Cost-Benefit Models WEBINAR : Digital Preservation Cost Models and Sustainability

1. Cost-Benefit Models WEBINAR: Digital Preservation Cost Models and Sustainability Digital Preservation Cost Models and Sustainability 13 June 2013 Kirnn Kaur, kirnn.kaur@bl.uk, British Library

2. Cost-Benefit Models: Introduction • Cost-benefit models research objectives • Cost models analysed • Costs vs benefits • Research – approach and methodology • Cost model analysis findings • Organizational infrastructure • Digital object management • Security risk management • Testing of cost models – overview • Survey results • Summary

3. Cost-benefit models research objectives The objectives were to evaluate and test cost models for the preservation of digital objects: • Cost parameters Review of published cost models and their cost parameters, to show how cost model parameters are concentrated or where areas of activity are not covered • Cost data collection and testing Collect cost information to test published cost models • Survey Current practice and requirements within research libraries (LIBER members)

4. Models analysed 1 of 4 CMDP - Cost Model for Digital Preservation • Estimates the costs of digital preservation (ingest, preservation planning and migrations, and archival storage), covers cultural heritage organisations • Still under development, tool available • Available on-line http://www.costmodelfordigitalpreservation.dk/ • Developed by the Royal Library of Denmark and the Danish National Archives DANS cost model • Calculates the costs of archiving datasets, based on activity based costing and balanced scorecard, covers research data archives • Validation to be undertaken • Paper published on the model http:/www.springerlink.com/content/v3r1282x328m607m//?MUD=MP • Developed by DANS, Data Archiving and Network Services, Netherlands

5. Models analysed 2 of 4 DP4lib - Digital Preservation for libraries • Calculates costs by a service model for long term preservation services to third parties, covers any sector • Validation taking place this year • Paper published on the model http://aparsen.digitalpreservation.eu/pub/Main/CostModels/DP4lib-Cost-By-Service-CostModel.docx • Developed by the DNB KRDS – Keeping research data safe (KRDS + KRDS 2) • Provides lists of benefits and potential metrics for research data, is applicable more widely. • Toolkits - benefits analysis, value and impact - for proposals, evaluation and planning • Published factsheet, user guide http://www.beagrie.com/krds.php • Development of toolkits funded by JISC partners in project include Charles Beagrie Ltd, UKOLN, DCC, UCL, UKDA, ADS, OCLC

6. Models analysed 3 of 4 LIFE3 – Life Cycle Information for E-literature • Looks at long-term costs of digital preservation for DP repositories • Third phase of the LIFE Project producing a predictive costing tool (not developed fully), excel version is available for use • Published excel tool and papers http://www.life.ac.uk/ • Developed by UCL and BL, project funded by JISC and RIN Presto PRIME – cost model for digital storage • Provides cost information and long term forecasting for mass digitisation of AV materials • Tools available and still under development • Published reporthttp://prestoprime.it-innovation.soton.ac.uk/planning-tool/accounts/login?next=/planning-tool/ • Developed within EC FP7 project http://www.prestoprime.eu/

7. Models analysed 4 of 4 ENSURE project • Estimates costs of digital preservation activities, assumes cloud storage is used, covers healthcare, clinical trials and financial sector, may be extended to manufacturing sector • Initial model to be developed further • Paper published “Towards a cost model for digital preservation” http://epubs.stfc.ac.uk/bitstream/7711/Towards%20a%20Cost%20Model%20for%20Long%20Term%20Digital%20Preservation.pdf • Being developed by EC FP7 project, ENSURE (Feb 11 – Jan 14) http://ensure-fp7-plone.fe.up.pt/site ISIS facility model • Applied specifically to long term preservation costs of data from ISIS facility at STFC (scientific research data) • Not applicable to other areas • Poster published http://ensure-fp7-plone.fe.up.pt/site/Poster.pdf • Developed as part of Cranfield University MSc project in collaboration with STFC

8. Costs vs benefits Cost models in digital preservation tend to focus on measuring activities related to specific processes Cost models reviewed provide a varied approach to measuring costs Generally, benefits are not considered as an element of a cost model. However, two cost models which do consider benefits are: • KRDS - provides benefits framework and value chain and benefits impact tool • DANS - provides Balanced Scorecard methodology and reports on benefits through success factor and performance indicator categories Overall, the relationship between costs and benefits falls within the framework of business cases

9. Approach and methodology Analysis of cost models • To determine scope, access, strengths and weaknesses of published cost models Mapping of cost parameters • ISO16363 to ascertain concentrationof parameters and identify gaps and areas for further investigation and development • Assessing parameters against the activities audited would show whether cost models included specific functional elements of the ISO Testing of cost models • Three cost models selected. Test data was used to carry out cost allocations across the models using various assumptions for apportionment of costs Survey • Online survey, 101 respondents. Results analysed

10. Analysis of cost models: main findings 1 of 4

11. ISO 16363 COST MODELS Analysis of cost models: main findings 2 of 4 • Overview • Some parameter definitions not easily aligned to standard • Some areas of standard not realistically measurable through a cost parameter • Inconsistent coverage of parameters across all areas. KRDS provides best coverage.

12. ISO 16363 COST MODELS Analysis of cost models: main findings 3 of 4 • Consistent coverage across most areas, as would be expected for cost models related to digital preservation • LIFE3 and ENSURE models provide parameters across all areas

13. ISO 16363 COST MODELS Analysis of cost models: main findings 4of 4 • Good coverage across the sections, although not consistent • Every cost model has cost parameters which match at least one of the two subheadings.

14. Testing of cost models – findings to date Overview: Methods of costing activities and cost parameters definitions do not match across models. A common basis for comparison and apportionment of costs wasdifficult to find Cost models tested: • DANS with DP4lib data • DP4lib with DANS data • LIFE3 with DP4lib data Findings to date: • Good understanding of costs as well as their breakdown is needed • Comparison or harmonisation of cost models is very difficult. Minimum conditions: • high level approach needs to be adopted • more detail needed to allow users to apportion costs • assumptions made during creation, testing and validation of cost models need to be explicit

15. Survey results Overview:Limited use of cost models although advantages widely accepted • Funding – cost models provide cost predictions • Management – involve key stakeholders in future development of cost models • Benefits and added value – cost models can provide tools • Controlling costs – very limited use although good understanding of the reasons for use of cost • Cost model requirements – validated, cover the digital preservation lifecycle, easy to use and freely available

16. Survey results: cost model requirementsQn: on what basis would you select a cost model (n=73)?

17. Summary Cost model work within APARSEN • Final report - end June 2013, published on project website • Brought together some of the cost model owners, some still in the development phase of their cost models • Fostered integration in the diverse area of cost modeling • Had a beneficial influence on future development of cost models What next ? • Integration within sustainability topic continues (booklet to be produced in Sept 2013) • 4C project interested in our findings (joint workshop later in year)

18. Network of Excellence