Cost of Services at UVa – Virginia Evans
The homegrown beginnings of TBM at UVa. Today approx. 35% of central IT budget charged back to units, whether that’s fee for service or a head count fee. Next year will be fully allocated out under new budget model.
Cost of Services Initiative: Central IT only; early discussions with Indiana & Cornell; Modeled with WTC on Indiana’s activity-based costing in 2011; continuing simplification.
181 cost categories, 50 lines of business likd “Hosting Services”, Network Connectivity, etc. Coordinated with service catalog.
Methodology: All coss (people time (72%of cost), other than personal services). Allocated to services for direct and indirect costs (shared costs and overhead).
Collect staff time through managers – once a year now.
Benefits so far: Awareness of all service costs and dependencies; streamlined rate setting; helpful for sourcing decisions (Help Desk, Cloud IaaS); Streamlined responding to surveys like Core Data, Hackett Consulting Engagement; Enabled greater transparency with constituents (schools, advisory groups, etc).
Limitations/Challenges: Takes a lot to maintain; Accuracy of time reporting; Capital costs hit budget in year they’re expended not over life; One time project costs hard to allocate; Benchmarking/Comparability – partial service costing makes it difficult. E.g. ERP costs are not just in central IT, so how do you pull it together?). Also how institutions budget (fringe benefits, power, space, etc).; Managing costs of services that go across org boundaries – how do you get managers to think about lowering costs – what’s good, what’s bad? Using data to manage is next step.
Possible Future uses: Benchmarking (what services? common tool?);
Jim Phelps – Washington
There’s a Technology Business Management Council – includes lots of big companies. They have a framework for what you call services, which Aptio can track and then you can benchmark. You can plug Aptio into CMDB and service catalog, or you can add data manually. Aptio has an integration with ServiceNow. One of the values of Aptio is that you are forced into using their definition of service towers.
There’s discussion about the value of understanding costs vs. the value of benchmarking. Several people have tried to benchmark, but it’s hard, and may even be impossible.
The data and process architecture side of this is not trivial.
Does this force simplification of service catalog? Need to keep service catalog in sync with costs of services. Services role up into portfolios.