At a time of growing IT expenditure and increased accountability for delivering value from IT, executives are asking if it makes sense to run IT as a business. The point is not to force IT to turn a profit but to make IT operate under the same economic constraints as any for-profit company by calling for rational resource allocation.
Yet the reality is very different for, in many cases, IT is still viewed as a cost center with little thought to whether IT is creating value. The case of information storage is especially vexing given the rise in storage spending and a lack of meaningful storage metrics. In this paper, we outline steps that firms can take to treat storage as a service-oriented business, moving IT from being a cost center to a value center.
Pricing is a basic fundamental premise of all rational economic decisions. The lack of pricing information can distort market behavior. For example, up until recently, cars buyers – unaware of car invoice pricing – were inclined to pay well above the odds for even the most basic model.
In a web-enabled world, a buyer with invoice-pricing information in hand can tell a car dealer how much they will spend on a car! Without accurate pricing data, information asymmetry can bias a purchase decision leading to buyer’s remorse (if a buyer overpays) or dealer losses (if prices are too low). Yet in regard to IT decision making, pricing information is conspicuous by its absence – the issue is not information asymmetry but rather the lack of information.
Users rarely know the true price of internally provided IT services and so there is a tendency to over-consume something that is seen as free. This is especially problematic in the case of information storage as users perceive data collection and retention as relatively inexpensive on account of the rapid fall in storage hardware costs and so vital resources are wasted. The firm (not the individual user) suffers as a result. If storage was run as a business, users would have to pay market prices for storage and waste would fall.
The fact that prices are consistently falling, not just in storage but in all areas of IT, is partially the reason why IT is not run as a business. This is like what happens when natural resources (such as drinking water) are plentiful; consumers do not want to pay for something that is seen as plentiful and so IT is reluctant to introduce a rationed services model that forces users to pay for something that they feel should be free.
A further reason is metering. Unless a service can be metered, users do not know what they are paying for or how to control their use. Equally, the CIO does not know what to charge. The only way that this situation can be reversed is if there is a realization that information is strategic, that storage is also strategic since it houses the information, and that storage spending has reached a point where it is stretching the IT budget. The latter is now a reality for many firms and so the time is ripe for running IT as a business.
Transitioning IT from a cost center to a value center will be challenging for many corporations. In Figure 1, we reveal some of the most glaring differences between running IT as a business and running IT as a cost center. CIOs rarely know how much it costs to do even the most basic IT activity because IT is not a job shop. Yet if chargebacks are used, this activity pricing data is vital. Some back-of-the-envelope prices exist for helpdesk activities such as account inquiries or password resets but few such calculations would pass muster if IT was run like a business with its own P&L.
In many cases, IT continues to serve as a cost center and so inaccurate pricing may not be fatal but it will lead to wasteful spending. Certainly, CIOs are accountable for the costs they incur under a cost center arrangement but without accurate pricing for the services performed by the IT function and an appreciation for the changing nature of these costs, there can be little understanding of whether IT is creating value or whether, in the words of Harvard Business Review editor Nick Carr, IT has become a cost-driven commodity and, as such, has ceased to matter.
As a cost center, the focus is on the supply side: bringing added capacity to the user rather than optimizing the resources that the user needs to do their job. It can be an attractive proposition for supply side advo-cates to seek better, faster, cheaper resource provisioning but the issue here is not the size of the IT faucet or the volume of the flow but the quality of the water – that is, whether the underlying IT resources are providing value for money.
Against this background has emerged the idea of running IT as a business. The implication may be that IT can sell its services to outside clients but more likely it is a way for IT to understand how services can be priced in order to ensure value for money. Any such understanding is clearly missing when IT resources and storage resources, especially, are provisioned on a just-in-case basis… buy more storage just in case it is needed.
To take a leaf from Dell’s experience with just-in-time product delivery, the concept of storage on-demand or of provisioning just-in-time has yet to reach the datacenter.
When storage is run as a cost center with capacity as a guiding principle, there is enormous potential for just-in-case provisioning to lead to massive resource underutilization in the form of buffer provisioning. The added fallacy of this approach is looking to utilization metrics to evaluate the performance of the storage environment: utilization is a supply side metric; it says nothing about whether storage utilization is in line with demand or whether IT overall is providing value to the firm.
Even at a time when a growing portion of IT budgets (as much as 70% in some cases) is spent keeping the lights on – there is ample reason to ask how much is being spent on providing various IT activities. What is clear, however, is that the notion of activity-based costing (ABC) has yet to make inroads into IT. ABC has its origins in manufacturing where the cost of each activity is compiled in a systematic manner.
Fixed costs use a predefined formula while variable costs are compiled from actual resource usage levels. When IT is run as a cost center, there is little incentive to compile accurate costing for each activity unless costs need to be charged-back to end-users. Even then, most firms expect a fudge factor, regarding chargebacks as an exchange of funny money rather than an arms length transaction.
Chargebacks do not motivate IT to run as a business or to be most efficient in how IT resources are allocated unless the underlying cost data are accurate and are perceived by users as providing value for money.
The above words, uttered by Intel co-founder Andy Grove, attest to the importance of having information (metrics) that managers can use to make rational resource allocation decisions. As a result, firms like Intel thrive in a metric-rich environment.
Most CIOs are not yet at a point where cost metrics can be computed and shared among key business-level decision makers. The problem may be that costs are not tracked and so are not available to be allocated back to users but the bigger problem is cost apportionment. For exam-ple, how can network or hardware costs be assigned to users?
What is the denominator? Data traffic? File size? We know that there is a basis in history for charging users for CPU cycles but in a networked world, this is clearly out of touch. Instead, the focus is on a differential pay for performance model. If users want sub-second response times, the cost is clearly different than users who are unconcerned with wait times.
If users want a superior storage service to manage critical data, the cost again should be different from the cost of providing a basic tactical service. Some progress has been made in storage with TCO (total cost of ownership) data but this is just the tip of the iceberg. It is also clear that a focus on supply side provision-ing distorts TCO since there is no appreciation of the underlying value generated by IT or the risks facing the firm from excessive efforts at reducing TCO to a more manageable number.
A relatively new idea for how to construct such metrics can be found in firms like Kaiser Permanente, the Hartford Financial Services Group, and Axa Financial. The IT department in each of these firms employs a CFO of IT – someone who is charged with managing IT cost metrics and who is otherwise instrumental in managing IT as a business.
Financial services firms spend over 7% of their annual revenues on IT. If only a small fraction of this spending is wasted but can be recovered through more effective management, a CFO of IT can have a direct effect on firm profitability. If this occurs, running IT as a business can have an immediate impact on the perceived performance of the IS organization and on the firm more broadly.
All of the above practices depend on access to information. Running IT as a business means securing data assets that might be compromised or otherwise incapable of supporting the business.
As many firms have learned, much to their surprise, lost or stolen data can lead to financial loss and damaged reputations. If IT is run as a cost center, there may be little consideration given to securing information from catastrophic IT outages but where IT is run as a business, there is a renewed sense of awareness that information must be protected just like a business would protect any of its physical assets from fire, theft or foul weather; IT is no different.
Protecting information means appreciating the value of information to the firm – and having a way to measure and attribute actual value to the data. As firms are now discovering, information is the starting point for the next strategic advantage. Firms such as Harrah’s and Capital One have had great success from information-based strategies that use data analytics to better understand their markets.
Such success can only come when IT is run as a business with a clear view of how IT creates value. The notion of a physical supply chain with resources passing from source to user is relevant to storage; a successful storage supply chain is key to running IT as a business.
Despite the surge in monitoring activities that followed the implementation of Sarbanes Oxley legislation, CFOs are resolute in calling for IT to maintain and improve service and to return even higher efficiency. To provide this sustained level of service, IT can no longer be viewed as a cost to be minimized.
When IT is run as a business, service performance can be judged according to service costs since users will now be able to appreciate the true cost and value of the IT services they receive. IT is not a free resource that end users can use and waste at will. If IT is run as a business, there will be clear visibility into how IT is used (or misused) and the value that IT provides to the firm.
Nowhere is this more true than in storage. Making the transition from cost center to business partner is not an easy step for storage or for those on the supply side who must come to acknowledge the strategic role of storage; moving away from supply side storage is a necessary step. At a time when the world is competitively flat and business is global, IT matters more than ever. Running IT and storage as a business is a must-have for succeeding in such a world.
This white paper was written by Professor Paul P Tallon of the Sellinger School of Business and Management, Loyola College, Baltimore, for CBR on behalf of GlassHouse Technologies. It outlines the steps companies can take to treat storage as a service-orientated business, moving IT from being a cost centre to a value centre.