Gartner published Gartner Fact Checks the Five Most-Common SaaS Assumptions this week. Let me share some of my experience driven insights on this subject.
“Assumption 1 — SaaS is less expensive than on-premises software. Gartner Fact Check: True during the first two years but may not be for a five-year TCO.”
In my experience, for on-premise implementations the devil is actually in details. Especially, in large enterprises adding a new enterprise application to the existing complex IT support context may mean steep learning curves, long lead times to procure, install and configure the necessary infrastructure components and hefty opportunity cost, exacerbating the risks and delays in launching the intended value to the organization. So, to compare apples to apples, the TCO for on-premise implementations also need to account for quantification of these risks, say the opportunity cost if the solution is delayed by six months due to the delays in procuring necessary infrastructure components multiplied by the probability of risk.
“Assumption 2 — SaaS is faster to implement than on-premises software. Gartner Fact Check: True for simple-requirement SaaS, which will be faster, but growing complexity and other factors are coming into play.”
While the business process standardization challenges are the same for both of these modes, I see that most of the on-premise infrastructure risks are jettisoned with SaaS, contributing to the shorter implementation times. So, when time is essence, say immediately reaching out to a underserved consumer market, in my opinion SaaS is the clear choice.
In addition, I noticed that most of the SaaS vendors have sound understanding of their product functionality and clearly defined interfaces to the major external systems, further shortening the implementation critical paths.
“Assumption 4 — SaaS does not integrate with on-premises application and/or data sources. Gartner Fact Check: False.”
While some big players, such as SaleForce and Workday, are successfully embracing web services and mashups for real-time and frontend integrations, many SaaS vendors are still paddling with FTP style full load batch interfaces. Integration capability is one area that is often overlooked by many SaaS vendors. Also, in addition to knowing about one/two primary systems that the solution must integrate with, SaaS vendors have imminent need to develop a comprehensive understanding of systems context that the SaaS solution should interoperate with.
“Assumption 5 — SaaS is only for simple, basic requirements. Gartner Fact Check: False.”
I totally agree. If solution is too simple, why go SaaS when it can be easily accommodated with simple forms-based internal solutions.
Also, without reading Garner’s detailed report, not sure whether they accounted for the pricing model differences across SaaS vendors/segments. For example, Talent Management/employee self-service style SaaS vendor pricing may be quite different from CPM style SaaS offerings. Perhaps, the TCO by segment may provide better insight into SaaS vs. on-premise value conundrum.
Overall Gartner findings are good read for customers considering SaaS solutions, and more importantly for SaaS vendors to reflect on their offerings and identify opportunities for further improvements.
Showing newest posts with label Project Management. Show older posts
Showing newest posts with label Project Management. Show older posts
Sunday, March 1, 2009
Insight to Gartner SaaS Fact Checks
Friday, September 26, 2008
Estimated Cost of Bailout: $700 Billion; Return of Investment: TBD
In Enterprise IT world no IT project gets funded without a formal Return of Investment analysis. Even in very beginning phases of project, often called initiation/qualified idea phase, a very high-level business case including ROI & risk assessment is predicament to fund the resources required for further definition. It is not uncommon for large Enterprise Initiatives to get approval for the overall budget soon after the ROI justification, however funding of budget will be done in a phased approach, only upon the successful conclusion and review of deliverables from earlier phases. In contrast, operational costs are recurring cost of business operations and do not necessarily require extensive ROI analysis for funding.
On corollary, most fiscal government policies are enacted into law and generally do not require additional approvals. Chairman Bernanke sometime back clarified that the proposed bailout plan is not a fiscal plan instead it’s a market-focused initiative that buys (and sells) assets using market-focused pricing strategies such as options. While everyone agrees that something need to be done to help the economy, market-focused initiative of this size demand clear quantification of associated costs, risks and value. However, these are the missing pieces in this puzzle currently.
Blogger H.J. Huneycutt wrote “It's a mystery to me as to why the concept of ROI is constantly evoked in the business world, yet almost never used when evaluating government policies. If the government can complete a task more efficiently than the private sector, then by all means it should do so. If it cannot, it should keep away.”
While no detailed ROI assessment in testimony, below are some interesting estimates from bloggers on cost of doing/not doing this.
Calculated Risk: “Price is still the key. Since Treasury doesn't plan on churning the $700 billion, the losses will be a portion of the amount invested - and the losses depend on how much Treasury pays (or overpays!) for the assets. The losses are unknowable at this point, but probably in the zero to $300 billion range”
Brett Arneds in his WSJ Blog: “The Federal government pays just 4.34% interest on long-term, 30-year loans. So the government could borrow this money for 30 years at a cost of just $30 billion in interest per year. Let's take a worst case scenario. Let's imagine Uncle Sam borrows $700 billion to buy these assets and never gets a single penny of it back. Let's imagine this paper ends up completely worthless. So instead he has to tap taxpayers to pay off part of the principal every year for 30 years, until the loan is all redeemed. How much would that cost per year? Try $42 billion. That's the interest and principal repayment.
That's less than one-third of 1% of our annual gross domestic product. That's the true, annual cost of this bailout.”
On cost of doing nothing (Benefits):
Washington Post: "If nothing is done, the potential for these markets to seize up in a big way is definitely there," said Frederic S. Mishkin, an economist at Columbia University who was a Federal Reserve governor until last month. "When you look at the history of these crises, when things spin out of control, the cost to fix it later goes up exponentially."
CNN Money: "Economists say that without a restoration of credit, unemployment would likely shoot up to over 10% from 6.1% today. And GDP could fall at an annual rate of between 2% and 4%"
The recent debacle of AIG, Lehman and other financial institutions providing profound evidence for the urgent need for detailed planning & risk management, it is imperative to have a detailed plan of action in place before expending valuable tax payer money.
On corollary, most fiscal government policies are enacted into law and generally do not require additional approvals. Chairman Bernanke sometime back clarified that the proposed bailout plan is not a fiscal plan instead it’s a market-focused initiative that buys (and sells) assets using market-focused pricing strategies such as options. While everyone agrees that something need to be done to help the economy, market-focused initiative of this size demand clear quantification of associated costs, risks and value. However, these are the missing pieces in this puzzle currently.
Blogger H.J. Huneycutt wrote “It's a mystery to me as to why the concept of ROI is constantly evoked in the business world, yet almost never used when evaluating government policies. If the government can complete a task more efficiently than the private sector, then by all means it should do so. If it cannot, it should keep away.”
While no detailed ROI assessment in testimony, below are some interesting estimates from bloggers on cost of doing/not doing this.
Calculated Risk: “Price is still the key. Since Treasury doesn't plan on churning the $700 billion, the losses will be a portion of the amount invested - and the losses depend on how much Treasury pays (or overpays!) for the assets. The losses are unknowable at this point, but probably in the zero to $300 billion range”
Brett Arneds in his WSJ Blog: “The Federal government pays just 4.34% interest on long-term, 30-year loans. So the government could borrow this money for 30 years at a cost of just $30 billion in interest per year. Let's take a worst case scenario. Let's imagine Uncle Sam borrows $700 billion to buy these assets and never gets a single penny of it back. Let's imagine this paper ends up completely worthless. So instead he has to tap taxpayers to pay off part of the principal every year for 30 years, until the loan is all redeemed. How much would that cost per year? Try $42 billion. That's the interest and principal repayment.
That's less than one-third of 1% of our annual gross domestic product. That's the true, annual cost of this bailout.”
On cost of doing nothing (Benefits):
Washington Post: "If nothing is done, the potential for these markets to seize up in a big way is definitely there," said Frederic S. Mishkin, an economist at Columbia University who was a Federal Reserve governor until last month. "When you look at the history of these crises, when things spin out of control, the cost to fix it later goes up exponentially."
CNN Money: "Economists say that without a restoration of credit, unemployment would likely shoot up to over 10% from 6.1% today. And GDP could fall at an annual rate of between 2% and 4%"
The recent debacle of AIG, Lehman and other financial institutions providing profound evidence for the urgent need for detailed planning & risk management, it is imperative to have a detailed plan of action in place before expending valuable tax payer money.
Subscribe to:
Posts (Atom)
Posts
