IaaS and depreciation

InsightaaS perspective: This is a bit of a "funny" post. I initially developed it as part of my ongoing series for the TELUS Talks Business blog site. It turned out that the folks there were more interested in having me expound on "migratory paths to the cloud," though - which I was happy to do, since it, too, is a really important issue in cloud. After I submitted the new piece (should be up later this month!), I asked if it would be okay to post this content here, and the folks at TELUS were kind enough to allow me to do so. I'm glad to have the opportunity to put this forward, and would be really interested in your feedback!

When prospective buyers and sellers of cloud services connect, they need to address a number of issues. Buyers need to know that their applications and information will be secure, and that they will have access to data whenever they need it; sellers need to demonstrate that they are sound, long-term partners, and that their fees represent a meaningful savings over traditional approaches to IT.

One issue that often goes underreported in the press — but which is a significant factor for both prospective customers and suppliers — is the impact of depreciation on cloud plans.

Let’s take a sample mid-sized business running on-prem servers, storage, networking and other infrastructure products. Our "X.co" invests $50,000 in infrastructure replacements and upgrades each year, and depreciates each infrastructure item over a five year period, meaning that it also has a $50,000 charge on its books for IT depreciation each year.

A hosting provider approaches X.co, explaining that it can get far more infrastructure for less than $50,000 if it moves to hosted virtual private cloud or IaaS. If X.co agrees that this is the year to move to IaaS and unplug its on-prem infrastructure, it avoids the $50,000 cash investment needed to refresh its on-prem infrastructure. However, there is still the ‘bow wave of depreciation’ to deal with: the books say that last year’s investment is still worth $40,000, the investment from the previous year is valued at $30,000, and the previous two years’ equipment is still valued at $20,000 and $10,000, respectively. If X.co turns off its local gear, it needs to write off all of the infrastructure it is no longer using, which results in a pretty significant accounting hit. X.co might be saving $50,000 in CAPEX cash, but it will have to recognize a $100,000 IT equipment depreciation charge — plus, of course, the fees associated with the IaaS production environment. From a budget perspective, the IaaS solution will actually increase year one costs by more than 100%.

Aha, you say — but what if they, like "Y.co," have been skipping refresh investments over the past 2-3 years? Doesn’t that reduce the depreciation hit? The answer is "yes — but…" They "yes" part applies to deprecation: if Y.co skipped two years of refreshes, its infrastructure book value is now only $30,000. However…this means that Y.co will only incur an accounting charge for infrastructure of $20,000 this year, and $10,000 next year. Will the IaaS service that compared favourably to a $50,000 CAPEX outlay still look like a bargain in this light?

Fortunately, there is a good answer to this conundrum: disaster recovery/business continuity. DR/BC in the cloud is an extremely compelling offering. There is generally a (pretty modest) fee for moving applications to the cloud and synching production data to cloud-based storage. Substantial IaaS charges are only incurred when you use the system — and in the case of a disaster, that’s exactly when you’d be eager to pay in exchange for service!

In the scenarios above, it’s unlikely that either X.co or Y.co has invested in building a parallel facility that can support their users in the event of some type of problem at the data centre — and yet, like virtually all companies, they are increasingly reliant on their infrastructure. Both will realize significant advantages from the establishment of a DR/BCaaS strategy. In Y.co’s case, the benefit is obvious: if you’re planning to use aging infrastructure, it’s smart to have a hedge against equipment failure. X.co benefits from this safety net, and also from the fact that it will have production apps and data in the cloud. If X.co chooses not to refresh infrastructure but instead simply turn off its oldest products at the end of the year, it can move a portion of its production workload to the cloud, maintain a portion on-prem, and avoid the accounting penalty (and realize the service vs. CAPEX savings) while migrating to this hybrid environment. In this way, DR/BCaaS helps buyers, sellers and accountants all see beyond the "bow wave of depreciation" to the value of hosted infrastructure!