Most technology providers have transitioned to offering software-as-a-service (SaaS) or cloud solutions. This means customers need to assess the impact of cloud vs on-premises before making an upgrade or purchase decision. TFP analysts support such pursuits and often prepare total cost of ownership (TCO) or return on investment (ROI) analyses that demonstrate the value of cloud economics. The results can be illuminating.
What is the Difference Between Cloud and On-Premises Software?
Most of these assessments involve software-as-a-service where the key difference is how the software is deployed. With SaaS, the business application lives on the vendor’s servers and the end-user accesses it using a web browser. On-premises software (also commonly referred to as on-prem or on-premise software) delivers a similar service but is licensed software deployed on servers in the customer’s data center or on end-user computers.
All of the same components (compute, network, data, storage, etc.) still exist, but they’re deployed at different locations.
On the surface, this makes comparing cloud vs on-premises applications seems like a simple request. Cloud solutions are typically subscribed to, based on usage. While on-premises solutions require a software license, vendor maintenance, hardware, networking, and IT support. Do the math for the number of users over a reasonable number of years, then make the call. Simple, right?
Not so much.
The Challenge with Comparing Cloud Economics to On-Premises Software
A litany of factors can easily make this a complex analysis or derail a cloud move based on one factor (like functionality or data security). Understanding and discussing the broader set of factors upfront can both speed up the analysis and, more importantly, the sales pursuit and buyer purchase.
For example, I recently worked the math on a cloud solution for a large enterprise that had the IT discipline of a profitable software company. After our analysis, it became clear that the basic cost savings hypothesis was flawed. There were only a few business impacts that could deliver real value. As painful as it was for my client, the value of moving to the cloud wasn’t there. The good news was that the analysis helped both parties move on from the “big opportunity” to a couple of smaller, but more meaningful areas of collaboration.
Conversely, I also worked the math on a large enterprise that was deep in technical debt (lots of incremental changes to an antiquated architecture) and had new management desirous of business agility. The math suggested that they couldn’t move to the cloud fast enough. It also helped to identify the lowest hanging fruit so they could prioritize the migration roadmap.
With that said, let’s take a (mostly) non-technical look at the factors you should consider when comparing the costs and benefits of cloud software or SaaS vs on-premises software. I’ve segmented these factors into the basic costs that most anyone would consider vs additional business considerations that can make or break an analysis. A more technical look could consider the virtues of various outsourcing models, private clouds, public clouds, and hybrid clouds. But we’ll reserve that discussion for another time.
Basic Cost Factors
The first set of factors pertain to the acquisition and deployment costs of the given software application. These are the basic components of cloud economics that are most commonly included in TCO analyses that compare SaaS to on-prem.
|Software||SaaS fees are typically paid monthly or annually based on peak concurrent use for the given period or as a fee for a given not-to-exceed volume.||A perpetual or term license for the peak number of users (or other metric) is purchased upfront or via an annual subscription.|
|Hardware||The server, most commonly a virtual server (aka virtual machine) hosted on the vendor’s (or outsourced partner’s) premises, is included in the cloud subscription.||Unless the software is only deployed on the end user devices, servers or virtual servers live in the customer’s data center(s). On-premises servers are purchased, maintained, and refreshed at a separate cost.|
|Network||Users access the software via a web-browser, so the cost of internet bandwidth as it pertains to concurrent usage of the application (and other SaaS applications) should be considered.||If access to the software is remote and through an intranet, then you might need to resize your routers at a cost.|
|Storage||Storage of production data or files (and backup) associated with the application is typically included or at least included up to a specified amount as part of the subscription.||Application files typically live on the user’s client or in the data center. Production and backup will impose additional costs if this is a high storage volume use case.|
|Deployment||SaaS can be plug and play and may just involve setting up user IDs and passwords. However, an integrated application can be much more complex and involve people, process, and technology integration costs.||On-premises deployment may be plug and play at the individual laptop or desktop level. But it can be much more protracted if you need to procure infrastructure (servers, storage, networking) and if you need to physically touch the end user devices to deploy the solution.|
|Implementation / Integration||For some SaaS, deployment = implementation. More often, though, there can be integration challenges. We’ve seen as short as less than a month to as long as a year, depending upon the level of integration, customization, and process change.||On-premises implementation and integration tends to take longer, be more costly, and is certainly much more IT-centric, so potentially victim to any IT bottlenecks.|
|Support||Tier 1 support is often handled by the customer’s support team or self-service. Basic tier 2 support is typically included in the subscription with options for premium+ support offered at an additional cost.||Tier 1 is mostly handled by the customer’s support team and some self-service is available. Support contracts are often required and available for a variety of service levels and costs. Network and hardware issues, however, are the responsibility of the customer’s IT team.|
|Updates / Maintenance||Cloud providers schedule and notify their customers regarding periodic updates or updates occur seamlessly at no additional cost or as part of a support contract (see above).||On-premises IT teams juggle the myriad of application updates (typically included in the support contract) and sometimes need access to all end-user clients to execute at the cost of labor, although this is more commonly automated and virtual.|
|Upgrades||Similar to updates, SaaS providers notify their customers regarding significant upgrades and may allow the customer to stay on a current version for a period of time if substantive integrations or customization exists.||Cost and complexity are the main reasons for delaying on-premises software upgrades. Due, in part, to customization, upgrades can come with a heavy cost in license and re-integration. Most TCO analyses of cloud vs on-premises factor such costly upgrades.|
Business Impact Factors
Now that we’ve reviewed the basic components of cloud economics, let’s look at factors that pertain to business impact. These factors offer more insight for executives who are looking to make sound business decisions. But they are often overlooked when creating comparisons.
|Business Agility||SaaS may be less flexible and less customizable, but is this a bug or feature? By offering self-service for reporting, analytics, configuration, and access (in-the-box customization), many cloud providers promote business agility. This can be tied to time-to-market acceleration for revenue growth and cost savings initiatives.||If “traditionally” on-premises with complex code-level customized processes and little or no automation, IT may simply be a bottleneck and impediment to upgrades and innovation. However, automation, private cloud, and hybrid cloud solutions are, in many cases, keeping IT relevant … and agile enough to meet the speed of business.|
|User Productivity||Many SaaS applications include the most current version, easy to follow user interfaces, and delivery from multiple data centers to reduce the latency caused by the distance between the application, the data, and the end user (see data gravity). Also, the fact that access is web-based means anywhere/anytime/any device access (including mobile friendly interfaces). Ease-of-use, performance, and accessibility yield higher productivity.|
I recently calculated a significant productivity gain for a group of developers in India that were accessing software via a VPN to a London data center. Moving to the cloud meant a ~10% productivity gain as the vendor’s data centers were better located.
|Many larger enterprises mimic the cloud experience via workspace applications that include single-sign-on and access to a unified set of applications. However, the proximity of the user to the data center can still impact latency and the need to access through a VPN can limit usability.
Companies that haven’t modernized may have less friendly legacy applications and multiple versions to contend with. This impacts user and IT productivity.
|User Onboarding / Removal||Many cloud applications put this activity in the hands of the business users—simple user interface, simple process of setting up or removing user IDs and passwords.||On-premises solutions may necessitate IT’s involvement and can result in delays. Time-to-onboard a new hire has become a critical factor for hiring and retaining new talent.|
|Reliability||SaaS providers must be reliable and must avoid outages—their recurring revenue relies on this. We have one SaaS client whose name is literally Five9, implying 99.999% uptime.|
Reliability is a prime reason to consider vendor reputation.
|On-premises solutions require their IT teams to architect solutions that do not fail. However, having worked on countless ROI analyses, it is rare to see on-premises at three 9s (99.9% uptime). Downtime can impact revenue, user productivity, and certainly comes with an IT cost.|
|Scalability||Scalability often just involves adding more user ID/password combinations (and ensuring bandwidth).||On-premises scalability may involve provisioning more hardware, more routers, etc., plus acquiring additional licenses—so IT and procurement heavy.|
|Control / Regulatory Compliance||The customer generally relinquishes control regarding cloud applications (and the data therein when downtime occurs).||The customer generally retains control for on-premises applications and data accessibility. In highly regulated industries or where data privacy is paramount (e.g., HIPAA compliance), this factor alone still has some companies choosing on-premises.|
|Security||Security was a huge initial deterrent for moving to the cloud—loss of control, internet access, a lack of security solutions. Much of that has changed and many now consider the level of security to be the same under either scenario. And some consider cloud security to be better—better technical expertise, physical separation, automated backup, and newer technology.||On-premises security is a mix of physical and technological—need to watch the front-doors and the “back-doors” as it were. Like much of this cloud vs on-premises comparison, it can come down to whether or not the company desires to have IT be a core competency and if they are willing to invest in current technology.|
|Functionality||This may fall into a number of the other factors but making an apples-to-apples comparison is paramount. When considering a move to the cloud, if the SaaS solution lacks current functionality, then that may be a con to moving to SaaS. If the SaaS solution provides more functionality, include the benefits of that functionality in your analysis.|
Other Business Factors
If you’ve had enough with factors, skip to the next section. What follows are some additional business factors that can impact a cloud vs on-premises software negotiation.
|Budget (OpEx vs. CapEx)||Cloud is generally considered an operating expense.||On-premises was traditionally considered capital intensive, but usage-based models and short-term subscriptions have changed this.|
|Cost Accounting (End-User)||True SaaS/cloud use results in fairly straight forward cost accounting—expenses are realized as the service is used. Having said that, if any substantive implementation costs are accrued, recent accounting rules have reinforced that such costs can be capitalized like software development costs.||On-premises costs have become murkier, especially with the addition of on-premises subscriptions. Perpetual licenses and hardware are still typically capitalized, but subscriptions were being expensed. Now, new accounting rules are suggesting that multi-year subscriptions be treated like perpetual or multi-year termed licenses and capitalized as well.|
|Financial Statements Impact||Because SaaS goes directly to operating expenses on the income statement, there is no balance sheet impact. And, the cash flow statement impact is part of operating activities. (Implementation costs may be capitalized.)||When on-premises software is a perpetual license it flows through the cash flow statement as a capital expenditure, lands on the balance sheet as property, plant, and equipment, and is then amortized over time, hitting the income statement as part of depreciation and amortization.|
|Key Financial Metrics (e.g., EPS, ROA, EBITDA)||Companies that look at the bottom line--net income, EPS, FCF (e.g., insurance, retail, manufacturing, hi-tech, and energy)—may be okay with an OpEx purchase model as long as the total cost of ownership (TCO) favors that approach.|
Note: Industries can favor multiple metrics, so it’s important to consider the specific company’s given KPI.
|Companies sensitive to increasing EBITDA (e.g., hi-tech and hospitality) may greatly prefer a CapEx purchase model as depreciation and amortization are not factored in EBITDA.
Also, companies focused on ROA (e.g., manufacturing, retail, finserv, and utilities) may also prefer a CapEx model if they believe that a core competency is getting the most return out of their assets.
|License/Usage Compliance||There is no license. Advanced SaaS/cloud providers rarely have usage compliance issues as usage is more easily monitored by the vendor.||On-premises solutions are notorious for having instances of license sprawl and a lack of clarity regarding current use. This often creates a sales opportunity but can also result in friction between the customer and the vendor.|
|Shared Services / Shadow IT||Cloud solutions can make it easier to monitor usage for shared services and internal cross-charges. They can also make it easier for business units to directly acquire technology outside the purview of IT (aka shadow IT).||On-premises solutions have been more challenging to manage from a cross-charge/shared services vantage. On-premises shadow IT tends to be limited to one-off or small deployments of desktop software.|
|Revenue Recognition (Vendor)||True SaaS revenue is typically recognized like service revenue, as the service is delivered.|
Note: The salesperson may optimize the subscription term based on their compensation plan.
|Perpetual license software revenue is typically recognized upfront. New rules have multi-year subscription software being recognized upfront. If the subscription includes support, then companies are extracting a portion of the multi-year subscription as upfront revenue.
Note: A perpetual license sale more readily lends itself to a volume purchase.
Quantifying the Difference Between Cloud and On-Premises Software
Given this rather exhaustive list of factors, it should be clear why a simple comparison of upfront purchase costs is not enough to justify a purchasing decision. Especially when you’re making a business case for an investment that will impact many areas of the business for years to come. You can use the above set of factors as a checklist to ensure that you’re putting together a proper business case when comparing on-premises software to SaaS. But first, do your homework to ensure that you understand the customer’s business strategy and investment goals.
I recently wrote two relevant blogs on this topic. In TCO vs ROI: Which is Best for Making Investment Decisions, I compare a simple TCO analysis to a complete ROI analysis where I pull in the business benefits. This post would be helpful if you need a concrete example (with numbers) of why it’s important to perform a comprehensive analysis.
In the second blog, How to Do an ROI Analysis in 4 Steps, I lay out the process (in great detail) for quantifying the value of a technology investment. These steps would also be applicable here.
Whether you’re comparing competitive technologies or considering the advantages of cloud vs on-premises solutions, discovery is critical. During discovery, you work to gain a clear understanding of the customer’s current state and their desired future state. Then you identify the metrics that will demonstrate that the given solution(s) are getting you from point A to B. You can use the factors listed above as your guide.
For the metrics gathering step, whether you’re comparing cloud to on-premises for the future state (or vice versa), gather the multi-year basic cost factors (aka “metrics”) for each solution. Some may only look at three years, but I prefer to at least look at five years. Then think through the business impact factors for each solution and add them to your comparison. If there’s more (or less) functionality in the future state solution, add that into the mix of quantifiable benefits.
If you’re more intently looking at a move to the cloud, the on-premises column may provide insight into possible current state challenges. While the cloud column will help you understand the ramifications (both good and bad) of making the change.
As you pull together your business case or cost comparison, the collaborative analysis step (working with the vendors or, if you’re selling, working with the prospect) keeps assumptions credible and ensures that the analysis is complete.
The final step, executive presentation, is simply a professionally delivered business case that provides clear justification (using C-suite language) for the recommended purchase.
The Bottom Line for Comparing Cloud vs On-Premises Software
Many technology companies are becoming SaaS or at least SaaS-y solution providers. This means learning curves are in play. This makes it critical, especially if you’re the buyer, to do the math to ensure that the multi-year value of cloud computing is there. Just because “everyone” is moving to the cloud, doesn’t mean you should move everything to the cloud or that new pricing is appropriate.
The bottom line for cloud vs on-premises is to consider all the factors of the change when compared to your current state. This will allow you to more completely quantify the multi-year impact of either approach and make the best decision for your company’s bottom line.