you use and spend the savings on tools that benefit the end users and not tools that you’d never need.
So I’m not exactly expecting a revolution here in the next year, but unless companies start thinking about the realities of how users are using their software or APIs, we are going to have to look elsewhere or create our own tools. Given what the licensing costs these days, there is money to get that done. Personally I’d rather just use an off the shelf solution and pass those savings on to my clients and get back to building great applications for them
The GeoMonkey only wants what is right
38 Comments
Of course ESRI’s solution to this is the ELA. Of course that is only a solution if you either work in an organization that is under it, or under contract with one.
Since our little group is neither, we pay “full price” and have to deal with both ESRI and Microsoft’s inability to change. Our solution of course is to stop paying maintenance on most of our products since we can’t afford it anymore.
@Lefty Even with ELA, it is a problem. We develop for one client who has everything and then how do we sell same service to client who has zero.
I’d wager most of ESRI, Microsoft and Oracle customers are NOT under a ELA and thus cannot have the freedom as James puts it, “to change the world”. Commercial users such as us are especially hurting because not only can we not pay for license upgrades/updates, but we cannot even pay for the correct support for tools such as PostGIS (which we replaced SQL Server internally with). I’d love to bring someone in to maintain the Postgres for us, but our costs go to ESRI for ArcSDE support.
As James puts it, it isn’t just ESRI, but they and others seem hellbent on ensuring the current pricing model stays put. Otherwise ELA’s would go away and we’d be more ala carte. Now that sounds like a plan to me.
I’m up way too late tonight (or up way too early this morning depending on my state of mind), but yes James… Something needs to be done!
@Barry:
Would these guys offer you any useful options for your Postgres support?
http://www.enterprisedb.com/
“but we cannot even pay for the correct support for tools such as PostGIS” Who pays a license fee for PostGIS?
Lets just say I’m not a Postgres/PostGIS expert. Consulting would be welcomed.
Two words: Manifold
I also like Manifold very much. Especially the long history of database support without middleware is very important. The webserver is nice and easy. I also like the spatial SQL-extensions.
But there are also some drawbacks: *) The cartographic possibilities of Manifold are quite limited.
*) Linear referencing with measures is not supported *) 3D-display features are not really satisfactory *) I am missing full datum transformation support
For my specific needs this means, that I need ArcInfo AND Manifold and the situation still remains quite unsatisfactory.
I wholeheartedly agree with your point here, and its great to hear someone else speaking up, especially in such a high-profile voice as yours is. I think a lot of people are in the same boat as your friend is – and a lot more are simply bypassing functionality all together, and that doesn’t end up benefiting these companies in the longer run. Eventually, either they are going to adapt their pricing model – or the open market (read open source) is going to develop free tools which can rival those they have created, and then where will they be? It behooves both consumers and producers to find a way to satisfy consumer demand at a pricepoint consumers can afford to keep the market share in the long run. I hope those in charge of the decision-making for the big guys hear what you’ve said here!
Dominic
open source
http://www.bigbilllister.com/images/capitalrecordlabel.jpg
Sure, open source solves the problem of licensing, but it still costs nearly as much to implement which I to in my post above.
when your whole argument is that licensing is too expensive with respect to your usage – how is that open source “cost nearly as much”
in any event – all you are arguing is that licensing plans for the software you use does not make you happy. you seem to feel you pay too much considering how much you use it .. you should factor this in to your cost of business. either it is justified or it is not. asking or expecting vendors to change licensing to reduce your costs – you must be expecting costs to increase on someone else to be revenue neutral – or do you expect vendors will just reduce income because it is inconvenient to you? per transaction is nice when you have relatively few transactions but if you actually use the software at the scale it is designed for – – then the current licensing might make sense.. if you want software unencumbered by these inconveniences – use open source – libre.
Consulting costs increase with FOSS. There is no way around that. Coupled with changed workflows the cost is pretty much the same initially.
HUH? Of course it is factored into the cost of business. As a consultant there isn’t anything else I do than that. And of course I’m going to let vendors know that their business model is broken. It is called feedback.
The current software doesn’t scale. It is stuck with unrealistic price points.
LMAO, sure if you live in a fantasy land. Here in reality we work within existing frameworks.
let me guess, the AI guy needed to generate square ended buffers.
my solution has been to take the 4 or 5 AI only functions that I need and wrap them into geoprocessing tools and publish them to ArcGIS Server….then the AV people just add them to their project. Of course this probably breaks some license buried deep deep deep in the EULA….or if it doesn’t I’m sure the pencilheads at ESRI are now furiously writing this exception into the 9.4 EULA.
JW, #1 reason to own ArcInfo.
Buy the ArcView license and write your own solution (or let develop with the money you save from the ArcInfo license) based on ArcView functionality.
Oh hell yea. Nothing pisses me off than having to figure out how to manage our licenses. OK, we get that ArcView is not an “editor”, but the analysis tools seem arbitrarily groups in Info and Editor vs View.
Whoever said ala carte licensing, I whole heartingly agree.
Good post James…
Things are changing pretty fast out on the bleeding/leading edge, and licensing is very slow to catch up. Interestingly, this leaves the leaders only one viable option – use open source. And as they push the software forward, it becomes more compelling for everyone, which is exactly what a COTS vendor should look to avoid. (GeoServer is a perfect example of this)
Maybe ESRI will offer some more options for those looking at building SaaS or data service solutions, and thus stem the tide…
Cheers,
Dave
The more I use ArcView the more I hit brick walls with unsupported features, of course the carrots are dangled in the Help file that if I use the posher versions of Arc these features are magically unlocked. At present I use ArcGIS mainly as a graphics package and leave the editor features to my other GIS software.
@James – Was the missing function feature linked annotation? Having to convert labels to essentially graphics frustrates the hell out of me.
So how would you price an offering if you were running the show? Everything ala carte (like an app store)? Capacity pricing (set price for “x” storage “y” cycles” “z” bandwidth”)? Per user/seat pricing (Basecamp style)? Transaction pricing (per click/tile)? Curious what would make folks do back flips other than what ESRI/Goog/MSFT/Oracle are doing now.
Sean -
From the perspective of building SaaS/Data Service solutions, a capacity type fee would be best. For me, I’m basically interested in a model that scales linearly.
For consumers of the SaaS, it depends on the type of functionality. Per user works if the functionality is tied to a person. For example, we use Rally to manage our agile dev’t process. It’s per user, and it makes sense because everyone manages THEIR tasks in it. Per seat works if it’s more “utility” functionality that’s not tied to an individual.
I think that Data Services are ideally metered, but many agencies can’t deal with “unknown” billing, so having the option for flat, fixed rates for “unlimited” transactions is also important (even if it costs more).
Cheers,
Dave
ESRI licensing may be the dumbest licensing system on the planet. I don’t even know what my company has or where. We are far too small for an enterprise license, and I’m beginning to think that ESRI press will soon publish a book: Getting To Know Arc Licensing so we can keep track of it all.
Like your client, there are functions I would like to use occasionally from editor or info licenses, but not enough to justify the cost. Some of these I do complicated work arounds importing and exporting data and coding some solutions myself.
My favorite is Maplex. Imagine I’ve just made a map someone requested, but it’s too ugly to give to the client. I can either sit around converting features to graphics and manually moving and adjusting them, send it out to an image and rework it with publishing tools, or I can by Maplex (or ArcInfor). Now how do I explain to my boss that I need to buy another $2500 software license because the software we bought to MAKE MAPS makes TERRIBLE maps! Where’s the logic in that?
You only have to explain why you did not think about the user needs before you bought the software.
Just to be defensive, I didn’t buy it, it was here when I got here. And of course, we need it for much more than just making maps, but what the boss sees is the output map, not the analysis tools that went into producing it, which really highlights the basic problem. But really, eventually people will start thinking about the user needs before buying the software, and with the explosion in the number of available tools, ESRI will have to fix their licensing schemes or eventually they will lose the market. I know, there was a smiley face on your comment, but it was another opportunity to harp about ESRI licensing.
@Dave @Sean Gorman: There are probably two models here. One for web services and one for “desktop”. Someone above mentioned Maplex, in that example at least I can license that product with ArcView. It is expensive, but it can be done and done cheaper than buying ArcInfo. I’m sure functionality can be grouped and licensed “by seat” on the desktop and metered on the server.
getting off topic – separate Maplex licensing is really hard to deal with as it really impacts portability of MXDs. For exmaple, author a nice map with Maplex labeling and bring it to an engnine app w/o the maplex extension and you get bad results. Or the other way around, maybe your engine app has maplex built in, but someone edits the MXD with an ArcMap license with no maplex and hosed up labeling. Really problematic to have the label engine, and then the good label engine that costs more – especially if you have thousands of engines apps deployed to many hundereds of customers like we do. It would really be better if labeling was just labeling, and it just came with the ArcMap license. All the other extensions I get, but not labeling as it really hurts basic MXD portabliltiy.
Anyways – variable pricing for usage is extremely problematic for our customer base which is public safety. My customers say, what if I run out of credits the hour before the huricane hits? We went through that back with ArcWeb Services, and it was always just a tough sell. For us, an all you can eat price model is important, as you have noted.
Hmmm. I wonder if I should take your twitter comments personally.
Only if you don’t plan to do anything about it. We get dealt bad cards, but sitting back and using it as an crutch as to why things don’t work, or why they are too expensive isn’t going to get things done.
Take charge and make it your own!
Boy I’d love to license GIS software this way. It is becoming way too hard to make business cases for SQL Server 2008, SDE and ArcGIS. All are critical (well maybe not SDE, but that is another comment) to our business process, but eat up so much costs.
There has to be a better way out.
Wow, excellent article. Throwing my hat into the ring wishing ESRI, Microsoft, Oracle would evaluate how people are actually using their software today. We do use open source as much as possible, but we still need to interface with existing products that our teams use. I’d love to be able to take advantage of functionality that I cannot because I can’t affort to license softare.
And of course we see open source trolls saying the answer is clear. It of course isn’t that simple and to just throw away millions of dollars of work is a good way to get you laughed out of the boardroom. This is a conversation where FOSS can’t help.
@wah
You are going to have to do better than terms such as libre to get FOSS in corporate America. I suppose if you are a euro and don’t care, so be it. If you are one of us then you’ll know that reality dictates that you can’t just abandon work that others have done and then spend hundreds of thousands of consulting dollars to get back to where you are now.
Or maybe you are in college and reality hasn’t hit you yet to what we have to deal with on a regular basis.
from reading the comments it doesn’t look like the esri business model is “hopelessly broken.” it looks like they are in fine shape. a network of dealers who see little alternative to their product – and scorn alternative business models as juvenile or worse -european. ESRI prices are high enough to cause some whining but not enough to cause you to give up a trip to san diego and some chicken wings for your trouble. Its the -i’m an esri implementer- consulting business model that’s broken. its like being an oldsmobile dealer trying to differentiate your dealership. you can put out balloons on saturdays but the problem is still with the product you are selling. i’m sure it will all be better with the next service pack ….
We’ve had long talks with our Microsoft rep about licensing that meets our needs. They can’t work in small enough chunks to actually assist us in being productive. We are stuck on SQL Server because our accounting department uses it. I’m trying to put together a business case for Postgresql so that we can abandon the limitations of SQL Server and continue developing Geowebish applications.
ESRI is a whole other matter as well, but if we can get out of MicroSoft, we can then get out of ESRI, eventually. I don’t want to do it, but $40,000 for a server, plus ArcEditor license to work with the data. Not going to happen here.
Dumb question: Isn’t part of the problem also that people tend to get locked into the ESRI (+ SQLServer) stack, instead of being able to mix in compatible alternatives to suit their needs/budget, as we can in mainstream application development? For example, ArcGIS on the desktop is great (even if the multi-layered licencing is not), but I wonder if so many people would necessarily pay through the nose for ArcServer if they had other options? The tendency towards lock-in surely reduces the pressure on ESRI to modify its licencing costs/conditions.
When you own a market niche (ESRI) or pretty much the whole thing (MSFT), your business model is subservient to the market share you already manage. Think http://en.wikipedia.org/wiki/ULCC instead of http://en.wikipedia.org/wiki/Cigarette_boat
When you’re subservient to your user base, and a revenue stream, changing course is often institutionally impractical (or impossible, if the exec. mgmt are happy with the status quo).
I believe the biggest impediment to price change and adjusted business models is not ESRI in the geospatial trades. I think the barrier is inside the ESRI channel. Implementations companies rely on ‘hard-to-program-means-more-hours-sold’ for a time-sold revenue curve. Perversely, a change to allow ‘by-the-drink’ licensing and SaaS resales effectively breaks the time-sold stream.
A while ago, a couple companies sprung up to do just this – offer ESRI tools through an ASP model. This was back in the day when the internet was still ‘kewl’. The folks at ESRI allowed it to happen, and watched the results. After watching adoption skyrocket at the edges of their customer influence, and seeing the traction the approach was getting, the Redlands folks changed their reseller license pronto. They stepped on that approach fast – and didn’t replace it with anything meaningful – so as not to cannibalize the support streams for themselves and their partners.
ESRI really, really, really likes the renewal stream from all those ELA’s scattered across the community, too.
We can wish, but the Gorilla must eat.
Nothing is free in a free Country.
And yes you ALL are still going to this years touchy “I love GIS” UC.
Me to ESRI Rep. “I don’t want free passes to your propaganda show aka UC, can I get a better price instead?”
Question: What makes for an expert in the field? It isn’t because you jumped some dopey hoops and became an GISP and attended every UC in the past 10 years. And the fact that you worked at ESRI tech support does not make you and expert either…
No – what makes you and expert in the field is the fact that you have made a lot of mistakes with many different products. This fact means you can steer clients away from making those same mistakes and into a product that they will truly use and appreciate.
And Yes sometimes it’s not a mapping product (Yes I said it – SOMETIMES IT’S NOT A MAPPING PRODUCT)
We all need to go on a ESRI diet.
Boycott this years UC and spent it learning something non-ESRI instead.
regarding desktop licensing: we have a deal with our ESRI reseller that allows us to switch between licensing models (view/editor/info) on a project basis. one month we work with arcinfo, followed by three months where we are fine with arcview only. same with extensions. that licensing model allows us to tailor licensing expenses very well to our projects – not quite SaaS but close, since we only pay what we need per project. not sure if it’s 100% ESRI approved what that reseller is doing though…
Another way to view the situation -
Think of SaaS vs fat licenses as Mass transport vs Cars
If mass transit were available wherever you wanted to go, most folks would be unlikely to buy cars. But since it ain’t – those folks buy cars to get from A to B, and for ‘potential’ mobility.
Now, since the infrastructure of ‘the internets’ is available pretty much anywhere, perhaps it’s only a matter of time until folks give up their resource-intensive tools and start working off mass services.
Or not. ;^)
If you have an arcview license of ArcGIS and want more functionality, look at ET Geotools. For $195, you get most of the ArcInfo functions. I highly recommend this software. http://www.ian-ko.com/
2 Trackbacks
[...] Spatially Adjusted (2067) – fairly technical GIS blog by professional James Fee – recommended: Licensing issues on the Geoweb [...]
[...] this is against a backdrop of web mapping (aka the geoweb) APIs from Google and Microsoft which are provided for free, along with fast, free, background [...]