One of the key drivers of the SaaS model has been the transfer of risks, burden and costs of ownership from the software buyer to the software vendor. In a way, by adopting SaaS the customers are telling the software industry to get their act together and bear the risks for their own doing.
For the longest time, customers have borne all the growing pains, quality issues, large capital expenditures in licensing, managing, updating and maintaining enterprise software. Now the tables are turned. With SaaS, the vendor is now saying, “You the customer, just pay for the use and we will take the responsibility of providing the same software on my premise and we will be responsible for upkeep, upgrade and up-time. All for a flat fee that comes out of your OPEX.” The statistics of SaaS adoption bear out that the customers have liked the message and acknowledged with a resounding “YES”. Vendors not embracing this SaaS movement are slowly seeing the truism of Richard Waters’ “The end of Software gravy train“.
With the economies of scale, the costs SaaS vendor incurs in managing the application will eventually trump the costs the customer incur if they did it themselves. But in the tough climate we are in, revenues hard to come by, coupled with the ongoing customer’s concerns around outage, data breaches and coincidentally, broader adoption of Cloud Services like Infrastructure-as-a-Service (IaaS) from Amazon EC2, Rackspace etc, a new hybrid model of delivering SaaS applications seems to be emerging and gaining interest.
SaaS vendors who support this hybrid model where the customer now has the option of subscribing to their service but choosing a cloud based infrastructure to host it.
Bruce Richardson of AMR Research wrote a post recently on his discussion with Don Klaiss CEO of Compiere around SaaS and Cloud Computing.
Don makes his observation about how Cloud Computing with its lower TCO, flexibility would become the next generation on-demand software delivery model. Specifically he went on to say
“Cloud computing is the next-generation of software-as-a service, in which a complete software environment is licensed as a subscription from a software vendor and low-cost, secure, and dependable IT hardware infrastructure is ‘rented’ from a utility-computing provider on demand. The customer has complete control over its own secure and private IT environment at a very low cost and without the hassle of procuring and managing its own data center. It can quickly scale IT resources up or down as computing needs change. And [the customer] has complete freedom to customize the solution as it sees fit and complete control over upgrade cycles and all other aspects of its IT environment.”
This symbiosis of Cloud Services and SaaS applications will definitely have arguments on both sides. So let us look at the merits and demerits of what this means, to customers, considering this model.
Merits
- As a customer you will no longer need to be concerned about losing control of the your critical applications, data and management. You will own the environment.
- If the vendor goes bust, you already have your application, data and hopefully with your IT resources already trained, you will be able to mitigate most of the risks to business continuity.
- Unlike a on-premise solution, you get the infrastructure that you can control at a fraction of the cost with Amazon Web Services (AWS), Rackspace or any other cloud based infrastructure provider – so you get the benefit of lower TCO. So all the benefits of your own environment minus the painful part of negotiating technology contracts with hardware vendors.
- One of the biggest challenges in most SaaS offerings is around business intelligence and reporting. Most offerings around reporting are merely a shadow of what you can do with your own data. Your end users who might have been trained on a corporate standard tool like Business Objects/Micro Strategy/Cognos would have had to get trained on the new reporting tool of the SaaS vendor. Customers typically work around this by negotiating the ability to get copies of their slice of data from SaaS multi-tenancy so they can run their own reports on-premise. With customer-owned cloud based infrastructure, and data under your control, now you get to use your own Business Intelligence, Reporting tool.
- With the source code being available to you, you will have luxury (depends on who you speak to) of making customizations which are not available in a true SaaS model. In true SaaS model you are limited to configuring business processes, UI changes/branding using the switches and knobs provided by the vendor.
Demerits
- As a customer you now have to worry about EVERYTHING outside of application changes. You will be responsible for all the maintenance, security, data management and the costs. The economies of scale that SaaS vendors achieve doing it for many many customers is not available to you.
- You will now be responsible for updates, change management to the application. This will over time introduce the need for SaaS vendor supporting multiple releases, considering that realistically, left to themselves customers would define their own schedule to upgrade. With that need comes additional costs that SaaS vendor will incur and that costs will be transferred over to the customer.
- Irrespective of what your corporate standard for hardware/IT is, you will have to make do with what the Cloud provider supports.
- You will now have train your IT to manage cloud infrastructure and include that under your corporate IT governance guidelines.
- You will be responsible to implement Disaster Recovery environment on your own. This almost doubles your cloud infrastructure cost. With a SaaS provider this would have (in most cases) been included as part of the service.
This short list should give you a frame of reference to evaluate this model.
To me, there are definitely slices of benefits when compared to the on-premise purely based on the transfer of management overhead and upfront costs to Cloud service provider and some benefits from agility of a SaaS vendor. But that said, one of the benefits of SaaS is if things went wrong you know who to go to but with a hybrid approach now you will invariably get ping-ponged between the two vendors.
A better model would be for the SaaS vendor to use cloud and also provide access to the data, ownership to the customer but still be responsible for the management, maintenance.
Would love to know what you think about this? Send me a note and I would be happy to discuss this more.
About Subraya Mallya
Subraya Mallya is a technology executive, entrepreneur involved in conceiving, building software businesses. In this blog, he focusses on business and technology strategies around Business Model, Product Development and Marketing. Connect with me on Twitter - @subrayamallya
August 20th, 2009 at 6:01 AM
Intriguing article and well written, thank you.
One interesting and oft missed aspect of (at least AWS) Cloud Computing is that the need for DR is somewhat mitigated in that data is replicated behind the scenes by the provider within their infrastructure. What that means IMO is that we are viewing the need for DR in a traditional enterprise model still while moving to the Cloud–notably a completely different approach to computing in general.
Cloud providers have SLAs and therefore have a vested interest in NOT losing your data. In that sense and finally, service providers have stepped up to the plate and have raised the availability bar in their IaaS. While it does not eliminate the need for DR and BC planning, it does not necessarily double the cost of your enterprise when considering a DR strategy.
I think that one of the biggest mistakes and misconceptions is that my enterprise and infrastructure costs will be drastically reduced by just moving my existing architecture into the Cloud.
In simple terms, owning a server means that you can ramp up the processors, memory, storage and network as much as you want and it won’t cost you a dime more than you have invested in it. Take that strategy to the Cloud and you’d better make sure that resource consumption leads to revenue generation or you’ll be supremely disappointed that you made the move.
We just need to think somewhat differently about computing. It’s more like a hotel: rent a resource (room), pay for other services (laundry, room service and even that $8 bottle of water), then check out. If friends come over, add another room to the bill for as long as it is needed. No long term contract. In the old days of the west, families used to live in hotels. The Cloud is not that different, heck, there are already hustlers and rustlers trying to break in an steal things–especially off the new cowboy who just bellied up to the hotel bar.
(Tipping hat)Mighty fine article. Much obliged.