29 July 2011
I was interested to read a recent article by Joe Francica in Directions Magazine – “The Cloud’s Disruptive Value Play”. In the article, Francica considers a statement about CAPEX and OPEX in relation to cloud computing, and applies it to the geospatial industry.
I completely agree with Francica’s end assertion that cloud-based GIS favours OPEX efficiencies. And while I would generally agree with his comments if they were only applied to large GIS software vendors, they seriously underestimate the strengths of our offering.
We’re not just ready for the cloud – we’re already there
Francica asserts that the vendor industry is not ready to provide cloud delivery. While this is true of the geospatial industry in general, there are exceptions.
eSpatial is one of those exceptions.
We have gone well beyond thinking about the cloud – and are actually delivering GIS software via the cloud to paying customers.
As Francica correctly asserts, the user community and technology buyers are ready for cloud-based GIS. Pioneers like eSpatial, who have embraced cloud delivery for GIS software, are reaping the rewards.
I understand why the majority of vendors in the geospatial industry have not moved to cloud-based delivery. Delivering a cloud-based solution involves making a significant technological investment that results in potential customer base cannibalisation – while simultaneously embracing a challenging business model.
It’s not for the faint-hearted.
Nor can we expect much interest from the large incumbents.
Experience in other sectors of the software industry shows us that they feel no need to lead the way into the cloud. Instead, the incumbents will pay lip service to the concept, and will only be dragged – kicking and screaming – into providing true cloud delivery when they have no other choice.
A notable exception to this rule is ESRI, who seem to be working towards a real cloud offering in a positive way.
Cloud delivery does not mean compromised functionality
While GIS software vendors may choose to provide less functionality in the cloud, it is simply that – a choice, not a fait accompli.
There is nothing inherent in cloud delivery that makes full GIS functionality impossible. Limitations are imposed by the vendor, not the delivery model.
At eSpatial we made the conscious decision not to cut down on what we offer customers, committing instead to provide full, enterprise-grade GIS functionality in the cloud. Our cloud customers access exactly the same GIS capabilities as our on-premise customers do in their corporate GIS.
Indeed, while we provide full functionality in the cloud, not all of our customers require it. Our biggest challenge is to simplify our offering, to accommodate smaller customers who have less comprehensive initial requirements.
Cloud delivery isn’t a flash in the pan
The geospatial industry may have been slow to embrace cloud delivery, but it’s not a fad: the cloud is a trusted, credible delivery method for business software.
Thousands of companies have become very successful by embracing the cloud – and hundreds of millions of users use cloud-based software in both personal and professional capacities.
The global trend for business software is towards cloud delivery, as the continued success of Google’s business offerings – and Microsoft’s increased commitment to cloud delivery – prove.
We don’t believe that the geospatial industry is so very different – cloud-based GIS is not just a possibility, but an eventuality. And the high level of interest in our GIS software only reinforces this opinion.
Not all GIS buyers even consider CAPEX vs OPEX
The geospatial industry tends to only consider current GIS users when talking about its market. This group of organisations and users are certainly interested in the CAPEX vs OPEX discussion – but they’re a very small percentage of the total potential GIS software buyers’ market.
The remainder of this potential market are new to the concept of GIS. They seek the insights that GIS can bring – but don’t view GIS software as a core application – instead, it’s seen as a useful business analytics tool.
These potential customers wouldn’t even consider the CAPEX investment that traditional GIS software involves. As we have discovered, a full-function GIS solution in the cloud is the perfect solution for them: full capabilities delivered in a trusted, reliable way; flexibility around user numbers; and low costs from the OPEX budget.
While the question of CAPEX vs OPEX is relevant to organisations who currently use traditional GIS software, it’s important to acknowledge that the vast majority of businesses simply won’t consider the traditional model.
Cloud delivery provides the geospatial industry with the opportunity to greatly expand its reach. The only question is – why are other vendors taking so long to embrace this opportunity?
Philip O’Doherty
CEO, eSpatial
Great stuff eSpatial! The one comment I would make is that it seems everyone confuses or automatically relates "Cloud GIS" with/to multi-tenant SaaS.
Cloud applications don't have to be shared (at any level: not the application level, not the infrastructure level, and not even data center level), in fact they can be more secure and stand-alone, than traditional IT/in-house hosted applications, all while maintaining the scalability the Cloud offers. Of course, as a customer, the more you are willing to share the better the economies of scale become and the Cloud application vendors will usually pass that cost savings achieved by sharing to their customers.
I am not proclaiming that it is insecure for customers to use a true multi-tenant SaaS application, in fact, as I alluded to before, many customers probably upgrade their physical and virtual security when they move from a self-hosted application to a cloud-based application; but, what we do see a lot of in the case of enterprise clients is multi-tenant SaaS applications limiting integration opportunities with customers other enterprise applications because they only make the data stored inside the SaaS app available via API call.
The problem with this is that if the customer has a true enterprise, multi-faceted geospatial workflow, where they would use the GIS SaaS app as just a part of their workflow larger workflow, it becomes hard to work that off-premise hosted, shared SaaS app into the enterprise GIS workflow because everything has to be passed via manual data upload or http API call.
So, if the customer wants to set up a Virtual Private Cloud Environment, whereas they have their own IP's/subnet and VPN access to an IT environment so they can securely pass data between servers, database, files, etc., the SaaS app that sits outside that environment becomes a limiting factor because the customer does not have full control of the back-end nor access to the data unless over http API call.
We have been very successful with helping large organizations overcome this by pushing the concept of single-tenant SaaS to our customers where we give them a software package (e.g. Esri ArcGIS Server, ERDAS APOLLO, OpenGeo, etc.) fully installed and configured on a cloud server (they can also choose to their cloud servers placed on dedicated hw nodes or move the cloud server internally to their own firewalled infrastructure), but the application is not shared in any way, so they may customize it (not just skin it) and have full back-end access to the root server also.
I would be curious to know if you at eSpatial found it easier to sell a multi-tenant hosted SaaS application or do you find that customers sometimes say "I want your app, but I don't want to share with others?"
Cheers! Thanks for the post… again, great stuff!
Phil: I think where you can help the community in the cloud discussion is to articulate the process eSpatial went through to arrive at its delivery architecture. When I heard it explained more completely at the Oracle Spatial user conference I more fully understood both benefits and limitations of the cloud infrastructure and why you deployed as you did. That kind of thorough due diligence is probably not appreciated very well.