Overselling ArcGIS Server
April 1, 2008 18 Comments
Dave Bouwman has written up a follow-up to his question at the Dev Summit closing session about ESRI overselling ArcGIS Server.
One issue I seem to run into a lot is the gap between client expectations of ArcGIS Server, and the reality. Although I did bring this up at the closing session of the Developer Summit, and we were told this would be addressed, I think it’s worth pointing out some specifics.
This is a HUGE issue to many of us in the consulting business as the realities of developing with ArcGIS Server sometimes doesn’t match up with expectations.
ESRI Salesman out the door to push ArcGIS Server capabilities to the masses.
About James Fee
Looks like he came with the briefcase and left with the bag of cash.
I’ve noticed this as I am subscribed to an ESRI list that e-mails me when there are free training seminars and they are almost entirely about ArcGIS server. I sent an e-mail to learnGIS@esri.com and asked this question:
Do you have any other software products besides ArcGIS Server ? It seems like all these live seminars are all about something I’m not using.
Their answer focused on a link to their products page. I guess my sarcasm was lost on them !
Thanks to Dave. I will mention that this is not just a huge issue in the consulting world, because in-house developers have “clients”, too. Clients who have certain expectation levels and who make commitments that we must implement.
Whenever I read about the fabulous geoprocessing capabilities of AGS, I always try to find out what people are actually, realistically accomplishing in that arena. I’m not so much interested in a REST api or a javascript API as I am in seeing a server with performance that makes me go, “Wow!”
i hear ya, i hear ya!
The sad thing is that ags really has some exceptional (if buggy) capabilities. Many of the performance issues would be present in any product that attempted to do complex cartography on the fly. IMO what has been oversold is the concept of web based GIS. Complex GIS simply work better on thick clients.
The overselling of ArcGIS Server has been an issue for a couple of clients that we have been dealing with. We finally got the network topology and securities approved @ a client’s shop for a production ArcGIS Server 9.2 implementation, and the sales rep wants to come in and start pimping the benefits of 9.3.
Plus, the Linux side still is kind of squirrley with poor support. But of course all that is fixed in 9.3 and THAT is why we need to move in that direction. Sorry, don’t want to get too far off this thread…
Everyone should remember this: “under sell and over deliver.”
For me, I think the overselling thing is no big deal. Or more accuratley, I don’t think they are overselling. Companies sell products, and pitch their value propositions – that’s what they do. They create value, and then sell that value to people who will pay them money for it because it solves their business problems. If they are not constantly selling what they have, they do not get to stay in business.
ESRI is really a platform company. They can’t afford to go vertically deep into every possible domain – nor should they. That would not be a good use of their time. Rather, they build a platform that enables us – users, developers, ISVs, etc – to take the thing as deep as we want to, based on our own , sometimes lifetimes of, domain specific expertise. How deep we go is usually dependant on our own needs, expertise, time, and talents, and resources.
That they have created a platform product with loads of potential capability just waiting to be exposed in new and remarkable ways by us (users, developers, ISVs, etc), just means there are loads of opportunities for people using this product. Its not a bad thing for your customers to hear about this new capability and then ask you about it. If you’re in business, its good for customers to aks you for new value. Thats not a bad thing.
If you can’t go as deep as you want out of the box, and actually you can get pretty far now – farther than you ever could before with ArcGIS Server, and don’t have the time, inclination, desire, to dive deep into it at the object level, find a developer, ISV, ESRI professional services, consultant, etc. to take you where you want to go. Because there are so many objects, there is really an infinite number of ways of putting it all together.
@Ron,
I like that.
John,
If I sell you are car and you decide to use it for racing in the Baja 500 but the car I sold you doesn’t have a solid chassy then good luck in modifying it to fit your needs. ESRI architecture is surely extensible but only so far. In my experience if you push too far ESRI technology tends to eventually break down. It’s one thing to extend the car a little or even moderately but if we as developers have to build lots of code on top of it then it really becomes necessary to build the car from the ground up. So far the car that ESRI has provided us is slow, can only be customized to a limited degree, breaks down a lot, etc. Mostly ESRI needs to be honest about this, not fill us full of hype. Now, in all honesty, my ESRI contacts have told me truthfully that AGS isn’t quite ready for “prime time” and I appreciate that but the marketing folks in Redlands are still out pushing AGS with the opposite story. It’s clear there’s a disconnect going on here which will eventually resolve itself when the rubber meets the pavement. In other words, when the technology catches up with the hype.
Ron
In many organizations, the sales department is completely disjointed from the engineering department. The sales folk make the promises without compunction, they just figure it’s up to the techies to actually keep those promises…
@Dave,
Very true – ESRI is definitely not the only company guilty of overselling.
What discourages me about AGS is that I don’t see many public sites using it, and the ones I have seen generally have maps the size of postage stamps.
I’ve recently installed ArcGIS Server at my work and I’m amazed how slow it is when only several users are on it.
I’ve heard that there are discussions of a class action lawsuit.
~Kenneth
http://www.kennethkay.com
ESRI is really a poorly developed, poorly conceived product. They took more time than anyone to move to dotnet. Try loading all streets for a state like Florida and they develop a graph (network) for the road. It is quite a simple though.
ESRI is really a poorly developed, poorly conceived product. Try loading all streets for a state like Florida and they develop a graph (network) for the road. It is quite a simple task though. What their software engineers thinking? ) or are they thinking at all?!
I hear Jack will be performing this at the UC this year.
Back in the Developer Conference in 2007, ESRI had commented that ArcGIS 10 was not going to be about new functionality but was going to be about improving what is already there. A new graphics and processing engine seem to be the core of ArcGIS 10. They know it’s slow and needs improving, here’s hoping for the best.
@ Rudy….I bet TransCAD and/or Maptitude can do it just fine?
KoS