Assessing Digitization Costs

[Jody DeRidder]

Over the years, I've worked to assess reasonable speeds and then costs for digitization of various kinds of materials. I think it may be possible to set up a reusable framework for making these estimates, to help in developing grant proposals, projecting project costs, and selecting digitization equipment.

At one point, we measured the costs for creating metadata by metadata librarians as opposed to archival staff (and then did a usability test to determine the extent to which this impacted the users).  We measured the actual time it took for both sets of people to describe the same content in 3 collections (each a different type of material).  (We did *not* try to measure the costs for organization and description of collections, even to the series level.   Do you know of anyone who's tried to do this?  It seems to be an intractable problem.)

When developing cost estimates for digitization requests, I have included the cost of the equipment and software (assuming it lasted 3 or 5 years with a certain level of usage, dividing the replacement cost by that number of hours, then multiplying by the number of hours used in the proposed project).  I also included overhead costs (I used the current percentage our Office of Sponsored Programs uses for this purpose on grant proposals), which includes building maintenance, heating/cooling, etc.  Employee costs had to include benefits costs, not just pay.  Of course, employee costs and overhead costs vary with the institution.

Speed of digitization is not just training or person dependent, it's also hardware dependent.  When I realized a few years ago that overhead capture was so much faster than flatbeds (see p. 153), I did a survey of overhead capture stations to help us select the best ones. Appalled at the costs, my Digitization Manager developed a low-cost alternative desktop capture station.  This certainly lowers our costs, while speeding capture.

More difficult to assess were costs for supervision, training, and communications. ( I don't think this can be standardized, but I may be wrong.)

 

Still, I think that it might be feasible, and possibly very helpful, if we were to develop a matrix into which people could plug in numbers that are appropriate for their institution, and it would output an expected cost for digitizing a specific set of material.   So, for example,  one could input that they'll be using students paid $9 an hour, with no benefits costs, an overhead of 18%, digitizing approximately 3,000  2-4 page letters in good condition on flatbeds that can be replaced for $1500, using Windows 7 stations that (with Excel and Photoshop or Silverlight) can be replaced for $1200.  They could specify entry of X number of metadata fields into the spreadsheet (weighted, as description takes longer), select the extent of optimization/quality control time, and enter in their own number for the time spent in uploading to their delivery system. 

The output would be the expected total costs for digitizing the material they're thinking about putting online.  This could then be used at multiple institutions.  

What do you think?

--jody

[Elizabeth Joan Kelly]

Yes! I would find this very useful.

[Joyce Chapman]

This is an interesting idea, though it would have to be clear that it was a rough estimation tool. A couple years ago I was project manager for a large-scale digitization effort of the TRLN consortium in NC, and we decided to capture time data about everything we did -- from transporting materials between institutions, to removing staples, conducting IP review, scanning, and quality control. We did this because we thought the information would be helpful to others in exactly the way you describe, particularly because we were working with 20th century materials, which are less frequently the content for large-scale digitization due to intellectual property concerns (see our IP strategy here http://www2.trln.org/ccc/ipstrategy.htm). And we imagined there would be significant differences in the amount of time spent conducting some of these activities for 20th century materials versus older materials (for example, perhaps more time on IP review or removing staples). I don't think anything has been done with that data yet; I left the project before it was complete, but I know that all the data is available, and I would be happy to analyze it for this purpose. Though it might take me a while to pull that together. We did not attempt to gather cost/time data for supervision, training, and communications.

Age of materials might be another piece of info folks would be able to input in such a calculator, if it was found to make a difference. Not to mention type of material, though that may be addressed enough by inputting type of scanner (?)

I have not ever tried to analyze cost for processing archival materials, but like you, I once was involved in a research project that gathered time data on metadata creation for finding aids. And like you, we had both catalogers and archival staff doing the work, though we did not have them replicate the work on the same collections. We did find some interesting differences in which sections of metadata each group was able to produce faster. We also collected data from two different institutions for metadata creation time, one in Georgia and one in NC.

I'd be interested to know how much more work there is out there on time and cost for digitization. It would be a fun exercise to try and pull it all together and see if a framework or a loose calculator could be created.

I'd be happy to volunteer to build/code the calculator, if it gets that far.

--
You received this message because you are subscribed to the Google Groups "Digital Library Assessment" group.
To unsubscribe from this group and stop receiving emails from it, send an email to digital-library-ass...@googlegroups.com.
To post to this group, send email to digital-libra...@googlegroups.com.
Visit this group at http://groups.google.com/group/digital-library-assessment.
To view this discussion on the web visit https://groups.google.com/d/msgid/digital-library-assessment/68e80273-4b35-44f5-993e-2ca109ed4509%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.