How GIS Lost the Web - Comments on 'Spatial Isn’t Special' by James Fee
James Fee's Original Post from the WeoGeo Blog
I was reading Paul’s blog post last week wondering if he had been putting the Jedi Mind Trick on me the last couple years. I’d like to think that my mind is stronger than that and began to reflect on why I was drawn to WeoGeo after working as a consultant all these years. In the first 10+ years of being a GIS consultant, things were really steady. Our clients didn’t change much (mostly DoD), our billing rates increased over the years as these clients saw value in our implementing professional GIS systems and we never felt like we had too much or too little work. Pure bliss…
Life before Google was easy for GIS professionals
Then came that fateful day in 2004 when Google bought Keyhole. Most of us GIS folks knew who Keyhole was. We’d seen their demo at tradeshows and conferences and thought, “If I only had money I’d so implement that”. But along came Google, who unlike the rest of us has money. They bought Keyhole and soon Google Earth was a household name. In fact, now when we were sitting next to someone on the airplane, we could now utter the phrase, “Like Google Earth” when describing what we do.
But that wasn’t all, we had one more surprise. Google had bought a company out of Australia in 2004, but not too many of us knew about it or cared. That small company was brought on to revolutionize how spatial data was shared (yes even more than Google Earth). Google Maps was everything us GIS pros wanted in our GIS web apps. Unlike our slow ArcIMS or WMS based solutions, this tool was slippy. Users weren’t interrupted as the page had to load and unlike what we were doing (and are still doing) it was intuitive. We all knew right now our existing web maps were broken and we’d have to change.
And change it did, I recall getting calls from so many more companies and organizations than before. They all wanted the same thing, their data on Google Maps. Of course these projects were all that simple, but the end result was the old points on a map routine (with some Google Chart API if you were lucky). It wasn’t so much that they didn’t get geospatial; it was that they didn’t view it as special. And wasn’t this what we’ve been fighting for all these years? Don’t lock the GIS crew in a back room; bring them up in the front so everyone can benefit from GIS. Except one thing, most consultants were built around delivering custom GIS solutions to clients who paid good money for it. These new GIS users didn’t want custom and they didn’t want to pay much.
On top of all it, big A&E firms started to squeeze out the small boutique consulting shops who used to provide much of the GIS support to government clients. Without these contracts, these firms had to look for work where they could find it, providing pushpins on maps showing tabular information. You went from average billing rate of over $100 to just over $50 ($75 if you were lucky). You still could find work at the higher rates, but that was usually far in between. To make up for the lower revenue, consultants started taking on more jobs. Working twice as hard for the same revenue as before, but happy to be working with JavaScript APIs rather than VB6 and Java.
So what happened? Wasn’t the “spatial isn’t special” supposed to free us all and make us more money? Now of course there are some of us still making a good living doing what we are doing, but when something becomes a commodity our standard of living (coding?) decreases. That isn’t to say that we should be wishing for the days integrating MapObjects onto FoxPro. No clearly that wasn’t as fun as you remembered. What GIS consultants did in the past to justify such higher billing rates is to bring something else to the table, something of value that you couldn’t get anywhere else. Converting shapefiles to KML or geocoding addresses from Microsoft Excel onto Google Maps isn’t special, nor are most of the other “GIS” applications being developed today. This means those consultants who either don’t have existing contracts or can’t provide value added services are left to the mercy of the market that doesn’t put much value on spatial analysis. And if you aren’t special, your revenue will be squeezed and you’ll be working very hard with little reward. Don Meltz sums it up very well when he describes GIS Analysts like word processors of the 70s and 80s. When everyone is using spatial technology, how can you be special?
This entry was posted on Tuesday, October 27th, 2009 at 8:30 am and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
The GEOpdx Take
First and foremost, desktop GIS isn't as dead as most would have us believe. Lets save that for a future discussion.
James Fee's post is spot on about spatially enabling the web. This is what we have been striving for in the Geospatial business. I can remember going to my first ESRI conference (probably 1998) and hearing about how MapOjects IMS would revolutionize mapping on the web.
The problem was that MapObjects IMS was short on value and high on price. A basic single server license cost 11k a year and the apps were very difficult to scale. So the dream of spatially enabling the web was dead instantly because not many outside of government could justify those costs. Here is a case of a business model throttling the stated goal in the long run.
Fast forward to today. The original geospatial movers have made distinctive efforts to update in the face of companies like Google/Microsoft and open source platforms like OSM. These competitors have basically created business models that provide a better value to cost relationship with customers for mass market apps. The problem is that the goal is no longer within reach because GIS vendors never created the capitol to get there. They are now relegated to the same market they started with.
The question is: does this situation signal a continuing trajectory? ESRI now offers ArcGIS server - a platform that nobody can afford outside of governments and large companies (again). It is decidedly more capable as a toolbox than MapObjects IMS ever was. Will a somewhat better cost to value ratio be enough to change things or will the business model continue to strangle the offering? Please feel free to make your thoughts known in the comments.
UPDATE: Related Links
Sean Gorman: The Once and Future Map, The Destiny of GIS
Bixel is Outside: Advice for the GIS Newbie
Geofoolery: The future Ain't What It Used to Be
6 comments
Spatial can be special at this price :)
(geomoose.org)
ArcGIS Server is certainly far more capable than its predecessors from ESRI and has a lot to offer, but I think that its cost to value ratio grows arguably poorer with the continued evolution of extremely capable open source applications that fill similar needs. While AGS' capabilities are impressive, it's easy to be distracted by its bells and whistles and lose sight of the geospatial functionality an organization actually needs. Buying a Ferrari just makes no sense when a Honda will fit the bill and do the job well. There are certainly plenty of occasions when AGS is the medicine the doctor ordered, but with the proliferation of alternatives, there are times when something else is the best solution.
In the sphere I know best, local government, my impression is that ESRI is living off of customers who are so invested in an ESRI stack that they're loathe to investigate other options. It's understandable given that migration is no easy feat, but with local government generally under severe financial strain, many agencies are at the limits of what they can spend on new software and additional maintenance. ESRI is going to need to be more creative and innovative than ever to hang on to the local government share it currently enjoys because at some point cost will be a far larger factor than any sense of customer loyalty.
