Re: Will the math add up as a customer?
I believe consumption based charging models to be THE approach to procuring storage (and a number of other services) as the mindset and culture of IT changes and evolves towards Cloud and the consumption models that it brings. Business units, customers and consumers expect to be able to simply want and have their IT services on demand. Companies need to be agile and able to flex to demand in order to keep winning new customers, and to retain existing customers who will inevitably benchmark the market place and are ever more able to simply take their business to your competitors.
Many organisations will utilise CAPEX and will like the idea of "owning" and asset. On the books, it looks to the accountants that the company has a greater worth as the assets are taken into consideration. In reality, that asset will need to be depreciated over a defined working life (e.g. 4 years). That depreciation is then reported across that 4 years as part of the capital budget. If that business is doing well and within that 4 years a need to scale out occurs, adding to the platform starts to add complexity with the accounting of the capital expense; ie do you coincide depreciation of capacity expansion to coincide with the 4 year planned end-date of the original investment (thus making the expansion more expensive on a per month basis against your budget), or do you start a new 4 year depreciation against that specific purchase (and if so, if you decide to replace the original asset after month 48, how do you handle depreciation remaining on the now isolated expansion that is no longer required) ? OPEX does not go away either. Many organisations will purchase an uplift to the standard warranty for support. If the warranty is 3 years, and your asset depreciation is 4 years, all is well for years 1 to 3. However, year 4 hits, you receive an OPEX bill for the full maintenance rather than an uplift over a warranty, and your perceived run-rate jumps for the device. There is also a level of complexity with some of the storage being still under warranty if expansion was purchased later in that initial 48 month depreciation cycle. You are also paying maintenance and depreciation against the full compliment of capacity installed from Day 1, rather than being charged for what you are actually consuming. You may well use that capacity and ideally run at a consumption rate that is efficient for maintaining the asset under a CAPEX budget, but what is often missed in the calculations at time of purchase, is the reduction in consumption as services and systems are migrated across the last 6 to 12 months of the device lifecycle to the latest and greatest infrastructure platform that comes in to replace it. If you replace the original device before the 4 year depreciation completes, you have to consider and account for the costs of the depreciation not realised against your new investment.
Now assume you have 100 such identical devices on CAPEX, purchased at different intervals, some of them expanded after a couple of years, some of them under utilised, some of them dedicated to a specific purpose or business line due to regulatory needs, some of them out of the initial warranty...... and now try and build your budget plan for the following year with little information on how the business might expand and having to factor in the unknown's of what may be the next big business win that requires a sudden hike in consumption. It's a minefield of complexity, expenditure peaks and troughs, capacity and performance risk, required outages for maintenance/expansion, arguments with vendors on support levels and response times......... the list goes on. Storage then becomes a major battleground for funding of a new project, and is often the roadblock to progress since the CAPEX costs involved tend to require a lengthy period of justification before gaining approval against the initial investment risk. Lastly, what else could that initial capital spend be used for that adds value to the company ?
The right type of subscription model simplifies this immensely. Everything supported to the same level. Costs known and easily calculated on a "price x quantity" unit cost basis. Knowing that as you grow, contract or evolve you can simply add capacity and scale as required to maintain performance expectations. Its consistent, easy to plan and articulate, and becomes a genuine running cost for the business (with the right utility model and contract in place). Your chosen vendor has a vested interest in ensuring the service runs well - they take a risk with the initial investment and installation services that they need to recover through the monthly subscription. A vendor will not dump excessive volumes of infrastructure upon you as that is their risk and their overhead that has to be carefully managed to ensure you have the required flexibility in consumption that allows you to not worry about the capacity availability, but that isn't placing volumes of expensive equipment on a datacentre floor to not be used and to therefore not generate any revenue and become THEIR depreciation headache that undermines the profit and their utility services business model.
Like Cloud, consumption based utility models may not be the cheapest method when measured across 4 or 5 years, but they can add huge value in terms of agility, flexibility, predictability (of cost) and in enabling a business to plan ahead or even get a new major project up and running before that project generates an income at minimal initial investment. Having those advantages is likely to increase revenue from new opportunities and winning new deals with a confidence to be able to meet commitments.
(I am the Global Storage Engineering Manager for a major company and consumer of IT services. I am not affiliated in any way with an IT services or infrastructure vendor/supplier/channel partner.)
Biting the hand that feeds IT
