IT Chargebacks: Yes or No?
Chargebacks or not, the real conversation settles around value. Value is a balance of costs and service. If you deploy chargebacks, simplicity is key.
I am a proponent of elegant, yet simple chargebacks tied to service levels. Costs and services are adjustable levers that create the value defined by the organization.
Having served community hospitals, academic juggernauts, and for-profit health systems, I’ve employed a variety of methodologies. Every institution is unique, with culture, strategy, profitability, and leadership all playing a role in determining best practice and overall value. There is no one-size-fits-all answer.
Gartner identifies seven common chargeback approaches. They do a great job of summarizing the pros and cons of each. I will skip the detail.
I avoid approaches that have significant maintenance costs associated with the operational mechanics. At one organization, I spent significant resources maintaining, defending, and negotiating our detailed chargebacks. We were able to charge back to the end user level, but with 300+ applications and associated hardware, we hit a point of diminishing returns. From an academic perspective, this looked beautiful, but costs eroded the benefit.
At another organization, we applied no chargebacks. Demand became insatiable. This led to the tragedy of the commons. All parties were unhappy. When demand went unchecked, services were perceived as free and IT costs spiraled out of control. Although clearer than a blue sky, the cause and effect cycle still became irrelevant due to a lack of associated sacrifice (cost to business unit) and defined value. Normally, a commons environment renders effective governance models impotent.
My preference: a simple model where costs are correlated with service levels at an enterprise level. Take total IT operational expenses and allocate costs on a single, rational, and easy to measure metric, i.e. the number of employees or number of end-user devices. Despite the limitations of these metrics, they are measurable and have a basis in logic. I do not go to the application level (per drink) or hardware cycle methodologies, where the complexity increases exponentially.
In the fictitious example below, the allocation of $1,000 per user comes with a negotiated service level of 4 on a 1-5 Likert scale. The service level is measured and the performance reported routinely. If the enterprise demands a higher service level, then there will be a negotiated fee increase. If the enterprise demands a price decrease, a revised service level is negotiated. Therefore, the service level becomes the lever for any discussion related to IT costs for the same basket of goods.
This model works well for demand management and governance as well. If the enterprise wants to make IT-enabled investments, the business case is developed. The business case includes a section on IT costs. In continuing the fictitious example, let us assume the enterprise wants to add an EHR at a cost of 25M annual operating expense. Using the model, IT costs increase from $1,000 to $1,250 per user at the current service level of 4.
During the EHR approval process, the enterprise understands and accepts that IT costs per user will increase and the allocation goes up. The elusive “business and IT convergence” is enabled at this intersection where strategy is executed knowing IT implications to include costs and service levels.
|ORGANIZATION OPERATING PRE EHR||$100,000,000|
|COST OF IT/SERVICE LEVEL 4||3%||$3,000,000|
|END USERS/COST PER||3,000||$1,000|
|HOSPITAL A USERS/ALLOCATION||1,000||$1,000,000|
|HOSPITAL B USERS/ALLOCATION||1,500||$1,500,000|
|TOTAL COST IT||$3,000,000|
|ORGANIZATION OPERATING POST EHR||$125,000,000|
|COST OF IT/SERVICE LEVEL 4||3%||$3,750,000|
|END USERS/COST PER||3,000||$1,250|
|HOSPITAL A USERS/ALLOCATION||1,000||$1,250,000|
|HOSPITAL B USERS/ALLOCATION||1,500||$1,875,000|
|TOTAL COST IT||$3,750,000|
To execute this model, you must nosse te ipsum. You must be well organized and have a published service catalog. You must know your costs and be able to measure and report your service levels.
Most importantly, you need to execute to those agreed-upon services and levels while keeping your costs in line with your enterprise commitment. There is the shared value.
No perfect solution exists. Work with your enterprise leaders. Pick something. Experiment. Adjust. You will know when it is working. Until then, keep refining.
(I will go into service levels as they relate to chargebacks in a subsequent post).
Ed Marx is a CIO currently working for a large integrated health system. Ed encourages your interaction through this blog. Add a comment by clicking the link at the bottom of this post. You can also connect with him directly through his profile pages on social networking sites LinkedIn and Facebook and you can follow him via Twitter — user name marxists.